Search

Recent advances in self-supervision and constrastive learning have brought the performance of foundation models to unprecedented levels in a variety of tasks. Fueled by this progress, these models are becoming the prevailing approach for a wide array of real-world vision problems, including risk-sensitive and high-stakes applications. However, ensuring safe deployment in these scenarios requires a more comprehensive understanding of their uncertainty modeling capabilities, which has been barely explored. In this work, we delve into the behavior of vision and vision-language foundation models under Conformal Prediction (CP), a statistical framework that provides theoretical guarantees of marginal coverage of the true class. Across extensive experiments including popular vision classification benchmarks, well-known foundation vision models, and three CP methods, our findings reveal that foundation models are well-suited for conformalization procedures, particularly those integrating Vision Transformers. Furthermore, we show that calibrating the confidence predictions of these models leads to efficiency degradation of the conformal set on adaptive CP methods. In contrast, few-shot adaptation to downstream tasks generally enhances conformal scores, where we identify Adapters as a better conformable alternative compared to Prompt Learning strategies. Our empirical study identifies APS as particularly promising in the context of vision foundation models, as it does not violate the marginal coverage property across multiple challenging, yet realistic scenarios.

Vision-language foundation models, such as CLIP, have shown unprecedented zero-shot performance across a wide range of tasks. Nevertheless, these models may be unreliable under distributional shifts, as their performance is significantly degraded. In this work, we explore how to efficiently leverage class text information to mitigate these distribution drifts encountered by large pre-trained vision-language models (VLMs) during test-time inference. In particular, we propose to generate pseudo-labels for the test-time samples by exploiting generic class text embeddings as fixed centroids of a label assignment problem, which is efficiently solved with Optimal Transport. Furthermore, the proposed adaptation method (CLIP-OT) integrates a multiple template knowledge distillation approach, which replicates multi-view contrastive learning strategies in unsupervised representation learning but without incurring additional computational complexity. Extensive experiments on multiple popular test-time adaptation benchmarks presenting diverse complexity empirically show the superiority of CLIP-OT, achieving performance gains of up to 7% over recent state-of-the-art methods, yet being computationally and memory efficient.

Integrating image and text data through multi-modal learning has emerged as a new approach in medical imaging research, following its successful deployment in computer vision. While considerable efforts have been dedicated to establishing medical foundation models and their zero-shot transfer to downstream tasks, the popular few-shot setting remains relatively unexplored. Following on from the currently strong emergence of this setting in computer vision, we introduce the first structured benchmark for adapting medical vision-language models (VLMs) in a strict few-shot regime and investigate various adaptation strategies commonly used in the context of natural images. Furthermore, we evaluate a simple generalization of the linear-probe adaptation baseline, which seeks an optimal blending of the visual prototypes and text embeddings via learnable class-wise multipliers. Surprisingly, such a text-informed linear probe yields competitive performances in comparison to convoluted prompt-learning and adapter-based strategies, while running considerably faster and accommodating the black-box setting. Our extensive experiments span three different medical modalities and specialized foundation models, nine downstream tasks, and several state-of-the-art few-shot adaptation methods. We made our benchmark and code publicly available to trigger further developments in this emergent subject: \url{https://github.com/FereshteShakeri/few-shot-MedVLMs}., Comment: MICCAI 2024 (Spotlight) - Code is available at https://github.com/FereshteShakeri/few-shot-MedVLMs.git

The rising integration of variable renewable energy sources (RES), like solar and wind power, introduces considerable uncertainty in grid operations and energy management. Effective forecasting models are essential for grid operators to anticipate the net load - the difference between consumer electrical demand and renewable power generation. This paper proposes a deep learning (DL) model based on long short-term memory (LSTM) networks for net load forecasting in renewable-based microgrids, considering both solar and wind power. The model's architecture is detailed, and its performance is evaluated using a residential microgrid test case based on a typical meteorological year (TMY) dataset. The results demonstrate the effectiveness of the proposed LSTM-based DL model in predicting the net load, showcasing its potential for enhancing energy management in renewable-based microgrids., Comment: Submitted to 2024 IEEE North American Power Symposium

