Search

Deep learning (DL) in orthopaedics has gained significant attention in recent years. Previous studies have shown that DL can be applied to a wide variety of orthopaedic tasks, including fracture detection, bone tumour diagnosis, implant recognition, and evaluation of osteoarthritis severity. The utilisation of DL is expected to increase, owing to its ability to present accurate diagnoses more efficiently than traditional methods in many scenarios. This reduces the time and cost of diagnosis for patients and orthopaedic surgeons. To our knowledge, no exclusive study has comprehensively reviewed all aspects of DL currently used in orthopaedic practice. This review addresses this knowledge gap using articles from Science Direct, Scopus, IEEE Xplore, and Web of Science between 2017 and 2023. The authors begin with the motivation for using DL in orthopaedics, including its ability to enhance diagnosis and treatment planning. The review then covers various applications of DL in orthopaedics, including fracture detection, detection of supraspinatus tears using MRI, osteoarthritis, prediction of types of arthroplasty implants, bone age assessment, and detection of joint-specific soft tissue disease. We also examine the challenges for implementing DL in orthopaedics, including the scarcity of data to train DL and the lack of interpretability, as well as possible solutions to these common pitfalls. Our work highlights the requirements to achieve trustworthiness in the outcomes generated by DL, including the need for accuracy, explainability, and fairness in the DL models. We pay particular attention to fusion techniques as one of the ways to increase trustworthiness, which have also been used to address the common multimodality in orthopaedics. Finally, we have reviewed the approval requirements set forth by the US Food and Drug Administration to enable the use of DL applications. As such, we aim to have this review function as a guide for researchers to develop a reliable DL application for orthopaedic tasks from scratch for use in the market., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 The Author(s). Published by Elsevier B.V. All rights reserved.)

With Artificial Intelligence (AI) increasingly permeating various aspects of society, including healthcare, the adoption of the Transformers neural network architecture is rapidly changing many applications. Transformer is a type of deep learning architecture initially developed to solve general-purpose Natural Language Processing (NLP) tasks and has subsequently been adapted in many fields, including healthcare. In this survey paper, we provide an overview of how this architecture has been adopted to analyze various forms of healthcare data, including clinical NLP, medical imaging, structured Electronic Health Records (EHR), social media, bio-physiological signals, biomolecular sequences. Furthermore, which have also include the articles that used the transformer architecture for generating surgical instructions and predicting adverse outcomes after surgeries under the umbrella of critical care. Under diverse settings, these models have been used for clinical diagnosis, report generation, data reconstruction, and drug/protein synthesis. Finally, we also discuss the benefits and limitations of using transformers in healthcare and examine issues such as computational cost, model interpretability, fairness, alignment with human values, ethical implications, and environmental impact., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 The Authors. Published by Elsevier B.V. All rights reserved.)

In the realm of pulmonary tracheal segmentation, the scarcity of annotated data stands as a prevalent pain point in most medical segmentation endeavors. Concurrently, most Deep Learning (DL) methodologies employed in this domain invariably grapple with other dual challenges: the inherent opacity of 'black box' models and the ongoing pursuit of performance enhancement. In response to these intertwined challenges, the core concept of our Human-Computer Interaction (HCI) based learning models (RS_UNet, LC_UNet, UUNet and WD_UNet) hinge on the versatile combination of diverse query strategies and an array of deep learning models. We train four HCI models based on the initial training dataset and sequentially repeat the following steps 1-4: (1) Query Strategy: Our proposed HCI models selects those samples which contribute the most additional representative information when labeled in each iteration of the query strategy (showing the names and sequence numbers of the samples to be annotated). Additionally, in this phase, the model selects the unlabeled samples with the greatest predictive disparity by calculating the Wasserstein Distance, Least Confidence, Entropy Sampling, and Random Sampling. (2) Central line correction: The selected samples in previous stage are then used for domain expert correction of the system-generated tracheal central lines in each training round. (3) Update training dataset: When domain experts are involved in each epoch of the DL model's training iterations, they update the training dataset with greater precision after each epoch, thereby enhancing the trustworthiness of the 'black box' DL model and improving the performance of models. (4) Model training: Proposed HCI model is trained using the updated training dataset and an enhanced version of existing UNet. Experimental results validate the effectiveness of this Human-Computer Interaction-based approaches, demonstrating that our proposed WD-UNet, LC-UNet, UUNet, RS-UNet achieve comparable or even superior performance than the state-of-the-art DL models, such as WD-UNet with only 15 %-35 % of the training data, leading to substantial reductions (65 %-85 % reduction of annotation effort) in physician annotation time., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 The Authors. Published by Elsevier B.V. All rights reserved.)

Glioblastoma, characterized as a grade 4 astrocytoma, stands out as the most aggressive brain tumor, often leading to dire outcomes. The challenge of treating glioblastoma is exacerbated by the convergence of genetic mutations and disruptions in gene expression, driven by alterations in epigenetic mechanisms. The integration of artificial intelligence, inclusive of machine learning algorithms, has emerged as an indispensable asset in medical analyses. AI is becoming a necessary tool in medicine and beyond. Current research on Glioblastoma predominantly revolves around non-omics data modalities, prominently including magnetic resonance imaging, computed tomography, and positron emission tomography. Nonetheless, the assimilation of omic data-encompassing gene expression through transcriptomics and epigenomics-offers pivotal insights into patients' conditions. These insights, reciprocally, hold significant value in refining diagnoses, guiding decision- making processes, and devising efficacious treatment strategies. This survey's core objective encompasses a comprehensive exploration of noteworthy applications of machine learning methodologies in the domain of glioblastoma, alongside closely associated research pursuits. The study accentuates the deployment of artificial intelligence techniques for both non-omics and omics data, encompassing a range of tasks. Furthermore, the survey underscores the intricate challenges posed by the inherent heterogeneity of Glioblastoma, delving into strategies aimed at addressing its multifaceted nature., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier B.V. All rights reserved.)

Background: Parkinson's Disease (PD) demands early diagnosis and frequent assessment of symptoms. In particular, analysing hand movements is pivotal to understand disease progression. Advancements in hand tracking using Deep Learning (DL) allow for the automatic and objective disease evaluation from video recordings of standardised motor tasks, which are the foundation of neurological examinations. In view of this scenario, this narrative review aims to describe the state of the art and the future perspective of DL frameworks for hand tracking in video-based PD assessment., Methods: A rigorous search of PubMed, Web of Science, IEEE Explorer, and Scopus until October 2023 using primary keywords such as parkinson, hand tracking, and deep learning was performed to select eligible by focusing on video-based PD assessment through DL-driven hand tracking frameworks RESULTS:: After accurate screening, 23 publications met the selection criteria. These studies used various solutions, from well-established pose estimation frameworks, like OpenPose and MediaPipe, to custom deep architectures designed to accurately track hand and finger movements and extract relevant disease features. Estimated hand tracking data were then used to differentiate PD patients from healthy individuals, characterise symptoms such as tremors and bradykinesia, or regress the Movement Disorder Society-Unified Parkinson's Disease Rating Scale (MDS-UPDRS) by automatically assessing clinical tasks such as finger tapping, hand movements, and pronation-supination., Conclusions: DL-driven hand tracking holds promise for PD assessment, offering precise, objective measurements for early diagnosis and monitoring, especially in a telemedicine scenario. However, to ensure clinical acceptance, standardisation and validation are crucial. Future research should prioritise large open datasets, rigorous validation on patients, and the investigation of new frontiers such as tracking hand-hand and hand-object interactions for daily-life tasks assessment., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 The Author(s). Published by Elsevier B.V. All rights reserved.)

