Search

• We propose a light-weight module, H-FC, which has a good ability to learn global dependencies. • H-FC can be easily substitute attention modules that lack the ability to learn global dependencies. • H-FC can be applied to high-resolution low-level feature maps. Because the appearance, shape and location of brain tumors vary greatly among different patients, brain tumor segmentation (BTS) is extremely challenging. Recently, many studies have used attention mechanisms to solve this problem, which can be roughly divided into two categories: the spatial attention based on convolution (with or without channel attention) and self-attention. Due to the limitation of convolution operations, the spatial attention based on convolution cannot learn global dependencies very well, resulting in poor performance in BTS. A simple improvement idea is to directly substitute it with self-attention, which has an excellent ability to learn global dependencies. Since self-attention is not friendly to GPU memory, this simple substitution will make the new attention mechanism unable to be applied to high-resolution low-level feature maps, which contain considerable geometric information and are also important for improving the performance of attention mechanism in BTS. In this paper, we propose a hierarchical fully connected module, named H-FC, to learn global dependencies. H-FC learns local dependencies at different feature map scales through fully connected layers hierarchically, and then combines these local dependencies as approximations of the global dependencies. H-FC requires very little GPU memory and can easily replace spatial attention module based on convolution operation, such as Attention Gate and SAM (in CBAM), to improve the performance of attention mechanisms in BTS. Adequate comparative experiments illustrate that H-FC performs better than Attention Gate and SAM (in CBAM), which lack the ability to learn global dependencies, in BTS, with improvements in most metrics and a larger improvement in Hausdorff Distance. By comparing the amount of calculation and parameters of the model before and after adding H-FC, it is prove that H-FC is light-weight. In this paper, we propose a novel H-FC to learn global dependencies. We illustrate the effectiveness of H-FC through experiments on BraTS2020 dataset. We mainly explore the influence of the region size and the number of steps on the performance of H-FC. We also confirm that the global dependencies of low-level feature maps are also important to BTS. We show that H-FC is light-weight through a time and space complexity analysis and the experimental results. [ABSTRACT FROM AUTHOR]

The modelling of complex workflows is an important problem-solving technique within healthcare settings. However, currently most of the workflow models use a simplified flow chart of patient flow obtained using on-site observations, group-based debates and brainstorming sessions, together with historic patient data. This paper presents a systematic and semi-automatic methodology for knowledge acquisition with detailed process representation using sequential interviews of people in the key roles involved in the service delivery process. The proposed methodology allows the modelling of roles, interactions, actions, and decisions involved in the service delivery process. This approach is based on protocol generation and analysis techniques such as: (i) initial protocol generation based on qualitative interviews of radiology staff, (ii) extraction of key features of the service delivery process, (iii) discovering the relationships among the key features extracted, and, (iv) a graphical representation of the final structured model of the service delivery process. The methodology is demonstrated through a case study of a magnetic resonance (MR) scanning service-delivery process in the radiology department of a large hospital. A set of guidelines is also presented in this paper to visually analyze the resulting process model for identifying process vulnerabilities. A comparative analysis of different workflow models is also conducted. [ABSTRACT FROM AUTHOR]

• Bone and soft tissue have significantly different vibration transmission capability. • There are many factors that affect the vibration transmission capability of the broken bone and soft tissue. • When the suitable frequency is chosen, the acceleration amplitude received by the proximal broken bone can be greatly different before and after collision. • The vibration acceleration amplitude can be used as a benchmark for judging whether the broken bones have collided. In the process of robotic fracture reduction, there is a risk of unintended collision of broken bones, which is not conducive to ensuring the safety of the reduction system. In order to solve this problem, this paper proposed a vibration-based collision detection method for fracture reduction process. Based on the two degree-of-freedom vibration response model, the factors affecting the respond of the vibration, including the excitation voltage, the clamping length at the proximal and distal ends, the mass and tensile force of the soft tissue, were obtained. The effects of these factors on the vibration transfer performance of broken bones and soft tissue were investigated by single factor experiments. The results showed that, in terms of peak value, the increase of excitation voltage would make the vibration amplitude increase linearly, and the increase of soft tissue mass and tension increased the vibration transmission capacity of soft tissue in the frequency range of 500–1000 Hz. In terms of peak frequency, the clamping length at the distal end had the greatest influence, which reached 74 Hz, followed by 45 Hz at the proximal end. While the influence of other factors was little. According to single factor experiments, the excitation frequency in the verification experiments was determined as 677 Hz. Under the vibration interference with the acceleration amplitude of 1.2 G, this method achieved correct detection. This research developed a broken bone collision detection method based on vibration excitation. The method can correctly detect the collision of broken bones with strong anti-interference ability. It is of great significance to improve the safety of fracture reduction process. [ABSTRACT FROM AUTHOR]