This paper addresses the critical issue of miscalibration in CLIP-based model adaptation, particularly in the challenging scenario of out-of-distribution (OOD) samples, which has been overlooked in the existing literature on CLIP adaptation. We empirically demonstrate that popular CLIP adaptation approaches, such as Adapters, Prompt Learning, and Test-Time Adaptation, substantially degrade the calibration capabilities of the zero-shot baseline in the presence of distributional drift. We identify the increase in logit ranges as the underlying cause of miscalibration of CLIP adaptation methods, contrasting with previous work on calibrating fully-supervised models. Motivated by these observations, we present a simple and model-agnostic solution to mitigate miscalibration, by scaling the logit range of each sample to its zero-shot prediction logits. We explore three different alternatives to achieve this, which can be either integrated during adaptation or directly used at inference time. Comprehensive experiments on popular OOD classification benchmarks demonstrate the effectiveness of the proposed approaches in mitigating miscalibration while maintaining discriminative performance, whose improvements are consistent across the three families of these increasingly popular approaches. The code is publicly available at: https://github.com/Bala93/CLIPCalib, Comment: ECCV 2024

We present a pioneering investigation into the application of deep learning techniques to analyze histopathological images for addressing the substantial challenge of automated prognostic prediction. Prognostic prediction poses a unique challenge as the ground truth labels are inherently weak, and the model must anticipate future events that are not directly observable in the image. To address this challenge, we propose a novel three-part framework comprising of a convolutional network based tissue segmentation algorithm for region of interest delineation, a contrastive learning module for feature extraction, and a nested multiple instance learning classification module. Our study explores the significance of various regions of interest within the histopathological slides and exploits diverse learning scenarios. The pipeline is initially validated on artificially generated data and a simpler diagnostic task. Transitioning to prognostic prediction, tasks become more challenging. Employing bladder cancer as use case, our best models yield an AUC of 0.721 and 0.678 for recurrence and treatment outcome prediction respectively., Comment: https://github.com/Biomedical-Data-Analysis-Laboratory/HistoPrognostics

In this work, we present a novel approach to calibrate segmentation networks that considers the inherent challenges posed by different categories and object regions. In particular, we present a formulation that integrates class and region-wise constraints into the learning objective, with multiple penalty weights to account for class and region differences. Finding the optimal penalty weights manually, however, might be unfeasible, and potentially hinder the optimization process. To overcome this limitation, we propose an approach based on Class and Region-Adaptive constraints (CRaC), which allows to learn the class and region-wise penalty weights during training. CRaC is based on a general Augmented Lagrangian method, a well-established technique in constrained optimization. Experimental results on two popular segmentation benchmarks, and two well-known segmentation networks, demonstrate the superiority of CRaC compared to existing approaches. The code is available at: https://github.com/Bala93/CRac/, Comment: Under review

As NASA prepares to carry out its Artemis lunar missions, the design and planning of robust power systems tailored to the lunar environment become necessary and urgent. Solar photovoltaic (PV) systems are among the most suitable power generators for lunar applications given the abundant solar irradiance the lunar surface receives as a result of the lack of an atmosphere. However, the vastly different environmental conditions of the moon compared to those on Earth call for special reconsiderations to the traditional PV power output models used for terrestrial applications. The substantially wider range of temperatures on the moon combined with the absence of atmosphere implies the PV will operate at very different thermal conditions. Therefore, this paper proposes a PV power output model that determines PV cell temperature on the lunar surface based on lunar ambient temperature as well as solar irradiance, while also capturing these special lunar conditions. The model is based on thermal exchange via the three main heat transfer paths: heat radiation, absorption, and convection. Moreover, four different array configurations are presented and compared to determine the most suitable choice based on energy generation and complexity for different locations around the moon.

