Valery Naranjo

Julio Silva-Rodríguez, Valery Naranjo, Jose Dolz

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.

Zahra Tabatabaei, Adrian colomer, Kjersti Engan et al.

According to GLOBOCAN 2020, prostate cancer is the second most common cancer in men worldwide and the fourth most prevalent cancer overall. For pathologists, grading prostate cancer is challenging, especially when discriminating between Grade 3 (G3) and Grade 4 (G4). This paper proposes a Self-Supervised Learning (SSL) framework to classify prostate histopathological images when labeled images are scarce. In particular, a tailored Convolutional Auto Encoder (CAE) is trained to reconstruct 128x128x3 patches of prostate cancer Whole Slide Images (WSIs) as a pretext task. The downstream task of the proposed SSL paradigm is the automatic grading of histopathological patches of prostate cancer. The presented framework reports promising results on the validation set, obtaining an overall accuracy of 83% and on the test set, achieving an overall accuracy value of 76% with F1-score of 77% in G4.

Claudio Fernandez-Martín, Umay Kiraz, Julio Silva-Rodríguez et al.

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.

Cristian Camilo Pulgarín-Ospina, Rocío del Amor, Adrián Colomera et al.

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.

Josep Zaragoza-Paredes, Javier Naranjo-Alcazar, Valery Naranjo et al.

Acoustic scene classification is an automatic listening problem that aims to assign an audio recording to a pre-defined scene based on its audio data. Over the years (and in past editions of the DCASE) this problem has often been solved with techniques known as ensembles (use of several machine learning models to combine their predictions in the inference phase). While these solutions can show performance in terms of accuracy, they can be very expensive in terms of computational capacity, making it impossible to deploy them in IoT devices. Due to the drift in this field of study, this task has two limitations in terms of model complexity. It should be noted that there is also the added complexity of mismatching devices (the audios provided are recorded by different sources of information). This technical report makes a comparative study of two different network architectures: conventional CNN and Conv-mixer. Although both networks exceed the baseline required by the competition, the conventional CNN shows a higher performance, exceeding the baseline by 8 percentage points. Solutions based on Conv-mixer architectures show worse performance although they are much lighter solutions.

Pablo Meseguer, Rocío del Amor, Valery Naranjo

Vision-language supervision has made remarkable strides in learning visual representations from textual guidance. In digital pathology, vision-language models (VLM), pre-trained on curated datasets of histological image-captions, have been adapted to downstream tasks, such as region of interest classification. Zero-shot transfer for slide-level prediction has been formulated by MI-Zero, but it exhibits high variability depending on the textual prompts. Inspired by prototypical learning, we propose MI-VisionShot, a training-free adaptation method on top of VLMs to predict slide-level labels in few-shot learning scenarios. Our framework takes advantage of the excellent representation learning of VLM to create prototype-based classifiers under a multiple-instance setting by retrieving the most discriminative patches within each slide. Experimentation through different settings shows the ability of MI-VisionShot to surpass zero-shot transfer with lower variability, even in low-shot scenarios. Code coming soon at thttps://github.com/cvblab/MIVisionShot.

Ilán Carretero, Roshni Mahtani, Silvia Perez-Deben et al.

Accurate diagnosis of spitzoid tumors (ST) is critical to ensure a favorable prognosis and to avoid both under- and over-treatment. Epigenetic data, particularly DNA methylation, provide a valuable source of information for this task. However, prior studies assume complete data, an unrealistic setting as methylation profiles frequently contain missing entries due to limited coverage and experimental artifacts. Our work challenges these favorable scenarios and introduces ReMAC, an extension of ReMasker designed to tackle classification tasks on high-dimensional data under complete and incomplete regimes. Evaluation on real clinical data demonstrates that ReMAC achieves strong and robust performance compared to competing classification methods in the stratification of ST. Code is available: https://github.com/roshni-mahtani/ReMAC.

Ilán Carretero, Pablo Meseguer, Rocío del Amor et al.