• The pose of the end-effector can be better controlled by IGPEF. • DGC can help solve the local minimum problem. • PRF can solve the issues that the robot can not reach the target nearby obstacles. • PDNN is used to solve the kinematics problem under safety constraints. • RTPVS is applied to prevent the accumulation of position errors. Safety and accuracy are essential for path planning in a surgical navigation system. In this paper, an improved path planning algorithm is proposed to increase the autonomous level of spine surgery robots for higher safety and accuracy. Firstly, the dynamic gravitational constant and piecewise repulsion function are adopted to improve the traditional Artificial Potential Field algorithm to solve the common issues of path planning, including local minimum, unable to reach the target near obstacles. To better control the pose of the end-effector in an operation space, the positions of the two endpoints of the end-effector are further constrained. Secondly, an improved Primal-Dual Neural Network with multiple constraints is proposed to minimize the joint angular velocity norm. The multiple constraints are formulated according to the planned path, the obstacle avoidance of the robot and the joint limits. Moreover, a real-time planned velocity scheme is applied to prevent the accumulation of position errors. The simulation results of the pedicle screw implantation demonstrate that the robot can find the collision-free trajectory and arrive at the target position in various complicated situations. More specifically, the error between two endpoints of the end-effector and the target pose is below 0.1 mm in reaching the surgical tool pose, while the maximum position error is around 0.05 mm when performing the planned path. Moreover, two experiments are conducted in the real-world to verify the proposed algorithm is effective in practice. [ABSTRACT FROM AUTHOR]

• Developed a progressive back-projection network (PBPN) for COVID-CT super-resolution, which improves the performance of COVID-CT super-resolution. • Designed an up-projection and down-projection residual modules (UD) to minimize the reconstruction error. • Designed a residual attention module to extract deep high-frequency information. Recently, the COVID-19 epidemic has become more and more serious around the world, how to improve the image resolution of COVID-CT is a very important task. The network based on progressive upsampling for COVID-CT super-resolution increases the reconstruction error. This paper proposes a progressive back-projection network (PBPN) for COVID-CT super-resolution to solve this problem. In this paper, we propose a progressive back-projection network (PBPN) for COVID-CT super-resolution. PBPN is divided into two stages, and each stage consists of back-projection, deep feature extraction and upscaling. We design an up-projection and down-projection residual module to minimize the reconstruction error and construct a residual attention module to extract deep features. In each stage, firstly, PBPN performs back-projection to extract shallow features by two up-projection and down-projection residual modules; then, PBPN extracts deep features from the shallow features by two residual attention modules; finally, PBPN upsamples the deep features through sub-pixel convolution. The proposed method achieves the improvements of about 0.14~0.47 dB/0.0012~0.0060 for × 2 scale factor, 0.02~0.08 dB/0.0024~0.0059 for × 3 scale factor, and 0.08~0.41 dB/ 0.0040~0.0147 for × 4 scale factor than state-of-the-art methods (Bicubic, SRCNN, FSRCNN, VDSR, LapSRN, DRCN and DSRN) in terms of PSNR/SSIM on benchmark datasets. The proposed mehtod obtains better performance for COVID-CT super-resolution and reconstructs high-quality high-resolution COVID-CT images that contain more details and edges. [ABSTRACT FROM AUTHOR]