Using deep learning models pre-trained on Imagenet is the traditional solution for medical image classification to deal with data scarcity. Nevertheless, relevant literature supports that this strategy may offer limited gains due to the high dissimilarity between domains. Currently, the paradigm of adapting domain-specialized foundation models is proving to be a promising alternative. However, how to perform such knowledge transfer, and the benefits and limitations it presents, are under study. The CGI-HRDC challenge for Hypertensive Retinopathy diagnosis on fundus images introduces an appealing opportunity to evaluate the transferability of a recently released vision-language foundation model of the retina, FLAIR. In this work, we explore the potential of using FLAIR features as starting point for fundus image classification, and we compare its performance with regard to Imagenet initialization on two popular transfer learning methods: Linear Probing (LP) and Fine-Tuning (FP). Our empirical observations suggest that, in any case, the use of the traditional strategy provides performance gains. In contrast, direct transferability from FLAIR model allows gains of 2.5%. When fine-tuning the whole network, the performance gap increases up to 4%. In this case, we show that avoiding feature deterioration via LP initialization of the classifier allows the best re-use of the rich pre-trained features. Although direct transferability using LP still offers limited performance, we believe that foundation models such as FLAIR will drive the evolution of deep-learning-based fundus image analysis., Comment: CGI 2023

The development of computer vision solutions for gigapixel images in digital pathology is hampered by significant computational limitations due to the large size of whole slide images. In particular, digitizing biopsies at high resolutions is a time-consuming process, which is necessary due to the worsening results from the decrease in image detail. To alleviate this issue, recent literature has proposed using knowledge distillation to enhance the model performance at reduced image resolutions. In particular, soft labels and features extracted at the highest magnification level are distilled into a model that takes lower-magnification images as input. However, this approach fails to transfer knowledge about the most discriminative image regions in the classification process, which may be lost when the resolution is decreased. In this work, we propose to distill this information by incorporating attention maps during training. In particular, our formulation leverages saliency maps of the target class via grad-CAMs, which guides the lower-resolution Student model to match the Teacher distribution by minimizing the l2 distance between them. Comprehensive experiments on prostate histology image grading demonstrate that the proposed approach substantially improves the model performance across different image resolutions compared to previous literature., Comment: EUSIPCO 2023 - Code available at https://github.com/cvblab/Mitosis-UTS

Efficient transfer learning (ETL) is receiving increasing attention to adapt large pre-trained language-vision models on downstream tasks with a few labeled samples. While significant progress has been made, we reveal that state-of-the-art ETL approaches exhibit strong performance only in narrowly-defined experimental setups, and with a careful adjustment of hyperparameters based on a large corpus of labeled samples. In particular, we make two interesting, and surprising empirical observations. First, to outperform a simple Linear Probing baseline, these methods require to optimize their hyper-parameters on each target task. And second, they typically underperform -- sometimes dramatically -- standard zero-shot predictions in the presence of distributional drifts. Motivated by the unrealistic assumptions made in the existing literature, i.e., access to a large validation set and case-specific grid-search for optimal hyperparameters, we propose a novel approach that meets the requirements of real-world scenarios. More concretely, we introduce a CLass-Adaptive linear Probe (CLAP) objective, whose balancing term is optimized via an adaptation of the general Augmented Lagrangian method tailored to this context. We comprehensively evaluate CLAP on a broad span of datasets and scenarios, demonstrating that it consistently outperforms SoTA approaches, while yet being a much more efficient alternative., Comment: CVPR 2024. Code: https://github.com/jusiro/CLAP

The water-energy nexus encompasses the interdependencies between water and energy resources identifying the existing links between the production and distribution of these resources. Therefore, understanding the water-energy nexus is crucial for developing sustainable and integrated resource management approaches. This paper proposes a decentralized co-optimization model for a micro water-energy nexus system (MWEN), aiming to optimize the combined supply of both resources to end consumers. The approach respects the separate ownership and management of the water and energy sectors while bridging the gap between their optimized operations. An enhanced version of the alternating direction method of multipliers (ADMM) is proposed, the objective-based ADMM (OB-ADMM), which is able to robustly optimize each system independently towards a common objective, only sharing information about the power consumption of water management, providing privacy for each resource provider.