Domain shift in the field of histopathological imaging is a common phenomenon due to the intra- and inter-hospital variability of staining and digitization protocols. The implementation of robust models, capable of creating generalized domains, represents a need to be solved. In this work, a new domain adaptation method to deal with the variability between histopathological images from multiple centers is presented. In particular, our method adds a training constraint to the supervised contrastive learning approach to achieve domain adaptation and improve inter-class separability. Experiments performed on domain adaptation and classification of whole-slide images of six skin cancer subtypes from two centers demonstrate the method's usefulness. The results reflect superior performance compared to not using domain adaptation after feature extraction or staining normalization.

Lucía Gómez-Zaragozá, Rocío del Amor, María José Castro-Bleda et al.

Spontaneous datasets for Speech Emotion Recognition (SER) are scarce and frequently derived from laboratory environments or staged scenarios, such as TV shows, limiting their application in real-world contexts. We developed and publicly released the Emotional Voice Messages (EMOVOME) dataset, including 999 voice messages from real conversations of 100 Spanish speakers on a messaging app, labeled in continuous and discrete emotions by expert and non-expert annotators. We evaluated speaker-independent SER models using acoustic features as baseline and transformer-based models. We compared the results with reference datasets including acted and elicited speech, and analyzed the influence of annotators and gender fairness. The pre-trained UniSpeech-SAT-Large model achieved the highest results, 61.64% and 55.57% Unweighted Accuracy (UA) for 3-class valence and arousal prediction respectively on EMOVOME, a 10% improvement over baseline models. For the emotion categories, 42.58% UA was obtained. EMOVOME performed lower than the acted RAVDESS dataset. The elicited IEMOCAP dataset also outperformed EMOVOME in predicting emotion categories, while similar results were obtained in valence and arousal. EMOVOME outcomes varied with annotator labels, showing better results and fairness when combining expert and non-expert annotations. This study highlights the gap between controlled and real-life scenarios, supporting further advancements in recognizing genuine emotions.

Pablo Meseguer, Rocío del Amor, Adrian Colomer et al.

Since the emergence of the ImageNet dataset, the pretraining and fine-tuning approach has become widely adopted in computer vision due to the ability of ImageNet-pretrained models to learn a wide variety of visual features. However, a significant challenge arises when adapting these models to domain-specific fields, such as digital pathology, due to substantial gaps between domains. To address this limitation, foundation models (FM) have been trained on large-scale in-domain datasets to learn the intricate features of histopathology images. In cancer diagnosis, whole-slide image (WSI) prediction is essential for patient prognosis, and multiple instance learning (MIL) has been implemented to handle the giga-pixel size of WSI. As MIL frameworks rely on patch-level feature aggregation, this work aims to compare the performance of various feature extractors developed under different pretraining strategies for cancer subtyping on WSI under a MIL framework. Results demonstrate the ability of foundation models to surpass ImageNet-pretrained models for the prediction of six skin cancer subtypes

Lucía Gómez Zaragozá, Rocío del Amor, Elena Parra Vargas et al.

Emotional Voice Messages (EMOVOME) is a spontaneous speech dataset containing 999 audio messages from real conversations on a messaging app from 100 Spanish speakers, gender balanced. Voice messages were produced in-the-wild conditions before participants were recruited, avoiding any conscious bias due to laboratory environment. Audios were labeled in valence and arousal dimensions by three non-experts and two experts, which were then combined to obtain a final label per dimension. The experts also provided an extra label corresponding to seven emotion categories. To set a baseline for future investigations using EMOVOME, we implemented emotion recognition models using both speech and audio transcriptions. For speech, we used the standard eGeMAPS feature set and support vector machines, obtaining 49.27% and 44.71% unweighted accuracy for valence and arousal respectively. For text, we fine-tuned a multilingual BERT model and achieved 61.15% and 47.43% unweighted accuracy for valence and arousal respectively. This database will significantly contribute to research on emotion recognition in the wild, while also providing a unique natural and freely accessible resource for Spanish.