Background: Alzheimer's disease (AD) is the most prevalent cause of dementia, characterized by a steady decline in mental, behavioral, and social abilities and impairs a person's capacity for independent functioning. It is a fatal neurodegenerative disease primarily affecting older adults., Objectives: The purpose of this literature review is to investigate various AD detection techniques, datasets, input modalities, algorithms, libraries, and performance evaluation metrics used to determine which model or strategy may provide superior performance., Method: The initial search yielded 807 papers, but only 100 research articles were chosen after applying the inclusion-exclusion criteria. This SLR analyzed research items published between January 2019 and December 2022. The ACM, Elsevier, IEEE Xplore Digital Library, PubMed, Springer and Taylor & Francis were systematically searched. The current study considers articles that used Magnetic Resonance Imaging (MRI), Positron Emission Tomography (PET), APOe4 genotype, Diffusion Tensor Imaging (DTI) and Cerebrospinal Fluid (CSF) biomarkers. The study was performed following Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines., Conclusion: According to the literature survey, most studies (n = 76) used the DL strategy. The datasets used by studies were primarily derived from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. The majority of studies (n = 73) used single-modality neuroimaging data, while the remaining used multi-modal input data. In a multi-modality approach, the combination of MRI and PET scans is commonly preferred. Also, Regarding the algorithm used, Convolution Neural Network (CNN) showed the highest accuracy, 100 %, in classifying AD vs. CN subjects whereas the SVM was the most common ML algorithm, with a maximum accuracy of 99.82 %., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier B.V. All rights reserved.)

Hematology is the study of diagnosis and treatment options for blood diseases, including cancer. Cancer is considered one of the deadliest diseases across all age categories. Diagnosing such a deadly disease at the initial stage is essential to cure the disease. Hematologists and pathologists rely on microscopic evaluation of blood or bone marrow smear images to diagnose blood-related ailments. The abundance of overlapping cells, cells of varying densities among platelets, non-illumination levels, and the amount of red and white blood cells make it more difficult to diagnose illness using blood cell images. Pathologists are required to put more effort into the traditional, time-consuming system. Nowadays, it becomes possible with machine learning and deep learning techniques, to automate the diagnostic processes, categorize microscopic blood cells, and improve the accuracy of the procedure and its speed as the models developed using these methods may guide an assisting tool. In this article, we have acquired, analyzed, scrutinized, and finally selected around 57 research papers from various machine learning and deep learning methodologies that have been employed in the diagnosis of leukemia and its classification over the past 20 years, which have been published between the years 2003 and 2023 by PubMed, IEEE, Science Direct, Google Scholar and other pertinent sources. Our primary emphasis is on evaluating the advantages and limitations of analogous research endeavors to provide a concise and valuable research directive that can be of significant utility to fellow researchers in the field., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier B.V. All rights reserved.)

Accurately measuring the evolution of Multiple Sclerosis (MS) with magnetic resonance imaging (MRI) critically informs understanding of disease progression and helps to direct therapeutic strategy. Deep learning models have shown promise for automatically segmenting MS lesions, but the scarcity of accurately annotated data hinders progress in this area. Obtaining sufficient data from a single clinical site is challenging and does not address the heterogeneous need for model robustness. Conversely, the collection of data from multiple sites introduces data privacy concerns and potential label noise due to varying annotation standards. To address this dilemma, we explore the use of the federated learning framework while considering label noise. Our approach enables collaboration among multiple clinical sites without compromising data privacy under a federated learning paradigm that incorporates a noise-robust training strategy based on label correction. Specifically, we introduce a Decoupled Hard Label Correction (DHLC) strategy that considers the imbalanced distribution and fuzzy boundaries of MS lesions, enabling the correction of false annotations based on prediction confidence. We also introduce a Centrally Enhanced Label Correction (CELC) strategy, which leverages the aggregated central model as a correction teacher for all sites, enhancing the reliability of the correction process. Extensive experiments conducted on two multi-site datasets demonstrate the effectiveness and robustness of our proposed methods, indicating their potential for clinical applications in multi-site collaborations to train better deep learning models with lower cost in data collection and annotation., Competing Interests: Declaration of competing interest DW, KK, LL, C-CS, HW, GZ and CW are employees at Sydney Neuroimaging Analysis Centre, Australia. AN used to be the employee of Sydney Neuroimaging Analysis Centre, Australia when the paper was built. LB and WO are now employees in Shanghai AI Lab, China. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. We affirm that the information provided in this declaration is accurate and complete to the best of our knowledge., (Copyright © 2024 The Authors. Published by Elsevier B.V. All rights reserved.)

Background and Objectives: Generating coherent reports from medical images is an important task for reducing doctors' workload. Unlike traditional image captioning tasks, the task of medical image report generation faces more challenges. Current models for generating reports from medical images often fail to characterize some abnormal findings, and some models generate reports with low quality. In this study, we propose a model to generate high-quality reports from medical images., Methods: In this paper, we propose a model called Hybrid Discriminator Generative Adversarial Network (HDGAN), which combines Generative Adversarial Network (GAN) with Reinforcement Learning (RL). The HDGAN model consists of a generator, a one-sentence discriminator, and a one-word discriminator. Specifically, the RL reward signals are judged on the one-sentence discriminator and one-word discriminator separately. The one-sentence discriminator can better learn sentence-level structural information, while the one-word discriminator can learn word diversity information effectively., Results: Our approach performs better on the IU-X-ray and COV-CTR datasets than the baseline models. For the ROUGE metric, our method outperforms the state-of-the-art model by 0.36 on the IU-X-ray, 0.06 on the MIMIC-CXR and 0.156 on the COV-CTR., Conclusions: The compositional framework we proposed can generate more accurate medical image reports at different levels., Competing Interests: Declaration of competing interest We declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted., (Copyright © 2024 Elsevier B.V. All rights reserved.)

An echocardiogram is a sophisticated ultrasound imaging technique employed to diagnose heart conditions. The transthoracic echocardiogram, one of the most prevalent types, is instrumental in evaluating significant cardiac diseases. However, interpreting its results heavily relies on the clinician's expertise. In this context, artificial intelligence has emerged as a vital tool for helping clinicians. This study critically analyzes key state-of-the-art research that uses deep learning techniques to automate transthoracic echocardiogram analysis and support clinical judgments. We have systematically organized and categorized articles that proffer solutions for view classification, enhancement of image quality and dataset, segmentation and identification of cardiac structures, detection of cardiac function abnormalities, and quantification of cardiac functions. We compared the performance of various deep learning approaches within each category, identifying the most promising methods. Additionally, we highlight limitations in current research and explore promising avenues for future exploration. These include addressing generalizability issues, incorporating novel AI approaches, and tackling the analysis of rare cardiac diseases., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier B.V. All rights reserved.)

The ability to reconstruct high-quality images from undersampled MRI data is vital in improving MRI temporal resolution and reducing acquisition times. Deep learning methods have been proposed for this task, but the lack of verified methods to quantify the uncertainty in the reconstructed images hampered clinical applicability. We introduce "NPB-REC", a non-parametric fully Bayesian framework, for MRI reconstruction from undersampled data with uncertainty estimation. We use Stochastic Gradient Langevin Dynamics during training to characterize the posterior distribution of the network parameters. This enables us to both improve the quality of the reconstructed images and quantify the uncertainty in the reconstructed images. We demonstrate the efficacy of our approach on a multi-coil MRI dataset from the fastMRI challenge and compare it to the baseline End-to-End Variational Network (E2E-VarNet). Our approach outperforms the baseline in terms of reconstruction accuracy by means of PSNR and SSIM (34.55, 0.908 vs. 33.08, 0.897, p<0.01, acceleration rate R=8) and provides uncertainty measures that correlate better with the reconstruction error (Pearson correlation, R=0.94 vs. R=0.91). Additionally, our approach exhibits better generalization capabilities against anatomical distribution shifts (PSNR and SSIM of 32.38, 0.849 vs. 31.63, 0.836, p<0.01, training on brain data, inference on knee data, acceleration rate R=8). NPB-REC has the potential to facilitate the safe utilization of deep learning-based methods for MRI reconstruction from undersampled data. Code and trained models are available at https://github.com/samahkh/NPB-REC., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier B.V. All rights reserved.)