• MR image reconstruction is an important inverse problem in medical imaging. • Deep learning methods can speed up and improve the quality in MRI reconstruction. • We propose a cascaded deep learning framework, which utilizes U-Nets as sub-blocks. • The proposed configuration is evaluated in detail by using the fastMRI dataset. • Simulations indicate improved reconstruction results compared to competing methods. Background and Objective: Recent studies in deep learning reveal that the U-Net stands out among the diverse set of deep models as an effective network structure, especially for imaging inverse problems. Initially, the U-Net model was developed to solve segmentation problems for biomedical images while using an annotated dataset. In this paper, we will study a novel application of the U-Net structure for the important inverse problem of MRI reconstruction. Deep networks are particularly efficient for the speed-up of the MR image reconstruction process by decreasing the data acquisition time, and they can significantly reduce the aliasing artifacts caused by the undersampling in the k-space. Our aim is to develop a novel and efficient cascaded U-Net framework for reconstructing MR images from undersampled k-space data. The new framework should have improved reconstruction performance when compared to competing methodologies. In this paper, a novel cascaded framework utilizing the U-Net as a sub-block is being proposed. The introduced U-Net cascade structure is applied to the magnetic resonance image reconstruction problem. The connection between the cascaded U-Nets is realized in the form of a recently developed projection-based updated data consistency layer. The novel structure is implemented in the PyTorch environment, which is one of the standards for deep learning implementations. The recently created fastMRI dataset which forms an important benchmark for MRI reconstruction is used for training and testing purposes. We present simulation results comparing the novel method with a variety of competitive deep networks. The new cascaded U-Net structures PSNR performance stands on average 1.28 dB higher than the baseline U-Net. The improvement, when compared to the standard CNN, is on average 3.32 dB. The proposed cascaded U-Net configuration results in an improved reconstruction performance when compared to the CNN, the cascaded CNN, and also the singular U-Net structures, where the singular U-Net forms the baseline reconstruction method from the fastMRI package. The use of the projection-based updated data consistency layer also leads to improved quantitative (including SSIM, PSNR, and NMSE results) and qualitative results when compared to the use of the conventional data consistency layer. [ABSTRACT FROM AUTHOR]

In this paper, we propose the fast and accurate registration method of partially scanned dental surfaces in a 3D dental laser scanning. To overcome the multiple point correspondence problems of conventional surface registration methods, we propose the novel depth map-based registration method to register 3D surface models. First, we convert a partially scanned 3D dental surface into a 2D image by generating the 2D depth map image of the surface model by applying a 3D rigid transformation into this model. Then, the image-based registration method using 2D depth map images accurately estimates the initial transformation between two consequently acquired surface models. To further increase the computational efficiency, we decompose the 3D rigid transformation into out-of-plane (i.e. x -, y -rotation, and z -translation) and in-plane (i.e. x -, y -translation, and z -rotation) transformations. For the in-plane transformation, we accelerate the transformation process by transforming the 2D depth map image instead of transforming the 3D surface model. For the more accurate registration of 3D surface models, we enhance iterative closest point (ICP) method for the subsequent fine registration. Our initial depth map-based registration well aligns each surface model. Therefore, our subsequent ICP method can accurately register two surface models since it is highly probable that the closest point pairs are the exact corresponding point pairs. The experimental results demonstrated that our method accurately registered partially scanned dental surfaces. Regarding the computational performance, our method delivered about 1.5 times faster registration than the conventional method. Our method can be successfully applied to the accurate reconstruction of 3D dental objects for orthodontic and prosthodontic treatment. [ABSTRACT FROM AUTHOR]

Abstract: This paper proposes a method for assessing the acrosome state of boar spermatozoa heads using digital image processing. We use gray level images in which spermatozoa have been labeled as acrosome-intact or acrosome damaged using the information of a coupled fluorescent image. The heads are segmented obtaining the outer head contour. A set of “n” inner contours separated by a logarithmic distance function is calculated later. For each point of the, in this case, seven contours a number of local texture features are computed. We have compared the classification performance of Relevance Learning Vector Quantization, class conditional means and KNN, employing cross-validation for the evaluation. Gradient magnitude data offer the best result with an overall test error of only 1%. This result outperforms previously applied methods and suggests this approach as an interesting automatized approach to this veterinarian problem. [Copyright &y& Elsevier]