Pablo Meseguer, Rocío del Amor, Valery Naranjo

Vision language models (VLM) pre-trained on datasets of histopathological image-caption pairs enabled zero-shot slide-level classification. The ability of VLM image encoders to extract discriminative features also opens the door for supervised fine-tuning for whole-slide image (WSI) classification, ideally using few labeled samples. Slide-level prediction frameworks require the incorporation of multiple instance learning (MIL) due to the gigapixel size of the WSI. Following patch-level feature extraction and aggregation, MIL frameworks rely on linear classifiers trained on top of the slide-level aggregated features. Classifier weight initialization has a large influence on Linear Probing performance in efficient transfer learning (ETL) approaches based on few-shot learning. In this work, we propose Zero-Shot Multiple-Instance Learning (ZS-MIL) to address the limitations of random classifier initialization that underperform zero-shot prediction in MIL problems. ZS-MIL uses the class-level embeddings of the VLM text encoder as the classification layer's starting point to compute each sample's bag-level probabilities. Through multiple experiments, we demonstrate the robustness of ZS-MIL compared to well-known weight initialization techniques both in terms of performance and variability in an ETL few-shot scenario for subtyping prediction.

Santiago Moreno, Pablo Meseguer, Rocío del Amor et al.

Accurate annotation of cutaneous neoplasm biopsies represents a major challenge due to their wide morphological variability, overlapping histological patterns, and the subtle distinctions between benign and malignant lesions. Vision-language foundation models (VLMs), pre-trained on paired image-text corpora, learn joint representations that bridge visual features and diagnostic terminology, enabling zero-shot localization and classification of tissue regions without pixel-level labels. However, most existing VLM applications in histopathology remain limited to slide-level tasks or rely on coarse interactive prompts, and they struggle to produce fine-grained segmentations across gigapixel whole-slide images (WSIs). In this work, we introduce a zero-shot visual-language segmentation pipeline for whole-slide images (ZEUS), a fully automated, zero-shot segmentation framework that leverages class-specific textual prompt ensembles and frozen VLM encoders to generate high-resolution tumor masks in WSIs. By partitioning each WSI into overlapping patches, extracting visual embeddings, and computing cosine similarities against text prompts, we generate a final segmentation mask. We demonstrate competitive performance on two in-house datasets, primary spindle cell neoplasms and cutaneous metastases, highlighting the influence of prompt design, domain shifts, and institutional variability in VLMs for histopathology. ZEUS markedly reduces annotation burden while offering scalable, explainable tumor delineation for downstream diagnostic workflows.

Andrés Laveda-Martínez, Natalia P. García-de-la-Puente, Fernando García-Torres et al.

Non-destructive testing (NDT) is essential in ceramic manufacturing to ensure the quality of components without compromising their integrity. In this context, Optical Coherence Tomography (OCT) enables high-resolution internal imaging, revealing defects such as pores, delaminations, or inclusions. This paper presents an automatic defect detection system based on Deep Learning (DL), trained on OCT images with manually segmented annotations. A neural network based on the U-Net architecture is developed, evaluating multiple experimental configurations to enhance its performance. Post-processing techniques enable both quantitative and qualitative evaluation of the predictions. The system shows an accurate behavior of 0.979 Dice Score, outperforming comparable studies. The inference time of 18.98 seconds per volume supports its viability for detecting inclusions, enabling more efficient, reliable, and automated quality control.

N. P. García-de-la-Puente, Rocío del Amor, Fernando García-Torres et al.

This paper aims to evaluate mid-infrared (MIR) Optical Coherence Tomography (OCT) systems as a tool to penetrate different materials and detect sub-surface irregularities. This is useful for monitoring production processes, allowing Non-Destructive Inspection Techniques of great value to the industry. In this exploratory study, several acquisitions are made on composite and ceramics to know the capabilities of the system. In addition, it is assessed which preprocessing and AI-enhanced vision algorithms can be anomaly-detection methodologies capable of detecting abnormal zones in the analyzed objects. Limitations and criteria for the selection of optimal parameters will be discussed, as well as strengths and weaknesses will be highlighted.

Miguel Camacho-Sánchez, Fernando García-Torres, Jesper John Lisegaard et al.