Foundation vision-language models are currently transforming computer vision, and are on the rise in medical imaging fueled by their very promising generalization capabilities. However, the initial attempts to transfer this new paradigm to medical imaging have shown less impressive performances than those observed in other domains, due to the significant domain shift and the complex, expert domain knowledge inherent to medical-imaging tasks. Motivated by the need for domain-expert foundation models, we present FLAIR, a pre-trained vision-language model for universal retinal fundus image understanding. To this end, we compiled 38 open-access, mostly categorical fundus imaging datasets from various sources, with up to 101 different target conditions and 288,307 images. We integrate the expert's domain knowledge in the form of descriptive textual prompts, during both pre-training and zero-shot inference, enhancing the less-informative categorical supervision of the data. Such a textual expert's knowledge, which we compiled from the relevant clinical literature and community standards, describes the fine-grained features of the pathologies as well as the hierarchies and dependencies between them. We report comprehensive evaluations, which illustrate the benefit of integrating expert knowledge and the strong generalization capabilities of FLAIR under difficult scenarios with domain shifts or unseen categories. When adapted with a lightweight linear probe, FLAIR outperforms fully-trained, dataset-focused models, more so in the few-shot regimes. Interestingly, FLAIR outperforms by a wide margin larger-scale generalist image-language models and retina domain-specific self-supervised networks, which emphasizes the potential of embedding experts' domain knowledge and the limitations of generalist models in medical imaging., Comment: Accepted in Medical Image Analysis. The pre-trained model is available at: https://github.com/jusiro/FLAIR

Digital pathology has become a standard in the pathology workflow due to its many benefits. These include the level of detail of the whole slide images generated and the potential immediate sharing of cases between hospitals. Recent advances in deep learning-based methods for image analysis make them of potential aid in digital pathology. However, a major limitation in developing computer-aided diagnostic systems for pathology is the lack of an intuitive and open web application for data annotation. This paper proposes a web service that efficiently provides a tool to visualize and annotate digitized histological images. In addition, to show and validate the tool, in this paper we include a use case centered on the diagnosis of spindle cell skin neoplasm for multiple annotators. A usability study of the tool is also presented, showing the feasibility of the developed tool., Comment: 11 pages, 9 figures, 6 tables

This paper proposes a decentralized energy management (DEM) strategy for a network of local microgrids, providing economically balanced energy schedules for all participating microgrids. The proposed DEM strategy can preserve the privacy of each microgrid by only requiring them to share network power exchange information. The proposed DEM strategy enhances the traditional alternating direction method of multipliers (ADMM) formulation for networks of microgrids by examining the global objective value as well as the solution quality. A novel stopping decision combining these two metrics is proposed for the enhanced objective-based ADMM (OB-ADMM) method. This paper also presents a centralized energy management (CEM) model as a benchmark, and a post-processing proportional exchange algorithm (PEA) to balance the economic benefit of each microgrid. The resulting proposed OB-ADMM model combined with the PEA delivers the final high-fidelity optimal solution for multiple microgrids in a grid-connected collaborative power exchange network. Moreover, the proposed decentralized operational strategy preserves the economic and privacy interests of individual microgrid participants. Different network cases are simulated to test the algorithm's performance, and the results validate the aforementioned claims., Comment: 10 Pages

The recent popularity of foundation models and the pre-train-and-adapt paradigm, where a large-scale model is transferred to downstream tasks, is gaining attention for volumetric medical image segmentation. However, current transfer learning strategies devoted to full fine-tuning for transfer learning may require significant resources and yield sub-optimal results when the labeled data of the target task is scarce. This makes its applicability in real clinical settings challenging since these institutions are usually constrained on data and computational resources to develop proprietary solutions. To address this challenge, we formalize Few-Shot Efficient Fine-Tuning (FSEFT), a novel and realistic scenario for adapting medical image segmentation foundation models. This setting considers the key role of both data- and parameter- efficiency during adaptation. Building on a foundation model pre-trained on open-access CT organ segmentation sources, we propose leveraging Parameter-Efficient Fine-Tuning and black-box Adapters to address such challenges. Furthermore, novel efficient adaptation methodologies are introduced in this work, which include Spatial black-box Adapters that are more appropriate for dense prediction tasks and constrained transductive inference, leveraging task-specific prior knowledge. Our comprehensive transfer learning experiments confirm the suitability of foundation models in medical image segmentation and unveil the limitations of popular fine-tuning strategies in few-shot scenarios., Comment: Journal Extension of MICCAI - MedAGI Workshop 2023. Code in https://github.com/jusiro/fewshot-finetuning