Abstract: Although the health care sector has already been subjected to a major computerization effort, this effort is often limited to the implementation of standalone systems which do not communicate with each other. Interoperability problems limit health care applications from achieving their full potential. In this paper, we propose the use of Semantic Web technologies to solve interoperability problems between data providers. Through the development of unifying health care ontologies, data from multiple health care providers can be aggregated, which can then be used as input for a decision support system. This way, more data is taken into account than a single health care provider possesses in his local setting. The feasibility of our approach is demonstrated by the creation of an end-to-end proof of concept, focusing on Belgian health care providers and medicinal decision support. [Copyright &y& Elsevier]

Abstract: Single-photon emission computed tomography (SPECT) images alone are difficult to understand in diagnosis, since anatomical structures are absent from the data. Studies on combination attempt to locate functional changes of the SPECT image by the magnetic resonance (MR) image. Due to the low similarity between original images, fused results are always darkened, obscured or loss some crucial anatomical structures. This paper has solved these problems by the variable-weight matrix which is estimated by minimizing the cost function using the simplex method. Under the generalized intensity-hue-saturation (GIHS) framework, the multiscaled analysis is presented for a better detail preservation. Besides, interactive approaches are discussed for the gradual variation between original images and the control of detail performance. The similarity assessment evaluates several different methods on a normal brain atlas. Two clips show the interactive property of the proposed method, while two medical cases demonstrate its clinical values. We conclude that the proposed method is superior to traditional methods, when considering the definition and the information capacity of fused results. [Copyright &y& Elsevier]

Background and objectives: Epidemiological models of epidemic spread are an essential tool for optimizing decision-making. The current literature is very extensive and covers a wide variety of deterministic and stochastic models. However, with the increase in computing resources, new, more general, and flexible procedures based on simulation models can assess the effectiveness of measures and quantify the current state of the epidemic. This paper illustrates the potential of this approach to build a new dynamic probabilistic model to estimate the prevalence of SARS-CoV-2 infections in different compartments. Methods: We propose a new probabilistic model in which, for the first time in the epidemic literature, parameter learning is carried out using gradient-free stochastic black-box optimization techniques simulating multiple trajectories of the infection dynamics in a general way, solving an inverse problem that is defined employing the daily information from mortality records. Results : After the application of the new proposal in Spain in the first and successive waves, the result of the model confirms the accuracy to estimate the seroprevalence and allows us to know the real dynamics of the pandemic a posteriori to assess the impact of epidemiological measures by the Spanish government and to plan more efficiently the subsequent decisions with the prior knowledge obtained. Conclusions: The model results allow us to estimate the daily patterns of COVID-19 infections in Spain retrospectively and examine the population's exposure to the virus dynamically in contrast to seroprevalence surveys. Furthermore, given the flexibility of our simulation framework, we can model situations —even using non-parametric distributions between the different compartments in the model— that other models in the existing literature cannot. Our general optimization strategy remains valid in these cases, and we can easily create other non-standard simulation epidemic models that incorporate more complex and dynamic structures. [ABSTRACT FROM AUTHOR]

Recent advances in wearable technology have facilitated the non-obtrusive monitoring of physiological signals, creating opportunities to monitor and predict stress. Researchers have utilized machine learning methods using these physiological signals to develop stress prediction models. Many of these prediction models have utilized objective stressor tasks (e.g., a public speaking task or solving math problems). Alternatively, the subjective user responses with self-reports have also been used for measuring stress. In this paper, we describe a methodological approach (a) to compare the prediction performance of models developed using objective markers of stress using participant-reported subjective markers of stress from self-reports; and (b) to develop personalized stress models by accounting for inter-individual differences. Towards this end, we conducted a laboratory-based study with 32 healthy volunteers. Participants completed a series of stressor tasks—social, cognitive and physical—wearing an instrumented commercial smartwatch that collected physiological signals and participant responses using timed self-reports. After extensive data preprocessing using a combination of signal processing techniques, we developed two types of models: objective stress models using the stressor tasks as labels; and subjective stress models using participant responses to each task as the label for that stress task. We trained and tested several machine learning algorithms—support vector machine (SVM), random forest (RF), gradient boosted trees (GBT), AdaBoost, and Logistic Regression (LR)—and evaluated their performance. SVM had the best performance for the models using the objective stressor (i.e., stressor tasks) with an AUROC of 0.790 and an F-1 score of 0.623. SVM also had the highest performance for the models using the subjective stress (i.e., participant self-reports) with an AUROC of 0.719 and an F-1 score of 0.520. Model performance improved with a personalized threshold model to an AUROC of 0.751 and an F-1 score of 0.599. The performance of the stress models using an instrumented commercial smartwatch was comparable to similar models from other state-of-the-art laboratory-based studies. However, the subjective stress models had a lower performance, indicating the need for further research on the use of self-reports for stress-related studies. The improvement in performance with the personalized threshold-based models provide new directions for building stress prediction models. [Display omitted] • Instrumented a commercial smartwatch for stress modeling using multi-stage data processing. • Compared stress prediction models using objective task-based stressors with subjective EMAs. • Developed personalized stress models with a personalized threshold accounting for inter-individual differences in stress. [ABSTRACT FROM AUTHOR]