Identifying COVID-19 through blood sample analysis is crucial in managing the disease and improving patient outcomes. Despite its advantages, the current test demands certified laboratories, expensive equipment, trained personnel, and 3-4 h for results, with a notable false-negative rate of 15%-20%. This study proposes a stacked deep-learning approach for detecting COVID-19 in blood samples to distinguish uninfected individuals from those infected with the virus. Three stacked deep learning architectures, namely the StackMean, StackMax, and StackRF algorithms, are introduced to improve the detection quality of single deep learning models. To counter the class imbalance phenomenon in the training data, the Synthetic Minority Oversampling Technique (SMOTE) algorithm is also implemented, resulting in increased specificity and sensitivity. The efficacy of the methods is assessed by utilizing blood samples obtained from hospitals in Brazil and Italy. Results revealed that the StackMax method greatly boosted the deep learning and traditional machine learning methods' capability to distinguish COVID-19-positive cases from normal cases, while SMOTE increased the specificity and sensitivity of the stacked models. Hypothesis testing is performed to determine if there is a significant statistical difference in the performance between the compared detection methods. Additionally, the significance of blood sample features in identifying COVID-19 is analyzed using the XGBoost (eXtreme Gradient Boosting) technique for feature importance identification. Overall, this methodology could potentially enhance the timely and precise identification of COVID-19 in blood samples., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier B.V. All rights reserved.)

The development of intelligent systems that use social media data for decision-making processes in numerous domains such as politics, business, marketing, and finance, has been made possible by the popularity of social media platforms. However, the utilization of textual data from social media in the healthcare management industry is still somewhat limited when it is compared to other industries. Investigating how current machine learning and natural language processing technologies can be used in the healthcare industry to gauge public sentiment is an important study. Earlier works on healthcare sentiment analysis have utilized traditional word embedding models trained on the general and medical corpus. However, integration of medical knowledge to pre-trained word embedding models has not been considered yet. Word embedding models trained on the general corpus led to the problem of lacking medical knowledge and the models trained on the small size of the medical corpus have limitations in capturing semantic and syntactic properties. This research proposes a new word embedding model named Word Embedding Integrated with Medical Knowledge Vector (WE-iMKVec). The proposed model integrates sentiment lexicons and medical knowledgebases into the pre-trained word embedding to enrich the properties of word embedding. A new medical-aware sentiment polarity score is proposed for the utilization in learning neural-network sentiment and these vectors incorporate with the original pre-trained word vectors. The resulting vectors are enriched with lexicon vectors and the medical knowledge vectors: Adverse Drug Reaction (ADR) vector and Unified Medical Language System (UMLS) vector are used to build the proposed WE-iMKVec model. WE-iMKVec is validated on the five different social media healthcare review datasets and the empirical results showed its superiority over traditional word embedding models in medical sentiment analysis. The highest improvement can be found in the patients.info medical condition dataset where the proposed model outperforms three conventional word2vec models (Google-News, PubMed-PMC, and Drug Reviews) by 12.7 %, 31.4 %, and 25.4 % respectively in terms of F1 score., Competing Interests: Declaration of competing interest The authors declare that the authors have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier B.V. All rights reserved.)

Motivation: Acute ischemic stroke is one of the leading causes of morbidity and disability worldwide, often followed by a long rehabilitation period. To improve and personalize stroke rehabilitation, it is essential to provide a reliable prognosis to caregivers and patients. Deep learning techniques might improve the predictions by incorporating different data modalities. We present a multimodal approach to predict the functional status of acute ischemic stroke patients after their discharge based on tabular data and CT perfusion imaging., Methods: We conducted experiments on tabular, imaging, and multimodal deep learning architectures to predict dichotomized mRS scores 3 months after the event. The dataset was collected from a Dutch hospital and includes 98 CVA patients with a visible occlusion on their CT perfusion scan. Tabular data is based on the Dutch Acute Stroke Audit data, and imaging data consists of summed-up CT perfusion maps., Results: On the tabular data, TabNet outperformed our baselines with an AUC of 0.71, while ResNet-10 on the imaging data performed comparably with an AUC of 0.70. Our implementation of the multimodal DAFT architecture outperforms baselines as well as comparable studies by achieving an 0.75 AUC, and 0.80 F1 score. This was achieved with a final model of less than a hundred thousand optimizable parameters, and a dataset less than half the size of reference papers., Conclusion: Overall, we demonstrate the feasibility of predicting the functional outcome for ischemic stroke patients and the usability of multimodal deep learning architectures for this purpose., Competing Interests: Declaration of competing interest None, (Copyright © 2023 The Authors. Published by Elsevier B.V. All rights reserved.)

Human accuracy in diagnosing psychiatric disorders is still low. Even though digitizing health care leads to more and more data, the successful adoption of AI-based digital decision support (DDSS) is rare. One reason is that AI algorithms are often not evaluated based on large, real-world data. This research shows the potential of using deep learning on the medical claims data of 812,853 people between 2018 and 2022, with 26,973,943 ICD-10-coded diseases, to predict depression (F32 and F33 ICD-10 codes). The dataset used represents almost the entire adult population of Estonia. Based on these data, to show the critical importance of the underlying temporal properties of the data for the detection of depression, we evaluate the performance of non-sequential models (LR, FNN), sequential models (LSTM, CNN-LSTM) and the sequential model with a decay factor (GRU-Δt, GRU-decay). Furthermore, since explainability is necessary for the medical domain, we combine a self-attention model with the GRU decay and evaluate its performance. We named this combination Att-GRU-decay. After extensive empirical experimentation, our model (Att-GRU-decay), with an AUC score of 0.990, an AUPRC score of 0.974, a specificity of 0.999 and a sensitivity of 0.944, proved to be the most accurate. The results of our novel Att-GRU-decay model outperform the current state of the art, demonstrating the potential usefulness of deep learning algorithms for DDSS development. We further expand this by describing a possible application scenario of the proposed algorithm for depression screening in a general practitioner (GP) setting-not only to decrease healthcare costs, but also to improve the quality of care and ultimately decrease people's suffering., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier B.V. All rights reserved.)

Electroencephalogram (EEG)-based Brain-Computer Interfaces (BCIs) build a communication path between human brain and external devices. Among EEG-based BCI paradigms, the most commonly used one is motor imagery (MI). As a hot research topic, MI EEG-based BCI has largely contributed to medical fields and smart home industry. However, because of the low signal-to-noise ratio (SNR) and the non-stationary characteristic of EEG data, it is difficult to correctly classify different types of MI-EEG signals. Recently, the advances in Deep Learning (DL) significantly facilitate the development of MI EEG-based BCIs. In this paper, we provide a systematic survey of DL-based MI-EEG classification methods. Specifically, we first comprehensively discuss several important aspects of DL-based MI-EEG classification, covering input formulations, network architectures, public datasets, etc. Then, we summarize problems in model performance comparison and give guidelines to future studies for fair performance comparison. Next, we fairly evaluate the representative DL-based models using source code released by the authors and meticulously analyse the evaluation results. By performing ablation study on the network architecture, we found that (1) effective feature fusion is indispensable for multi-stream CNN-based models. (2) LSTM should be combined with spatial feature extraction techniques to obtain good classification performance. (3) the use of dropout contributes little to improving the model performance, and that (4) adding fully connected layers to the models significantly increases their parameters but it might not improve their performance. Finally, we raise several open issues in MI-EEG classification and provide possible future research directions., Competing Interests: Declaration of competing interest The authors declared that they have no conflicts of interest to this work. We declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted., (Copyright © 2023 The Authors. Published by Elsevier B.V. All rights reserved.)

Accurate measurement of blood flow velocity is important for the prevention and early diagnosis of atherosclerosis. However, due to the uncertainty of parameter settings, the autocorrelation velocimetry methods based on clutter filtering are prone to incorrectly filter out the near-wall blood flow signal, resulting in poor velocimetric accuracy. In addition, the Doppler coherent compounding acts as a low-pass filter, which also leads to low values of blood flow velocity estimated by the above methods. Motivated by this status quo, here we propose a deep learning estimator that combines clutter filtering and blood flow velocimetry based on the adaptive property of one-dimensional convolutional neural network (1DCNN). The estimator is operated by first extracting the blood flow signal from the original Doppler echo signal through an affine transformation of the 1D convolution, and then converting the extracted signal into the desired blood flow velocity using a linear transformation function. The effectiveness of the proposed method is verified by simulation as well as in vivo carotid artery data. Compared with typical velocimetry methods such as high-pass filtering (HPF) and singular value decomposition (SVD), the results show that the normalized root means square error (NRMSE) obtained by 1DCNN is reduced by 54.99 % and 53.50 % for forward blood flow velocimetry, and 70.99 % and 69.50 % for reverse blood flow velocimetry, respectively. Consistently, the in vivo measurements demonstrate that the goodness-of-fit of the proposed estimator is improved by 8.72 % and 4.74 % for five subjects. Moreover, the estimation time consumed by 1DCNN is greatly reduced, which costs only 2.91 % of the time of HPF and 12.83 % of the time of SVD. In conclusion, the proposed estimator is a better alternative to the current blood flow velocimetry, and is capable of providing more accurate diagnosis information for vascular diseases in clinical applications., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier B.V. All rights reserved.)