This paper investigates how to efficiently and economically deliver offshore energy generated by offshore wind turbines to onshore. Both power cables and hydrogen energy facilities are analyzed. Each method is examined with the associated proposed optimal sizing model. A cost-benefit analysis is conducted, and these methods are compared under different scenarios. Three long-distance energy transmission methods are proposed to transmit offshore energy onshore, considering the capital/operation costs for related energy conversion and transmission facilities. The first method that deploys power cables only is a benchmark method. The other two methods utilize a hydrogen supercenter (HSC) and transmit offshore energy to and onshore substation through hydrogen pipelines. For the second method, electrolyzers are placed at offshore wind farms connected to HSC with low-pressure hydrogen pipelines. For the third method, offshore wind power is distributed to the HSC and then converted into hydrogen using electrolyzers placed at the HSC. An offshore scenario including three wind farms located in the Gulf of Mexico and one onshore substation is used as the base test case. Numerical simulations are conducted over a planning period of 30 years using the proposed optimization models for each method separately. Simulation results show that the proposed hydrogen-based methods can effectively transmit offshore energy to the onshore substation. The distances between the wind farms can influence the selection and configuration of the offshore wind energy system. Moreover, the proposed study analyzes the cost and benefits for various systems with different energy carriers. It can also suggest which method could be the best candidate for offshore energy delivery system under which scenario and demonstrate a promising future for offshore wind power generation.

Mitotic activity is a crucial proliferation biomarker for the diagnosis and prognosis of different types of cancers. Nevertheless, mitosis counting is a cumbersome process for pathologists, prone to low reproducibility, due to the large size of augmented biopsy slides, the low density of mitotic cells, and pattern heterogeneity. To improve reproducibility, deep learning methods have been proposed in the last years using convolutional neural networks. However, these methods have been hindered by the process of data labelling, which usually solely consist of the mitosis centroids. Therefore, current literature proposes complex algorithms with multiple stages to refine the labels at pixel level, and to reduce the number of false positives. In this work, we propose to avoid complex scenarios, and we perform the localization task in a weakly supervised manner, using only image-level labels on patches. The results obtained on the publicly available TUPAC16 dataset are competitive with state-of-the-art methods, using only one training phase. Our method achieves an F1-score of 0.729 and challenges the efficiency of previous methods, which required multiple stages and strong mitosis location information., Comment: Presented at IDEAL 2022

This paper proposes a Networked Micro Water-Energy Nexus (NetMicroWEN) capable of co-optimizing and simultaneously supplying water and energy to local consumers in nearby communities. The system manages different water and energy inputs of different communities in a local network to cooperatively meet their demands. This paper considers a centralized network topology that connects all members of the network under one control system. This paper also proposes a Proportional Exchange Algorithm (PEA) that allows members to benefit equally from exchanging both resources among other members of the NetMicroWEN and the main water and power distribution systems. The co-optimization model is a mixed-integer linear program, involving all necessary power and water related constraints for the network to achieve a feasible and practical solution. The economic benefits of the NetMicroWEN are illustrated by a comparison with separate Micro Water- Energy Nexus (MWEN) systems meeting their own demands individually. The case studies demonstrate that the proposed NetMicroWEN achieves substantially lower operating costs compared to the operation of separate MWEN systems., Comment: 6 Pages. 3 Figures, 4 Tables

This paper proposes a novel resilience index, a microgrid survivability rate (SR) under extreme events, and then proposes a novel Resilient Operational Planning (ROP) algorithm to maximize the proposed resilience index SR. The proposed ROP algorithm can incorporate predetermined inverter failure probabilities and generate multiple scenarios accordingly to optimize resilient operations during an extreme event. The implemented ROP algorithm consists of two main steps: (i) optimization of resilient operational planning, and (ii) preventive resilience enhancement if minimum SR is not met per the analysis in step 1. A typical microgrid (MG) is studied to compare the proposed ROP algorithm against a traditional microgrid energy management (MEM) model. Results indicate that an enhanced resilience operation is achieved by the ROP algorithm, which is demonstrated by the quantification of resilience via the SR. Moreover, the proposed ROP algorithm is able to obtain a greater SR overall compared to that achieved by the traditional MEM, and this benefit of using the proposed ROP increases as the inverter failure probabilities increase., Comment: 7 pages, 4 figures