• A melanoma detection framework Mask-DenseNet+ is proposed based on the Mask-RCNN. It combines Mask R-CNN networks to add mask processing operations, introduces integrated learning ideas, and adjusts and improves the original DenseNet network structure. • Mask-DenseNet+ is compared with some ablation experiments. • The accuracy of our proposed method reaches 90.61%, the sensitivity reaches 78.00%, and the specificity reaches 93.43%. • The method is feasible and effective, and achieves the preliminary goal of melanoma detection. It is greatly improved the detection accuracy and reached the level of visual diagnosis of doctors. Melanoma is a type of malignant skin cancer with high mortality, and its incidence is increasing rapidly in recent years. At present, the best treatment is surgical resection after early diagnosis. However, due to the high visual similarity between melanoma and benign melanocytic nevus, coupled with the scarcity and imbalance of data, traditional methods are difficult to achieve good recognition and detection results. Similarly, many machine learning methods have been applied to the task of skin disease detection and classification. However, the accuracy and sensitivity of the experiments are still not satisfactory. Therefore, this paper proposed a method to identify melanoma more efficiently and accurately. We implemented a Mixed Skin Lesion Picture Generate method based on Mask R-CNN (MSLP-MR) to solve the problem of data imbalance. Besides, we designed a melanoma detection framework of Mask-DenseNet+ based on MSLP-MR. This method used Mask R-CNN to introduce the method of mask segmentation, and combined with the idea of ensemble learning to integrate multiple classifiers for weighted prediction. Compared with the ablation experiments, the accuracy, sensitivity and AUC of the proposed network classification are improved by 2.56%, 29.33% and 0.0345. The experimental results on the ISIC dataset shown that the accuracy of the algorithm is 90.61%, the sensitivity reaches 78.00%, which is higher than the original methods; the specificity reaches 93.43%; and the AUC reaches 0.9502. The method is feasible and effective, and achieves the preliminary goal of melanoma detection. It is greatly improved the detection accuracy and reached the level of visual diagnosis of doctors. [ABSTRACT FROM AUTHOR]

• A novel U-Net based architecture guided by the multi-scale attention mechanism with input image pyramid and deep supervised output layers is developed for cardiac segmentation in short-axis MRI images. • The Focal Tversky Loss function is incorporated into the attention mechanism gated U-Net, which can effectively tackle the problem of high imbalance between the background class and the target class. • The multi-scale input image pyramid is improved in order to obtain better intermediate image features. Background and Objective: Automatic cardiac segmentation plays an utmost role in the diagnosis and quantification of cardiovascular diseases. Methods: This paper proposes a new cardiac segmentation method in short-axis Magnetic Resonance Imaging (MRI) images, called attention U-Net architecture with input image pyramid and deep supervised output layers (AID), which can fully-automatically learn to pay attention to target structures of various sizes and shapes. During each training process, the model continues to learn how to emphasize the desired features and suppress irrelevant areas in the original images, effectively improving the accuracy of cardiac segmentation. At the same time, we introduce the Focal Tversky Loss (FTL), which can effectively solve the problem of high imbalance in the amount of data between the target class and the background class during cardiac image segmentation. In order to obtain a better representation of intermediate features, we add a multi-scale input pyramid to the attention network. Results: The proposed cardiac segmentation technique is tested on the public Left Ventricle Segmentation Challenge (LVSC) dataset, which is shown to achieve 0.75, 0.87 and 0.92 for Jaccard Index, Sensitivity and Specificity, respectively. Experimental results demonstrate that the proposed method is able to improve the segmentation accuracy compared with the standard U-Net, and achieves comparable performance to the most advanced fully-automated methods. Conclusions: Given its effectiveness and advantages, the proposed method can facilitate cardiac segmentation in short-axis MRI images in clinical practice. [ABSTRACT FROM AUTHOR]