Resin infusion (RI) and resin transfer moulding (RTM) are critical processes for the manufacturing of high-performance fibre-reinforced polymer composites, particularly for large-scale applications such as wind turbine blades. Controlling the resin flow dynamics in these processes is critical to ensure the uniform impregnation of the fibre reinforcements, thereby preventing residual porosities and dry spots that impact the consequent structural integrity of the final component. This paper presents a reinforcement learning (RL) based strategy, established using process simulations, for synchronising the different resin flow fronts in an infusion scenario involving two resin inlets and a single outlet. Using Proximal Policy Optimisation (PPO), our approach addresses the challenge of managing the fluid dynamics in a partially observable environment. The results demonstrate the effectiveness of the RL approach in achieving an accurate flow convergence, highlighting its potential towards improving process control and product quality in composites manufacturing.

Pablo Meseguer, Rocío del Amor, Valery Naranjo

Pretraining on large-scale, in-domain datasets grants histopathology foundation models (FM) the ability to learn task-agnostic data representations, enhancing transfer learning on downstream tasks. In computational pathology, automated whole slide image analysis requires multiple instance learning (MIL) frameworks due to the gigapixel scale of the slides. The diversity among histopathology FMs has highlighted the need to design real-world challenges for evaluating their effectiveness. To bridge this gap, our work presents a novel benchmark for evaluating histopathology FMs as patch-level feature extractors within a MIL classification framework. For that purpose, we leverage the AI4SkIN dataset, a multi-center cohort encompassing slides with challenging cutaneous spindle cell neoplasm subtypes. We also define the Foundation Model - Silhouette Index (FM-SI), a novel metric to measure model consistency against distribution shifts. Our experimentation shows that extracting less biased features enhances classification performance, especially in similarity-based MIL classifiers.

Miguel López-Pérez, Marco Miani, Valery Naranjo et al.

Generative models are popular for medical imaging tasks such as anomaly detection, feature extraction, data visualization, or image generation. Since they are parameterized by deep learning models, they are often sensitive to distribution shifts and unreliable when applied to out-of-distribution data, creating a risk of, e.g. underrepresentation bias. This behavior can be flagged using uncertainty quantification methods for generative models, but their availability remains limited. We propose SLUG: A new UQ method for VAEs that combines recent advances in Laplace approximations with stochastic trace estimators to scale gracefully with image dimensionality. We show that our UQ score -- unlike the VAE's encoder variances -- correlates strongly with reconstruction error and racial underrepresentation bias for dermatological images. We also show how pixel-wise uncertainty can detect out-of-distribution image content such as ink, rulers, and patches, which is known to induce learning shortcuts in predictive models.

Alvaro Pastor-Naranjo, Pablo Meseguer, Rocío del Amor et al.

Ewing's sarcoma (ES), characterized by a high density of small round blue cells without structural organization, presents a significant health concern, particularly among adolescents aged 10 to 19. Artificial intelligence-based systems for automated analysis of histopathological images are promising to contribute to an accurate diagnosis of ES. In this context, this study explores the feature extraction ability of different pre-training strategies for distinguishing ES from other soft tissue or bone sarcomas with similar morphology in digitized tissue microarrays for the first time, as far as we know. Vision-language supervision (VLS) is compared to fully-supervised ImageNet pre-training within a multiple instance learning paradigm. Our findings indicate a substantial improvement in diagnostic accuracy with the adaption of VLS using an in-domain dataset. Notably, these models not only enhance the accuracy of predicted classes but also drastically reduce the number of trainable parameters and computational costs.

Saul Fuster, Farbod Khoraminia, Julio Silva-Rodríguez et al.

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.

Zahra Tabatabaei, Adrián Colomer, JAvier Oliver Moll et al.

Computer Aid Diagnosis (CAD) has developed digital pathology with Deep Learning (DL)-based tools to assist pathologists in decision-making. Content-Based Histopathological Image Retrieval (CBHIR) is a novel tool to seek highly correlated patches in terms of similarity in histopathological features. In this work, we proposed two CBHIR approaches on breast (Breast-twins) and skin cancer (Skin-twins) data sets for robust and accurate patch-level retrieval, integrating a custom-built Siamese network as a feature extractor. The proposed Siamese network is able to generalize for unseen images by focusing on the similar histopathological features of the input pairs. The proposed CBHIR approaches are evaluated on the Breast (public) and Skin (private) data sets with top K accuracy. Finding the optimum amount of K is challenging, but also, as much as K increases, the dissimilarity between the query and the returned images increases which might mislead the pathologists. To the best of the author's belief, this paper is tackling this issue for the first time on histopathological images by evaluating the top first retrieved images. The Breast-twins model achieves 70% of the F1score at the top first, which exceeds the other state-of-the-art methods at a higher amount of K such as 5 and 400. Skin-twins overpasses the recently proposed Convolutional Auto Encoder (CAE) by 67%, increasing the precision. Besides, the Skin-twins model tackles the challenges of Spitzoid Tumors of Uncertain Malignant Potential (STUMP) to assist pathologists with retrieving top K images and their corresponding labels. So, this approach can offer a more explainable CAD tool to pathologists in terms of transparency, trustworthiness, or reliability among other characteristics.