Coronavirus (COVID-19) is a newly discovered viral disease from the SARS-CoV-2 family. This has caused a moral panic resulting in the spread of informative and uninformative information about COVID-19 and its effects. Twitter is a popular social media platform used extensively during the current outbreak. This paper aims to predict informative tweets related to COVID-19 on Twitter using a novel set of fuzzy rules involving deep learning techniques. This study focuses on identifying informative tweets during the pandemic to provide the public with trustworthy information and forecast how quickly diseases could spread. In this case, we have implemented RoBERTa and CT-BERT models using the fuzzy methodology to identify COVID-19 patient tweets. The proposed architecture combines deep learning transformer models RoBERTa and CT-BERT with the fuzzy technique to categorize posts as INFORMATIVE or UNINFORMATIVE. We performed a comparative analysis of our method with machine learning models and deep learning approaches. The results show that our proposed model can classify informative and uninformative tweets with an accuracy of 91.40% and an F1-score of 91.94% using the COVID-19 English tweet dataset. The proposed model is accurate and ready for real-world application., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier B.V. All rights reserved.)

Attention Deficit/Hyperactivity Disorder (ADHD) is a prevalent neurodevelopmental disorder in childhood that often persists into adulthood. Objectively diagnosing ADHD can be challenging due to the reliance on subjective questionnaires in clinical assessment. Fortunately, recent advancements in artificial intelligence (AI) have shown promise in providing objective diagnoses through the analysis of medical images or activity recordings. These AI-based techniques have demonstrated accurate ADHD diagnosis; however, the growing complexity of deep learning models has introduced a lack of interpretability. These models often function as black boxes, unable to offer meaningful insights into the data patterns that characterize ADHD., Objective: This paper proposes a methodology to interpret the output of an AI-based diagnosis system for combined ADHD in age and gender-stratified populations., Methods: Our system is based on the analysis of 24 hour-long activity records using Convolutional Neural Networks (CNNs) to classify spectrograms of activity windows. These windows are interpreted using occlusion maps to highlight the time-frequency patterns explaining ADHD activity., Results: Significant differences in the frequency patterns between ADHD and controls both in diurnal and nocturnal activity were found for all the populations. Temporal dispersion also presented differences in the male population., Conclusion: The proposed interpretation techniques for CNNs highlighted gender- and age-related differences between ADHD patients and controls. Leveraging these differences could potentially lead to improved diagnostic accuracy, especially if a larger and more balanced dataset is utilized., Significance: Our findings pave the way for the development of an AI-based diagnosis system for ADHD that offers interpretability, thereby providing valuable insights into the underlying etiology of the disease., Competing Interests: Declaration of competing interest The authors declare no conflict of interest., (Copyright © 2023 The Author(s). Published by Elsevier B.V. All rights reserved.)

Current models on Explainable Artificial Intelligence (XAI) have shown a lack of reliability when evaluating feature-relevance for deep neural biomarker classifiers. The inclusion of reliable saliency-maps for obtaining trustworthy and interpretable neural activity is still insufficiently mature for practical applications. These limitations impede the development of clinical applications of Deep Learning. To address, these limitations we propose the RemOve-And-Retrain (ROAR) algorithm which supports the recovery of highly relevant features from any pre-trained deep neural network. In this study we evaluated the ROAR methodology and algorithm for the Face Emotion Recognition (FER) task, which is clinically applicable in the study of Autism Spectrum Disorder (ASD). We trained a Convolutional Neural Network (CNN) from electroencephalography (EEG) signals and assessed the relevance of FER-elicited EEG features from individuals diagnosed with and without ASD. Specifically, we compared the ROAR reliability from well-known relevance maps such as Layer-Wise Relevance Propagation, PatternNet, Pattern-Attribution, and Smooth-Grad Squared. This study is the first to bridge previous neuroscience and ASD research findings to feature-relevance calculation for EEG-based emotion recognition with CNN in typically-development (TD) and in ASD individuals., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023. Published by Elsevier B.V.)

Evolutionary algorithms have been successfully employed to find the best structure for many learning algorithms including neural networks. Due to their flexibility and promising results, Convolutional Neural Networks (CNNs) have found their application in many image processing applications. The structure of CNNs greatly affects the performance of these algorithms both in terms of accuracy and computational cost, thus, finding the best architecture for these networks is a crucial task before they are employed. In this paper, we develop a genetic programming approach for the optimization of CNN structure in diagnosing COVID-19 cases via X-ray images. A graph representation for CNN architecture is proposed and evolutionary operators including crossover and mutation are specifically designed for the proposed representation. The proposed architecture of CNNs is defined by two sets of parameters, one is the skeleton which determines the arrangement of the convolutional and pooling operators and their connections and one is the numerical parameters of the operators which determine the properties of these operators like filter size and kernel size. The proposed algorithm in this paper optimizes the skeleton and the numerical parameters of the CNN architectures in a co-evolutionary scheme. The proposed algorithm is used to identify covid-19 cases via X-ray images., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 The Author(s). Published by Elsevier B.V. All rights reserved.)

Early melanoma diagnosis is the most important factor in the treatment of skin cancer and can effectively reduce mortality rates. Recently, Generative Adversarial Networks have been used to augment data, prevent overfitting and improve the diagnostic capacity of models. However, its application remains a challenging task due to the high levels of inter and intra-class variance seen in skin images, limited amounts of data, and model instability. We present a more robust Progressive Growing of Adversarial Networks based on residual learning, which is highly recommended to ease the training of deep networks. The stability of the training process was increased by receiving additional inputs from preceding blocks. The architecture is able to produce plausible photorealistic synthetic 512 × 512 skin images, even with small dermoscopic and non-dermoscopic skin image datasets as problem domains. In this manner, we tackle the lack of data and the imbalance problems. Additionally, the proposed approach leverages a skin lesion boundary segmentation algorithm and transfer learning to enhance the diagnosis of melanoma. Inception score and Matthews Correlation Coefficient were used to measure the performance of the models. The architecture was evaluated qualitatively and quantitatively through the use of an extensive experimental study on sixteen datasets, illustrating its effectiveness in the diagnosis of melanoma. Finally, four state-of-the-art data augmentation techniques applied in five convolutional neural network models were significantly outperformed. The results indicated that a bigger number of trainable parameters will not necessarily obtain a better performance in melanoma diagnosis., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 The Author(s). Published by Elsevier B.V. All rights reserved.)

Deep learning has become a thriving force in the computer aided diagnosis of liver cancer, as it solves extremely complicated challenges with high accuracy over time and facilitates medical experts in their diagnostic and treatment procedures. This paper presents a comprehensive systematic review on deep learning techniques applied for various applications pertaining to liver images, challenges faced by the clinicians in liver tumour diagnosis and how deep learning bridges the gap between clinical practice and technological solutions with an in-depth summary of 113 articles. Since, deep learning is an emerging revolutionary technology, recent state-of-the-art research implemented on liver images are reviewed with more focus on classification, segmentation and clinical applications in the management of liver diseases. Additionally, similar review articles in literature are reviewed and compared. The review is concluded by presenting the contemporary trends and unaddressed research issues in the field of liver tumour diagnosis, offering directions for future research in this field., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier B.V. All rights reserved.)