• Training three U-nets in different directions (axial plane, coronal plane and sagittal plane) of TOF-MRA volumes. • Adopt the focal loss function to U-net to solve the problem of imbalance between the target points and background points in TOF-MRA images. • Proposed a voting method for two-dimensional and three-dimensional feature fusion. Accurate cerebrovascular segmentation plays an important role in the diagnosis of cerebrovascular diseases. Considering the complexity and uncertainty of doctors' manual segmentation of cerebral vessels, this paper proposed an automatic segmentation algorithm based on Multiple-U-net (M-U-net) to segment cerebral vessel structures from the Time-of-flight Magnetic Resonance Angiography (TOF-MRA) data. First, the TOF-MRA data was normalized by volume and then divided into three groups through slices of axial, coronal and sagittal directions respectively. Three single U-nets were trained by divided dataset. To solve the problem of uneven distribution of positive and negative samples, the focal loss function was adopted in training. After obtaining the prediction results of three single U-nets, the voting feature fusion and the post-processing process based on connected domain analysis would be performed. 95 volumes of TOF-MRA provided by the MIDAS platform were applied to the experiment, among which 20 volumes were treated as the training dataset, 5 volumes were used as the validation dataset and the remaining 70 volumes were divided into 10 groups to test the trained model respectively. Experiments showed that the proposed M-U-net based algorithm achieved 88.60% and 87.93% Dice Similarity Coefficient (DSC) on the verification dataset and testing dataset, which performed better than any single U-net. Compared with other existing algorithms, our algorithm reached the state of the art level. The feature fusion of three single U-nets could effectively complement the segmentation results. [ABSTRACT FROM AUTHOR]

• To our best knowledge, we are the first to use deep learning for fully automatic functional region annotation of liver, demonstrating its ability to meet the challenge of classifying hepatic segments. • The anatomy class-specific information was used to solve the problem of the class crossing between slices. • A 2.5D spatial adaptation input module to overcome the problem of inconsistency in slices spatial information was designed. • We manually annotated 112 CT cases from the Task08_HepaticVessel in the Medical segmentation Decathlon study [25] over 3 months, and validated this pipeline on our data. Automatic functional region annotation of liver should be very useful for preoperative planning of liver resection in the clinical domain. However, many traditional computer-aided annotation methods based on anatomical landmarks or the vascular tree often fail to extract accurate liver segments. Furthermore, these methods are difficult to fully automate and thus remain time-consuming. To address these issues, in this study we aim to develop a fully-automated approach for functional region annotation of liver using deep learning based on 2.5D class-aware deep neural networks with spatial adaptation. 112 CT scans were fed into our 2.5D class-aware deep neural network with spatial adaptation for automatic functional region annotation of liver. The proposed model was built upon the ResU-net architecture, which adaptively selected a stack of adjacent CT slices as input and, generating masks corresponding to the center slice, automatically annotated the liver functional region from abdominal CT images. Furthermore, to minimize the problem of class-level ambiguity between different slices, the anatomy class-specific information was used. The final algorithm performance for automatic functional region annotation of liver showed large overlap with that of manual reference segmentation. The dice similarity coefficient of hepatic segments achieved high scores and an average dice score of 0.882. The entire calculation time was quite fast (~5 s) compared to manual annotation (~2.5 hours). The proposed models described in this paper offer a feasible solution for fully-automated functional region annotation of liver from CT images. The experimental results demonstrated that the proposed method can attain a high average dice score and low computational time. Therefore, this work should allow for improved liver surgical resection planning by our precise segmentation and simple fully-automated method. [ABSTRACT FROM AUTHOR]