Rocío del Amor, Julio Silva-Rodríguez, Adrián Colomer et al.

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.

Zahra Tabatabaei, Adrian Colomer, Javier Oliver Moll et al.

Digital pathology has revolutionized cancer diagnosis by leveraging Content-Based Medical Image Retrieval (CBMIR) for analyzing histopathological Whole Slide Images (WSIs). CBMIR enables searching for similar content, enhancing diagnostic reliability and accuracy. In 2020, breast and prostate cancer constituted 11.7% and 14.1% of cases, respectively, as reported by the Global Cancer Observatory (GCO). The proposed Unsupervised CBMIR (UCBMIR) replicates the traditional cancer diagnosis workflow, offering a dependable method to support pathologists in WSI-based diagnostic conclusions. This approach alleviates pathologists' workload, potentially enhancing diagnostic efficiency. To address the challenge of the lack of labeled histopathological images in CBMIR, a customized unsupervised Convolutional Auto Encoder (CAE) was developed, extracting 200 features per image for the search engine component. UCBMIR was evaluated using widely-used numerical techniques in CBMIR, alongside visual evaluation and comparison with a classifier. The validation involved three distinct datasets, with an external evaluation demonstrating its effectiveness. UCBMIR outperformed previous studies, achieving a top 5 recall of 99% and 80% on BreaKHis and SICAPv2, respectively, using the first evaluation technique. Precision rates of 91% and 70% were achieved for BreaKHis and SICAPv2, respectively, using the second evaluation technique. Furthermore, UCBMIR demonstrated the capability to identify various patterns in patches, achieving an 81% accuracy in the top 5 when tested on an external image from Arvaniti.

Zahra Tabatabaei, Yuandou Wang, Adrián Colomer et al.

The paper proposes a Federated Content-Based Medical Image Retrieval (FedCBMIR) platform that utilizes Federated Learning (FL) to address the challenges of acquiring a diverse medical data set for training CBMIR models. CBMIR assists pathologists in diagnosing breast cancer more rapidly by identifying similar medical images and relevant patches in prior cases compared to traditional cancer detection methods. However, CBMIR in histopathology necessitates a pool of Whole Slide Images (WSIs) to train to extract an optimal embedding vector that leverages search engine performance, which may not be available in all centers. The strict regulations surrounding data sharing in medical data sets also hinder research and model development, making it difficult to collect a rich data set. The proposed FedCBMIR distributes the model to collaborative centers for training without sharing the data set, resulting in shorter training times than local training. FedCBMIR was evaluated in two experiments with three scenarios on BreaKHis and Camelyon17 (CAM17). The study shows that the FedCBMIR method increases the F1-Score (F1S) of each client to 98%, 96%, 94%, and 97% in the BreaKHis experiment with a generalized model of four magnifications and does so in 6.30 hours less time than total local training. FedCBMIR also achieves 98% accuracy with CAM17 in 2.49 hours less training time than local training, demonstrating that our FedCBMIR is both fast and accurate for both pathologists and engineers. In addition, our FedCBMIR provides similar images with higher magnification for non-developed countries where participate in the worldwide FedCBMIR with developed countries to facilitate mitosis measuring in breast cancer diagnosis. We evaluate this scenario by scattering BreaKHis into four centers with different magnifications.

Gabriel García, Adrián Colomer, Rafael Verdú-Monedero et al.