Secondary hypertension is associated with higher risks of target organ damage and cardiovascular and cerebrovascular disease events. Early aetiology identification can eliminate aetiologies and control blood pressure. However, inexperienced doctors often fail to diagnose secondary hypertension, and comprehensively screening for all causes of high blood pressure increases health care costs. To date, deep learning has rarely been involved in the differential diagnosis of secondary hypertension. Relevant machine learning methods cannot combine textual information such as chief complaints with numerical information such as the laboratory examination results in electronic health records (EHRs), and the use of all features increases health care costs. To reduce redundant examinations and accurately identify secondary hypertension, we propose a two-stage framework that follows clinical procedures. The framework carries out an initial diagnosis process in the first stage, on which basis patients are recommended for disease-related examinations, followed by differential diagnoses of different diseases based on the different characteristics observed in the second stage. We convert the numerical examination results into descriptive sentences, thus blending textual and numerical characteristics. Medical guidelines are introduced through label embedding and attention mechanisms to obtain interactive features. Our model was trained and evaluated using a cross-sectional dataset containing 11,961 patients with hypertension from January 2013 to December 2019. The F1 scores of our model were 0.912, 0.921, 0.869 and 0.894 for primary aldosteronism, thyroid disease, nephritis and nephrotic syndrome and chronic kidney disease, respectively, which are four kinds of secondary hypertension with high incidence rates. The experimental results show that our model can powerfully use the textual and numerical data contained in EHRs to provide effective decision support for the differential diagnosis of secondary hypertension., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier B.V. All rights reserved.)

In recent years, deep learning has been used to develop an automatic breast cancer detection and classification tool to assist doctors. In this paper, we proposed a three-stage deep learning framework based on an anchor-free object detection algorithm, named the Probabilistic Anchor Assignment (PAA) to improve diagnosis performance by automatically detecting breast lesions (i.e., mass and calcification) and further classifying mammograms into benign or malignant. Firstly, a single-stage PAA-based detector roundly finds suspicious breast lesions in mammogram. Secondly, we designed a two-branch ROI detector to further classify and regress these lesions that aim to reduce the number of false positives. Besides, in this stage, we introduced a threshold-adaptive post-processing algorithm with dense breast information. Finally, the benign or malignant lesions would be classified by an ROI classifier which combines local-ROI features and global-image features. In addition, considering the strong correlation between the task of detection head of PAA and the task of whole mammogram classification, we added an image classifier that utilizes the same global-image features to perform image classification. The image classifier and the ROI classifier jointly guide to enhance the feature extraction ability and further improve the performance of classification. We integrated three public datasets of mammograms (CBIS-DDSM, INbreast, MIAS) to train and test our model and compared our framework with recent state-of-the-art methods. The results show that our proposed method can improve the diagnostic efficiency of radiologists by automatically detecting and classifying breast lesions and classifying benign and malignant mammograms., Competing Interests: Declaration of competing interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2022. Published by Elsevier B.V.)

In assisted reproductive technology (ART), embryos produced by in vitro fertilization (IVF) are graded according to their live birth potential, and high-grade embryos are preferentially transplanted. However, rates of live birth following clinical ART remain low worldwide. Grading is based on the embryo shape at a limited number of stages and does not consider the shape of embryos and intracellular structures, e.g., nuclei, at various stages important for normal embryogenesis. Here, we developed a Normalized Multi-View Attention Network (NVAN) that directly predicts live birth potential from the nuclear structure in live-cell fluorescence images of mouse embryos from zygote to across a wide range of stages. The input is morphological features of cell nuclei, which were extracted as multivariate time-series data by using the segmentation algorithm for mouse embryos. The classification accuracy of our method (83.87%) greatly exceeded that of existing machine-learning methods and that of visual inspection by embryo culture specialists. Our method also has a new attention mechanism that allows us to determine which values of multivariate time-series data, used to describe nuclear morphology, were the basis for the prediction. By visualizing the features that contributed most to the prediction of live birth potential, we found that the size and shape of the nucleus at the morula stage and at the time of cell division were important for live birth prediction. We anticipate that our method will help ART and developmental engineering as a new basic technology for IVF embryo selection., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2022 The Authors. Published by Elsevier B.V. All rights reserved.)

The adoption of electronic health records in hospitals has ensured the availability of large datasets that can be used to predict medical complications. The trajectories of patients in real-world settings are highly variable, making longitudinal data modeling challenging. In recent years, significant progress has been made in the study of deep learning models applied to time series; however, the application of these models to irregular medical time series (IMTS) remains limited. To address this issue, we developed a generic deep-learning-based framework for modeling IMTS that facilitates the comparative studies of sequential neural networks (transformers and long short-term memory) and irregular time representation techniques. A validation study to predict retinopathy complications was conducted on 1207 patients with type 1 diabetes in a French database using their historical glycosylated hemoglobin measurements, without any data aggregation or imputation. The transformer-based model combined with the soft one-hot representation of time gaps achieved the highest score: an area under the receiver operating characteristic curve of 88.65%, specificity of 85.56%, sensitivity of 83.33% and an improvement of 11.7% over the same architecture without time information. This is the first attempt to predict retinopathy complications in patients with type 1 diabetes using deep learning and longitudinal data collected from patient visits. This study highlighted the significance of modeling time gaps between medical records to improve prediction performance and the utility of a generic framework for conducting extensive comparative studies., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2022 Elsevier B.V. All rights reserved.)

Computer-aided detection systems based on deep learning have shown great potential in breast cancer detection. However, the lack of domain generalization of artificial neural networks is an important obstacle to their deployment in changing clinical environments. In this study, we explored the domain generalization of deep learning methods for mass detection in digital mammography and analyzed in-depth the sources of domain shift in a large-scale multi-center setting. To this end, we compared the performance of eight state-of-the-art detection methods, including Transformer based models, trained in a single domain and tested in five unseen domains. Moreover, a single-source mass detection training pipeline was designed to improve the domain generalization without requiring images from the new domain. The results show that our workflow generalized better than state-of-the-art transfer learning based approaches in four out of five domains while reducing the domain shift caused by the different acquisition protocols and scanner manufacturers. Subsequently, an extensive analysis was performed to identify the covariate shifts with the greatest effects on detection performance, such as those due to differences in patient age, breast density, mass size, and mass malignancy. Ultimately, this comprehensive study provides key insights and best practices for future research on domain generalization in deep learning based breast cancer detection., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2022 The Author(s). Published by Elsevier B.V. All rights reserved.)

Chest X-ray (CXR) imaging is one of the most common diagnostic imaging techniques in clinical diagnosis and is usually used for radiological examinations to screen for thorax diseases. In this paper, we propose a novel computer-aided diagnosis (CAD) system based on a hybrid deep learning model composed of a convolutional neural network (CNN) and a graph neural network (GNN). The system is intended to explore implicit correlations between thorax diseases to aid in the multilabel chest X-ray image classification task, which we term ‶CheXGAT‶. Specifically, the proposed CheXGAT framework comprises two main modules: an image representation learning (IRL) module and a graph representation learning (GRL) module. We employ the IRL module to learn high-level visual representation features from the CXR image. From the GRL module, the self-attention mechanism aggregates neighborhood features from the graphic structure to enhance the implicit correlation between thorax diseases. We adopted a data-driven method to create a disease correlation matrix that works on the message passing and aggregation process for the nodes in the GRL module. After end-to-end training, the GRL module enhances the correlation between thorax diseases to improve diagnosis performance. We performed experiments on the NIH Chest X-ray14 dataset, which contains 112,120 frontal-view radiographs; each image has multiple thorax disease labels. In the experimental results, the average AUC score of our proposed CheXGAT model reached 0.8266, and the AUC scores of Emphysema and Hernia reached 0.9447 and 0.9313, respectively. In addition, we visualized explanations via a gradient-based localization method in our proposed CheXGAT framework. Compared with previous studies, the experimental results show the competitive performance of our framework. Specifically, we propose a CAD system that uses a hybrid model to help radiologists identify CXR images from 14 chest diseases for clinical diagnosis., Competing Interests: Declaration of competing interest The authors declare that they have no financial and personal relationships with other people or organizations that could inappropriately influence their work., (Copyright © 2022 Elsevier B.V. All rights reserved.)