Current unsupervised anomaly localization approaches rely on generative models to learn the distribution of normal images, which is later used to identify potential anomalous regions derived from errors on the reconstructed images. However, a main limitation of nearly all prior literature is the need of employing anomalous images to set a class-specific threshold to locate the anomalies. This limits their usability in realistic scenarios, where only normal data is typically accessible. Despite this major drawback, only a handful of works have addressed this limitation, by integrating supervision on attention maps during training. In this work, we propose a novel formulation that does not require accessing images with abnormalities to define the threshold. Furthermore, and in contrast to very recent work, the proposed constraint is formulated in a more principled manner, leveraging well-known knowledge in constrained optimization. In particular, the equality constraint on the attention maps in prior work is replaced by an inequality constraint, which allows more flexibility. In addition, to address the limitations of penalty-based functions we employ an extension of the popular log-barrier methods to handle the constraint. Last, we propose an alternative regularization term that maximizes the Shannon entropy of the attention maps, reducing the amount of hyperparameters of the proposed model. Comprehensive experiments on two publicly available datasets on brain lesion segmentation demonstrate that the proposed approach substantially outperforms relevant literature, establishing new state-of-the-art results for unsupervised lesion segmentation, and without the need to access anomalous images., Comment: arXiv admin note: substantial text overlap with arXiv:2109.00482 Accepted at Medical Image Analysis

This paper presents an acoustic echo canceler based on a U-Net convolutional neural network for single-talk and double-talk scenarios. U-Net networks have previously been used in the audio processing area for source separation problems because of their ability to reproduce the finest details of audio signals, but to our knowledge, this is the first time they have been used for acoustic echo cancellation (AEC). The U-Net hyperparameters have been optimized to obtain the best AEC performance, but using a reduced number of parameters to meet a latency restriction of 40 ms. The training and testing of our model have been carried out within the framework of the 'ICASSP 2021 AEC Challenge' organized by Microsoft. We have trained the optimized U-Net model with a synthetic dataset only (S-U-Net) and with a synthetic dataset and the single-talk set of a real dataset (SR-U-Net), both datasets were released for the challenge. The S-U-Net model presented better results for double-talk scenarios, thus their inferred near-end signals from the blind testset were submitted to the challenge. Our canceler ranked 12th among 17 teams, and 5th among 10 academia teams, obtaining an overall mean opinion score of 3.57., Comment: 6 pages, 2 figures, submitted to the 2021 IEEE International Conference on Acoustics, Speech and Signal Processing on October 2020

Current unsupervised anomaly localization approaches rely on generative models to learn the distribution of normal images, which is later used to identify potential anomalous regions derived from errors on the reconstructed images. However, a main limitation of nearly all prior literature is the need of employing anomalous images to set a class-specific threshold to locate the anomalies. This limits their usability in realistic scenarios, where only normal data is typically accessible. Despite this major drawback, only a handful of works have addressed this limitation, by integrating supervision on attention maps during training. In this work, we propose a novel formulation that does not require accessing images with abnormalities to define the threshold. Furthermore, and in contrast to very recent work, the proposed constraint is formulated in a more principled manner, leveraging well-known knowledge in constrained optimization. In particular, the equality constraint on the attention maps in prior work is replaced by an inequality constraint, which allows more flexibility. In addition, to address the limitations of penalty-based functions we employ an extension of the popular log-barrier methods to handle the constraint. Comprehensive experiments on the popular BRATS'19 dataset demonstrate that the proposed approach substantially outperforms relevant literature, establishing new state-of-the-art results for unsupervised lesion segmentation., Comment: Accepted at BMVC'21

Prostate cancer is one of the most prevalent cancers worldwide. One of the key factors in reducing its mortality is based on early detection. The computer-aided diagnosis systems for this task are based on the glandular structural analysis in histology images. Hence, accurate gland detection and segmentation is crucial for a successful prediction. The methodological basis of this work is a prostate gland segmentation based on U-Net convolutional neural network architectures modified with residual and multi-resolution blocks, trained using data augmentation techniques. The residual configuration outperforms in the test subset the previous state-of-the-art approaches in an image-level comparison, reaching an average Dice Index of 0.77.