In this paper, we present a self-training-based framework for glaucoma grading using OCT B-scans under the presence of domain shift. Particularly, the proposed two-step learning methodology resorts to pseudo-labels generated during the first step to augment the training dataset on the target domain, which is then used to train the final target model. This allows transferring knowledge-domain from the unlabeled data. Additionally, we propose a novel glaucoma-specific backbone which introduces residual and attention modules via skip-connections to refine the embedding features of the latent space. By doing this, our model is capable of improving state-of-the-art from a quantitative and interpretability perspective. The reported results demonstrate that the proposed learning strategy can boost the performance of the model on the target dataset without incurring in additional annotation steps, by using only labels from the source examples. Our model consistently outperforms the baseline by 1-3% across different metrics and bridges the gap with respect to training the model on the labeled target data.

Julio Silva-Rodríguez, Valery Naranjo, Jose Dolz

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.

Gabriel García, Anna Esteve, Adrián Colomer et al.

Recently, bladder cancer has been significantly increased in terms of incidence and mortality. Currently, two subtypes are known based on tumour growth: non-muscle invasive (NMIBC) and muscle-invasive bladder cancer (MIBC). In this work, we focus on the MIBC subtype because it is of the worst prognosis and can spread to adjacent organs. We present a self-learning framework to grade bladder cancer from histological images stained via immunohistochemical techniques. Specifically, we propose a novel Deep Convolutional Embedded Attention Clustering (DCEAC) which allows classifying histological patches into different severity levels of the disease, according to the patterns established in the literature. The proposed DCEAC model follows a two-step fully unsupervised learning methodology to discern between non-tumour, mild and infiltrative patterns from high-resolution samples of 512x512 pixels. Our system outperforms previous clustering-based methods by including a convolutional attention module, which allows refining the features of the latent space before the classification stage. The proposed network exceeds state-of-the-art approaches by 2-3% across different metrics, achieving a final average accuracy of 0.9034 in a multi-class scenario. Furthermore, the reported class activation maps evidence that our model is able to learn by itself the same patterns that clinicians consider relevant, without incurring prior annotation steps. This fact supposes a breakthrough in muscle-invasive bladder cancer grading which bridges the gap with respect to train the model on labelled data.

Gabriel García, Rocío del Amor, Adrián Colomer et al.

Glaucoma is one of the leading causes of blindness worldwide and Optical Coherence Tomography (OCT) is the quintessential imaging technique for its detection. Unlike most of the state-of-the-art studies focused on glaucoma detection, in this paper, we propose, for the first time, a novel framework for glaucoma grading using raw circumpapillary B-scans. In particular, we set out a new OCT-based hybrid network which combines hand-driven and deep learning algorithms. An OCT-specific descriptor is proposed to extract hand-crafted features related to the retinal nerve fibre layer (RNFL). In parallel, an innovative CNN is developed using skip-connections to include tailored residual and attention modules to refine the automatic features of the latent space. The proposed architecture is used as a backbone to conduct a novel few-shot learning based on static and dynamic prototypical networks. The k-shot paradigm is redefined giving rise to a supervised end-to-end system which provides substantial improvements discriminating between healthy, early and advanced glaucoma samples. The training and evaluation processes of the dynamic prototypical network are addressed from two fused databases acquired via Heidelberg Spectralis system. Validation and testing results reach a categorical accuracy of 0.9459 and 0.8788 for glaucoma grading, respectively. Besides, the high performance reported by the proposed model for glaucoma detection deserves a special mention. The findings from the class activation maps are directly in line with the clinicians' opinion since the heatmaps pointed out the RNFL as the most relevant structure for glaucoma diagnosis.

Julio Silva-Rodríguez, Elena Payá-Bosch, Gabriel García et al.

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.

Julio Silva-Rodríguez, Adrián Colomer, María A. Sales et al.

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.

Julio Silva-Rodríguez, Adrián Colomer, Valery Naranjo

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.

Julio Silva-Rodríguez, Adrián Colomer, Jose Dolz et al.

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.

Rocío del Amor, Laëtitia Launet, Adrián Colomer et al.