Segmentation of specific brain tissue from MRI volumes is of great significance for brain disease diagnosis, progression assessment, and monitoring of neurological conditions. Manual segmentation is time-consuming, laborious, and subjective, which significantly amplifies the need for automated processes. Over the last decades, the active development in the field of deep learning, especially convolutional neural networks (CNNs), and the associated performance improvements have increased the demand for the application of CNN-based methods to provide consistent measurements and quantitative analyses. In this paper, we present an efficient deep learning approach for the segmentation of brain tissue. More specifically, we address the problem of segmentation of the posterior limb of the internal capsule (PLIC) in preterm neonates. To this end, we propose a CNN-based pipeline comprised of slice-selection modules and a multi-view segmentation model, which exploits the 3D information contained in the MRI volumes to improve segmentation performance. One special feature of the proposed method is its ability to identify one desired slice out of the whole image volume, which is relevant for pediatricians in terms of prognosis. To increase computational efficiency, we apply a strategy that automatically reduces the information contained in the MRI volumes to its relevant parts. Finally, we conduct an expert rating alongside standard evaluation metrics, such as dice score, to evaluate the performance of the proposed framework. We demonstrate the benefit of the multi-view technique by comparing it with its single-view counterparts, which reveals that the proposed method strikes a good balance between exploiting the available image information and reducing the required computing power compared to 3D segmentation networks. Standard evaluation metrics as, well as expert-based assessment, confirm the good performance of the proposed framework, with the latter being more relevant in terms of clinical applicability. We demonstrate that the proposed deep learning pipeline can compete with the experts in terms of accuracy. To prove the generalisability of the proposed method, we additionally assess our deep learning pipeline to data from the Developing Human Connectome Project (dHCP)., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2022 The Author(s). Published by Elsevier B.V. All rights reserved.)

The segmentation of magnetic resonance (MR) images is a crucial task for creating pseudo computed tomography (CT) images which are used to achieve positron emission tomography (PET) attenuation correction. One of the main challenges of creating pseudo CT images is the difficulty to obtain an accurate segmentation of the bone tissue in brain MR images. Deep convolutional neural networks (CNNs) have been widely and efficiently applied to perform MR image segmentation. The aim of this work is to propose a segmentation approach that combines multiresolution handcrafted features with CNN-based features to add directional properties and enrich the set of features to perform segmentation. The main objective is to efficiently segment the brain into three tissue classes: bone, soft tissue, and air. The proposed method combines non subsampled Contourlet (NSCT) and non subsampled Shearlet (NSST) coefficients with CNN's features using different mechanisms. The entropy value is calculated to select the most useful coefficients and reduce the input's dimensionality. The segmentation results are evaluated using fifty clinical brain MR and CT images by calculating the precision, recall, dice similarity coefficient (DSC), and Jaccard similarity coefficient (JSC). The results are also compared to other methods reported in the literature. The DSC of the bone class is improved from 0.6179 ± 0.0006 to 0.6416 ± 0.0006. The addition of multiresolution features of NSCT and NSST with CNN's features demonstrates promising results. Moreover, NSST coefficients provide more useful information than NSCT coefficients., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2022 Elsevier B.V. All rights reserved.)

Alzheimer's Disease (AD) is an irreversible neurodegenerative disease that results in a progressive decline in cognitive abilities. Since AD starts several years before the onset of the symptoms, its early detection is challenging due to subtle changes in biomarkers mainly detectable in different neuroimaging modalities. Developing computer-aided diagnostic models based on deep learning can provide excellent opportunities for the analysis of different neuroimage modalities along with other non-image biomarkers. In this survey, we perform a comparative analysis of about 100 published papers since 2019 that employ basic deep architectures such as CNN, RNN, and generative models for AD diagnosis. Moreover, about 60 papers that have applied a trending topic or architecture for AD are investigated. Explainable models, normalizing flows, graph-based deep architectures, self-supervised learning, and attention mechanisms are considered. The main challenges in this body of literature have been categorized and explained from data-related, methodology-related, and clinical adoption aspects. We conclude our paper by addressing some future perspectives and providing recommendations to conduct further studies for AD diagnosis., (Copyright © 2021. Published by Elsevier B.V.)

Deep learning-based methods, in particular, convolutional neural networks and fully convolutional networks are now widely used in the medical image analysis domain. The scope of this review focuses on the analysis using deep learning of focal liver lesions, with a special interest in hepatocellular carcinoma and metastatic cancer; and structures like the parenchyma or the vascular system. Here, we address several neural network architectures used for analyzing the anatomical structures and lesions in the liver from various imaging modalities such as computed tomography, magnetic resonance imaging and ultrasound. Image analysis tasks like segmentation, object detection and classification for the liver, liver vessels and liver lesions are discussed. Based on the qualitative search, 91 papers were filtered out for the survey, including journal publications and conference proceedings. The papers reviewed in this work are grouped into eight categories based on the methodologies used. By comparing the evaluation metrics, hybrid models performed better for both the liver and the lesion segmentation tasks, ensemble classifiers performed better for the vessel segmentation tasks and combined approach performed better for both the lesion classification and detection tasks. The performance was measured based on the Dice score for the segmentation, and accuracy for the classification and detection tasks, which are the most commonly used metrics., (Copyright © 2022 The Authors. Published by Elsevier B.V. All rights reserved.)

Glioma is a relatively common brain tumor disease with high mortality rate. Humans have been seeking a more effective therapy. In the course of treatment, the specific location of the tumor needs to be determined first in any case. Therefore, how to segment tumors from brain tissue accurately and quickly is a persistent problem. In this paper, a new dual-stream decoding CNN architecture combined with U-net for automatic segmentation of brain tumor on MR images namely DDU-net is proposed. Two edge-based optimization strategies are used to enhance the performance of brain tumor segmentation. First, we design a separate branch to process edge stream information. Here, high level edge features are reduced in dimension of channel and integrated into the conventional semantic stream in the way of residual. Second, a regularization loss function is used to encourage the predicted segmentation mask to align with ground truth around the edge mainly by penalizing pixels where the predicted segmentation masks and labels do not match around the edge. In training, we employ a novel edge extraction algorithm for providing edge labels with higher quality. Moreover, we add a self-adaptive balancing class weight coefficient into the cross entropy loss function for solving the serious class imbalance problem in the backpropagation of edge extraction. Our experiments show that this leads to a very efficient architecture which can produce clearer prediction at the edge of the tumor. Our method achieves ideal performance on BraTS2017 and BraTS2018 in terms of Dice coefficient., (Copyright © 2021 Elsevier B.V. All rights reserved.)

Developing deep learning models to analyze histology images has been computationally challenging, as the massive size of the images causes excessive strain on all parts of the computing pipeline. This paper proposes a novel deep learning-based methodology for improving the computational efficiency of histology image classification. The proposed approach is robust when used with images that have reduced input resolution, and it can be trained effectively with limited labeled data. Moreover, our approach operates at either the tissue- or slide-level, removing the need for laborious patch-level labeling. Our method uses knowledge distillation to transfer knowledge from a teacher model pre-trained at high resolution to a student model trained on the same images at a considerably lower resolution. Also, to address the lack of large-scale labeled histology image datasets, we perform the knowledge distillation in a self-supervised fashion. We evaluate our approach on three distinct histology image datasets associated with celiac disease, lung adenocarcinoma, and renal cell carcinoma. Our results on these datasets demonstrate that a combination of knowledge distillation and self-supervision allows the student model to approach and, in some cases, surpass the teacher model's classification accuracy while being much more computationally efficient. Additionally, we observe an increase in student classification performance as the size of the unlabeled dataset increases, indicating that there is potential for this method to scale further with additional unlabeled data. Our model outperforms the high-resolution teacher model for celiac disease in accuracy, F1-score, precision, and recall while requiring 4 times fewer computations. For lung adenocarcinoma, our results at 1.25× magnification are within 1.5% of the results for the teacher model at 10× magnification, with a reduction in computational cost by a factor of 64. Our model on renal cell carcinoma at 1.25× magnification performs within 1% of the teacher model at 5× magnification while requiring 16 times fewer computations. Furthermore, our celiac disease outcomes benefit from additional performance scaling with the use of more unlabeled data. In the case of 0.625× magnification, using unlabeled data improves accuracy by 4% over the tissue-level baseline. Therefore, our approach can improve the feasibility of deep learning solutions for digital pathology on standard computational hardware and infrastructures., (Copyright © 2021 The Author(s). Published by Elsevier B.V. All rights reserved.)