The Gleason scoring system is the primary diagnostic and prognostic tool for prostate cancer. In recent years, with the development of digitisation devices, the use of computer vision techniques for the analysis of biopsies has increased. However, to the best of the authors' knowledge, the development of algorithms to automatically detect individual cribriform patterns belonging to Gleason grade 4 has not yet been studied in the literature. The objective of the work presented in this paper is to develop a deep-learning-based system able to support pathologists in the daily analysis of prostate biopsies. The methodological core of this work is a patch-wise predictive model based on convolutional neural networks able to determine the presence of cancerous patterns. In particular, we train from scratch a simple self-design architecture. The cribriform pattern is detected by retraining the set of filters of the last convolutional layer in the network. From the reconstructed prediction map, we compute the percentage of each Gleason grade in the tissue to feed a multi-layer perceptron which provides a biopsy-level score.mIn our SICAPv2 database, composed of 182 annotated whole slide images, we obtained a Cohen's quadratic kappa of 0.77 in the test set for the patch-level Gleason grading with the proposed architecture trained from scratch. Our results outperform previous ones reported in the literature. Furthermore, this model reaches the level of fine-tuned state-of-the-art architectures in a patient-based four groups cross validation. In the cribriform pattern detection task, we obtained an area under ROC curve of 0.82. Regarding the biopsy Gleason scoring, we achieved a quadratic Cohen's Kappa of 0.81 in the test subset. Shallow CNN architectures trained from scratch outperform current state-of-the-art methods for Gleason grades classification.

Prostate cancer is one of the main diseases affecting men worldwide. The Gleason scoring system is the primary diagnostic tool for prostate cancer. This is obtained via the visual analysis of cancerous patterns in prostate biopsies performed by expert pathologists, and the aggregation of the main Gleason grades in a combined score. Computer-aided diagnosis systems allow to reduce the workload of pathologists and increase the objectivity. Recently, efforts have been made in the literature to develop algorithms aiming the direct estimation of the global Gleason score at biopsy/core level with global labels. However, these algorithms do not cover the accurate localization of the Gleason patterns into the tissue. In this work, we propose a deep-learning-based system able to detect local cancerous patterns in the prostate tissue using only the global-level Gleason score during training. The methodological core of this work is the proposed weakly-supervised-trained convolutional neural network, WeGleNet, based on a multi-class segmentation layer after the feature extraction module, a global-aggregation, and the slicing of the background class for the model loss estimation during training. We obtained a Cohen's quadratic kappa (k) of 0.67 for the pixel-level prediction of cancerous patterns in the validation cohort. We compared the model performance for semantic segmentation of Gleason grades with supervised state-of-the-art architectures in the test cohort. We obtained a pixel-level k of 0.61 and a macro-averaged f1-score of 0.58, at the same level as fully-supervised methods. Regarding the estimation of the core-level Gleason score, we obtained a k of 0.76 and 0.67 between the model and two different pathologists. WeGleNet is capable of performing the semantic segmentation of Gleason grades similarly to fully-supervised methods without requiring pixel-level annotations.

Prostate cancer is one of the main diseases affecting men worldwide. The gold standard for diagnosis and prognosis is the Gleason grading system. In this process, pathologists manually analyze prostate histology slides under microscope, in a high time-consuming and subjective task. In the last years, computer-aided-diagnosis (CAD) systems have emerged as a promising tool that could support pathologists in the daily clinical practice. Nevertheless, these systems are usually trained using tedious and prone-to-error pixel-level annotations of Gleason grades in the tissue. To alleviate the need of manual pixel-wise labeling, just a handful of works have been presented in the literature. Motivated by this, we propose a novel weakly-supervised deep-learning model, based on self-learning CNNs, that leverages only the global Gleason score of gigapixel whole slide images during training to accurately perform both, grading of patch-level patterns and biopsy-level scoring. To evaluate the performance of the proposed method, we perform extensive experiments on three different external datasets for the patch-level Gleason grading, and on two different test sets for global Grade Group prediction. We empirically demonstrate that our approach outperforms its supervised counterpart on patch-level Gleason grading by a large margin, as well as state-of-the-art methods on global biopsy-level scoring. Particularly, the proposed model brings an average improvement on the Cohen's quadratic kappa (k) score of nearly 18% compared to full-supervision for the patch-level Gleason grading task.