Melanoma is an aggressive neoplasm responsible for the majority of deaths from skin cancer. Specifically, spitzoid melanocytic tumors are one of the most challenging melanocytic lesions due to their ambiguous morphological features. The gold standard for its diagnosis and prognosis is the analysis of skin biopsies. In this process, dermatopathologists visualize skin histology slides under a 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 daily clinical practice. Nevertheless, no automatic CAD systems have yet been proposed for the analysis of spitzoid lesions. Regarding common melanoma, no proposed system allows both the selection of the tumoral region and the prediction of the diagnosis as benign or malignant. Motivated by this, we propose a novel end-to-end weakly-supervised deep learning model, based on inductive transfer learning with an improved convolutional neural network (CNN) to refine the embedding features of the latent space. The framework is composed of a source model in charge of finding the tumor patch-level patterns, and a target model focuses on the specific diagnosis of a biopsy. The latter retrains the backbone of the source model through a multiple instance learning workflow to obtain the biopsy-level scoring. To evaluate the performance of the proposed methods, we perform extensive experiments on a private skin database with spitzoid lesions. Test results reach an accuracy of 0.9231 and 0.80 for the source and the target models, respectively. Besides, the heat map findings are directly in line with the clinicians' medical decision and even highlight, in some cases, patterns of interest that were overlooked by the pathologist due to the huge workload.

Gabriel García, Rocío del Amor, Adrián Colomer et al.

Nowadays, glaucoma is the leading cause of blindness worldwide. We propose in this paper two different deep-learning-based approaches to address glaucoma detection just from raw circumpapillary OCT images. The first one is based on the development of convolutional neural networks (CNNs) trained from scratch. The second one lies in fine-tuning some of the most common state-of-the-art CNNs architectures. The experiments were performed on a private database composed of 93 glaucomatous and 156 normal B-scans around the optic nerve head of the retina, which were diagnosed by expert ophthalmologists. The validation results evidence that fine-tuned CNNs outperform the networks trained from scratch when small databases are addressed. Additionally, the VGG family of networks reports the most promising results, with an area under the ROC curve of 0.96 and an accuracy of 0.92, during the prediction of the independent test set.

Amartya Kalapahar, Julio Silva-Rodríguez, Adrián Colomer et al.

Worldwide, prostate cancer is one of the main cancers affecting men. The final diagnosis of prostate cancer is based on the visual detection of Gleason patterns in prostate biopsy by pathologists. Computer-aided-diagnosis systems allow to delineate and classify the cancerous patterns in the tissue via computer-vision algorithms in order to support the physicians' task. The methodological core of this work is a U-Net convolutional neural network for image segmentation modified with residual blocks able to segment cancerous tissue according to the full Gleason system. This model outperforms other well-known architectures, and reaches a pixel-level Cohen's quadratic Kappa of 0.52, at the level of previous image-level works in the literature, but providing also a detailed localisation of the patterns.

José Ignacio Orlando, Huazhu Fu, João Barbossa Breda et al.

Glaucoma is one of the leading causes of irreversible but preventable blindness in working age populations. Color fundus photography (CFP) is the most cost-effective imaging modality to screen for retinal disorders. However, its application to glaucoma has been limited to the computation of a few related biomarkers such as the vertical cup-to-disc ratio. Deep learning approaches, although widely applied for medical image analysis, have not been extensively used for glaucoma assessment due to the limited size of the available data sets. Furthermore, the lack of a standardize benchmark strategy makes difficult to compare existing methods in a uniform way. In order to overcome these issues we set up the Retinal Fundus Glaucoma Challenge, REFUGE (\url{https://refuge.grand-challenge.org}), held in conjunction with MICCAI 2018. The challenge consisted of two primary tasks, namely optic disc/cup segmentation and glaucoma classification. As part of REFUGE, we have publicly released a data set of 1200 fundus images with ground truth segmentations and clinical glaucoma labels, currently the largest existing one. We have also built an evaluation framework to ease and ensure fairness in the comparison of different models, encouraging the development of novel techniques in the field. 12 teams qualified and participated in the online challenge. This paper summarizes their methods and analyzes their corresponding results. In particular, we observed that two of the top-ranked teams outperformed two human experts in the glaucoma classification task. Furthermore, the segmentation results were in general consistent with the ground truth annotations, with complementary outcomes that can be further exploited by ensembling the results.