Subcutaneous Implantable Cardioverter-Defibrillators (S-ICDs) are used for prevention of sudden cardiac death triggered by ventricular arrhythmias. T Wave Over Sensing (TWOS) is an inherent risk with S-ICDs which can lead to inappropriate shocks. A major predictor of TWOS is a high T:R ratio (the ratio between the amplitudes of the T and R waves). Currently, patients' Electrocardiograms (ECGs) are screened over 10 s to measure the T:R ratio to determine the patients' eligibility for S-ICD implantation. Due to temporal variations in the T:R ratio, 10 s is not a long enough window to reliably determine the normal values of a patient's T:R ratio. In this paper, we develop a convolutional neural network (CNN) based model utilising phase space reconstruction matrices to predict T:R ratios from 10-second ECG segments without explicitly locating the R or T waves, thus avoiding the issue of TWOS. This tool can be used to automatically screen patients over a much longer period and provide an in-depth description of the behavior of the T:R ratio over that period. The tool can also enable much more reliable and descriptive screenings to better assess patients' eligibility for S-ICD implantation., (Copyright © 2021 University of Southampton. Published by Elsevier B.V. All rights reserved.)

Obstructive Sleep Apnea Syndrome (OSAS) is the most common sleep-related breathing disorder. It is caused by an increased upper airway resistance during sleep, which determines episodes of partial or complete interruption of airflow. The detection and treatment of OSAS is particularly important in patients who suffered a stroke, because the presence of severe OSAS is associated with higher mortality, worse neurological deficits, worse functional outcome after rehabilitation, and a higher likelihood of uncontrolled hypertension. The gold standard test for diagnosing OSAS is polysomnography (PSG). Unfortunately, performing a PSG in an electrically hostile environment, like a stroke unit, on neurologically impaired patients is a difficult task; moreover, the number of strokes per day vastly outnumbers the availability of polysomnographs and dedicated healthcare professionals. Hence, a simple and automated recognition system to identify OSAS cases among acute stroke patients, relying on routinely recorded vital signs, is highly desirable. The vast majority of the work done so far focuses on data recorded in ideal conditions and highly selected patients, and thus it is hardly exploitable in real-life circumstances, where it would be of actual use. In this paper, we propose a novel convolutional deep learning architecture able to effectively reduce the temporal resolution of raw waveform data, like physiological signals, extracting key features that can be used for further processing. We exploit models based on such an architecture to detect OSAS events in stroke unit recordings obtained from the monitoring of unselected patients. Unlike existing approaches, annotations are performed at one-second granularity, allowing physicians to better interpret the model outcome. Results are considered to be satisfactory by the domain experts. Moreover, through tests run on a widely-used public OSAS dataset, we show that the proposed approach outperforms current state-of-the-art solutions., (Copyright © 2021 Elsevier B.V. All rights reserved.)

This paper presents a method for automatic segmentation of tympanic membranes (TMs) from video-otoscopic images based on deep fully convolutional neural network. Built upon the UNet architecture, the proposed EAR scheme is based on three main paradigms: EfficientNet for the encoder, Attention gate for the skip connection path, and Residual blocks for the decoder. The paper also introduces a new loss function term for the neural networks to perform segmentation tasks. Particularly, we propose to integrate EfficientNet-B4 into the encoder part of the UNet. In addition, the decoder part of the proposed network is constructed based on residual blocks from ResNet architecture. By this way, the proposed approach could take advantages of the EfficientNet and ResNet architectures such as preserving efficient reception field size for the model and avoiding overfitting problem. In addition, in the skip connection path, we employ the attention gate that can handle the varieties in shapes and sizes of interested objects, which are common issues in TM regions. Moreover, for network training, we proposed a new loss function term based on the shape distance between predicted and ground truth masks, and exploited the stochastic weight averaging to avoid being trapped in local minima. We evaluate the proposed approach on a TM dataset which includes 1012 otoscopic images from patients diagnosed with and without otitis media. Experimental results show that the proposed approach achieves high segmentation performance with the average Dice similarity coefficient of 0.929, without any pre- or post-processing steps, that outperforms other state-of-the-art methods., (Copyright © 2021 Elsevier B.V. All rights reserved.)

The tumor proliferation, which is correlated with tumor grade, is a crucial biomarker indicative of breast cancer patients' prognosis. The most commonly used method in predicting tumor proliferation speed is the counting of mitotic figures in Hematoxylin and Eosin (H&E) histological slides. Manual mitosis counting is known to suffer from reproducibility problems. This paper presents a fully automated system for tumor proliferation prediction from whole slide images via mitosis counting. First, by considering the epithelial tissue as mitosis activity regions, we build a deep-learning-based region of interest detection method to select the high mitosis activity regions from whole slide images. Second, we learned a set of deep neural networks to detect mitosis detection from selected areas. The proposed mitosis detection system is designed to effectively overcome the mitosis detection challenges by two novel deep preprocessing and two-step hard negative mining approaches. Third, we trained a Support Vector Machine (SVM) classifier to predict the final tumor proliferation score. The proposed method was evaluated on the dataset of the Tumor Proliferation Assessment Challenge (TUPAC16) and achieved a 73.81 % F-measure and 0.612 weighted kappa score, respectively, outperforming all previous approaches significantly. Experimental results demonstrate that the proposed system considerably improves the tumor proliferation prediction accuracy and provides a reliable automated tool to support health care make-decisions., (Copyright © 2021 Elsevier B.V. All rights reserved.)

Artificial intelligence is a broad field that comprises a wide range of techniques, where deep learning is presently the one with the most impact. Moreover, the medical field is an area where data both complex and massive and the importance of the decisions made by doctors make it one of the fields in which deep learning techniques can have the greatest impact. A systematic review following the Cochrane recommendations with a multidisciplinary team comprised of physicians, research methodologists and computer scientists has been conducted. This survey aims to identify the main therapeutic areas and the deep learning models used for diagnosis and treatment tasks. The most relevant databases included were MedLine, Embase, Cochrane Central, Astrophysics Data System, Europe PubMed Central, Web of Science and Science Direct. An inclusion and exclusion criteria were defined and applied in the first and second peer review screening. A set of quality criteria was developed to select the papers obtained after the second screening. Finally, 126 studies from the initial 3493 papers were selected and 64 were described. Results show that the number of publications on deep learning in medicine is increasing every year. Also, convolutional neural networks are the most widely used models and the most developed area is oncology where they are used mainly for image analysis., (Copyright © 2021 Elsevier B.V. All rights reserved.)

In clinical settings, a lot of medical image datasets suffer from the imbalance problem which hampers the detection of outliers (rare health care events), as most classification methods assume an equal occurrence of classes. In this way, identifying outliers in imbalanced datasets has become a crucial issue. To help address this challenge, one-class classification, which focuses on learning a model using samples from only a single given class, has attracted increasing attention. Previous one-class modeling usually uses feature mapping or feature fitting to enforce the feature learning process. However, these methods are limited for medical images which usually have complex features. In this paper, a novel method is proposed to enable deep learning models to optimally learn single-class-relevant inherent imaging features by leveraging the concept of imaging complexity. We investigate and compare the effects of simple but effective perturbing operations applied to images to capture imaging complexity and to enhance feature learning. Extensive experiments are performed on four clinical datasets to show that the proposed method outperforms four state-of-the-art methods., (Copyright © 2020 Elsevier B.V. All rights reserved.)

In this paper, we review the state-of-the-art approaches for knee articular cartilage segmentation from conventional techniques to deep learning (DL) based techniques. Knee articular cartilage segmentation on magnetic resonance (MR) images is of great importance in early diagnosis of osteoarthritis (OA). Besides, segmentation allows estimating the articular cartilage loss rate which is utilised in clinical practice for assessing the disease progression and morphological changes. It has been traditionally applied in quantifying longitudinal knee OA progression pattern to detect and assess the articular cartilage thickness and volume. Topics covered include various image processing algorithms and major features of different segmentation techniques, feature computations and the performance evaluation metrics. This paper is intended to provide researchers with a broad overview of the currently existing methods in the field, as well as to highlight the shortcomings and potential considerations in the application at clinical practice. The survey showed that state-of-the-art techniques based on DL outperform the other segmentation methods. The analysis of the existing methods reveals that integration of DL-based algorithms with other traditional model-based approaches has achieved the best results (mean Dice similarity coefficient (DSC) between 85.8% and 90%)., (Copyright © 2020 Elsevier B.V. All rights reserved.)

Lung cancer is the leading cause of cancer death worldwide. Early detection of lung cancer is helpful to provide the best possible clinical treatment for patients. Due to the limited number of radiologist and the huge number of chest x-ray radiographs (CXR) available for observation, a computer-aided detection scheme should be developed to assist radiologists in decision-making. While deep learning showed state-of-the-art performance in several computer vision applications, it has not been used for lung nodule detection on CXR. In this paper, a deep learning-based lung nodule detection method was proposed. We employed patch-based multi-resolution convolutional networks to extract the features and employed four different fusion methods for classification. The proposed method shows much better performance and is much more robust than those previously reported researches. For publicly available Japanese Society of Radiological Technology (JSRT) database, more than 99% of lung nodules can be detected when the false positives per image (FPs/image) was 0.2. The FAUC and R-CPM of the proposed method were 0.982 and 0.987, respectively. The proposed approach has the potential of applications in clinical practice., (Copyright © 2019 Elsevier B.V. All rights reserved.)

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) plays an important role in diagnosis and grading of brain tumors. Although manual DCE biomarker extraction algorithms boost the diagnostic yield of DCE-MRI by providing quantitative information on tumor prognosis and prediction, they are time-consuming and prone to human errors. In this paper, we propose a fully-automated, end-to-end system for DCE-MRI analysis of brain tumors. Our deep learning-powered technique does not require any user interaction, it yields reproducible results, and it is rigorously validated against benchmark and clinical data. Also, we introduce a cubic model of the vascular input function used for pharmacokinetic modeling which significantly decreases the fitting error when compared with the state of the art, alongside a real-time algorithm for determination of the vascular input region. An extensive experimental study, backed up with statistical tests, showed that our system delivers state-of-the-art results while requiring less than 3 min to process an entire input DCE-MRI study using a single GPU., (Copyright © 2019 Elsevier B.V. All rights reserved.)

Paralyzed patients were increasing day by day. Some of the neurodegenerative diseases like amyotrophic lateral sclerosis, Brainstem Leison, Stupor and Muscular dystrophy affect the muscle movements in the body. The affected persons were unable to migrate. To overcome from their problem they need some assistive technology with the help of bio signals. Electrooculogram (EOG) based Human Computer Interaction (HCI) is one of the technique used in recent days to overcome such problem. In this paper we clearly check the possibilities of creating nine states HCI by our proposed method. Signals were captured through five electrodes placed on the subjects face around the eyes. These signals were amplified with ADT26 bio amplifier, filtered with notch filter, and processed with reference power and band power techniques to extract features to detect the eye movements and mapped with Time Delay Neural Network to classify the eye movements to generate control signal to control external hardware devices. Our experimental study reports that maximum average classification of 91.09% for reference power feature and 91.55%-for band power feature respectively. The obtained result confirms that band power features with TDNN network models shows better performance than reference features for all subjects. From this outcome we conclude that band power features with TDNN network models was more suitable for classifying the eleven difference eye movements for individual subjects. To validate the result obtained from this method we categorize the subjects in age wise to check the accuracy of the system. Single trail analysis was conducted in offline to identify the recognizing accuracy of the proposed system. The result summarize that band power features with TDNN network models exceed the reference power with TDNN network model used in this study. Through the outcome we conclude that that band power features with TDNN network was more suitable for designing EOG based HCI in offline mode., (Copyright © 2019 Elsevier B.V. All rights reserved.)

Cancer is the second leading cause of death after cardiovascular diseases. Out of all types of cancer, brain cancer has the lowest survival rate. Brain tumors can have different types depending on their shape, texture, and location. Proper diagnosis of the tumor type enables the doctor to make the correct treatment choice and help save the patient's life. There is a high need in the Artificial Intelligence field for a Computer Assisted Diagnosis (CAD) system to assist doctors and radiologists with the diagnosis and classification of tumors. Over recent years, deep learning has shown an optimistic performance in computer vision systems. In this paper, we propose an enhanced approach for classifying brain tumor types using Residual Networks. We evaluate the proposed model on a benchmark dataset containing 3064 MRI images of 3 brain tumor types (Meningiomas, Gliomas, and Pituitary tumors). We have achieved the highest accuracy of 99% outperforming the other previous work on the same dataset., (Copyright © 2019 Elsevier B.V. All rights reserved.)

Diabetic retinopathy (DR) is an eye disease that victimize the people suffering from diabetes from many years. The severe form of DR results in form of the blindness that can initially be controlled by the DR-screening oriented treatment. The effective screening programs require the trained human resource that manually grade the fundus images to understand the severity of the disease. But due to the complexity of this process, and the insufficient number of the trained workers, the precise manual grading is an expensive process. The CAD-based solutions try to address these limitations but most of the existing DR detection systems are as evaluated over small sets and become ineffective when applied in real scenarios. Therefore, in this paper we proposed a novel technique to precisely detect the various stages of the DR by extending the research of the content-based image retrieval domain. To achieve the human-level performance over the large-scale DR-datasets (i.e. Kaggle-DR), the fundus images are represented by the novel tetragonal local octa pattern (T-LOP) features, that are then classified through the extreme learning machine (ELM). To justify the significance of the method, the proposed scheme is compared against several state-of-the-art methods including the deep learning-based methods over four DR-datasets of variational lengths (i.e. Kaggle-DR, DRIVE, Review-DB, STARE). The experimental results confirm the significance of the DR-detection scheme to serve as a stand-alone solution for providing the precise information of the severity of the DR in an efficient manner., (Copyright © 2019 Elsevier B.V. All rights reserved.)

The automated analysis of retinal images is a widely researched area which can help to diagnose several diseases like diabetic retinopathy in early stages of the disease. More specifically, separation of vessels and lesions is very critical as features of these structures are directly related to the diagnosis and treatment process of diabetic retinopathy. The complexity of the retinal image contents especially in images with severe diabetic retinopathy makes detection of vascular structure and lesions difficult. In this paper, a novel framework based on morphological component analysis (MCA) is presented which benefits from the adaptive representations obtained via dictionary learning. In the proposed Bi-level Adaptive MCA (BAMCA), MCA is extended to locally deal with sparse representation of the retinal images at patch level whereas the decomposition process occurs globally at the image level. BAMCA method with appropriately offline learnt dictionaries is adopted to work on retinal images with severe diabetic retinopathy in order to simultaneously separate vessels and exudate lesions as diagnostically useful morphological components. To obtain the appropriate dictionaries, K-SVD dictionary learning algorithm is modified to use a gated error which guides the process toward learning the main structures of the retinal images using vessel or lesion maps. Computational efficiency of the proposed framework is also increased significantly through some improvement leading to noticeable reduction in run time. We experimentally show how effective dictionaries can be learnt which help BAMCA to successfully separate exudate and vessel components from retinal images even in severe cases of diabetic retinopathy. In this paper, in addition to visual qualitative assessment, the performance of the proposed method is quantitatively measured in the framework of vessel and exudate segmentation. The reported experimental results on public datasets demonstrate that the obtained components can be used to achieve competitive results with regard to the state-of-the-art vessel and exudate segmentation methods., (Copyright © 2019 Elsevier B.V. All rights reserved.)

Diabetic retinopathy (DR) results in vision loss if not treated early. A computer-aided diagnosis (CAD) system based on retinal fundus images is an efficient and effective method for early DR diagnosis and assisting experts. A computer-aided diagnosis (CAD) system involves various stages like detection, segmentation and classification of lesions in fundus images. Many traditional machine-learning (ML) techniques based on hand-engineered features have been introduced. The recent emergence of deep learning (DL) and its decisive victory over traditional ML methods for various applications motivated the researchers to employ it for DR diagnosis, and many deep-learning-based methods have been introduced. In this paper, we review these methods, highlighting their pros and cons. In addition, we point out the challenges to be addressed in designing and learning about efficient, effective and robust deep-learning algorithms for various problems in DR diagnosis and draw attention to directions for future research., (Copyright © 2019 Elsevier B.V. All rights reserved.)