A systematic comparison of deep learning strategies for weakly supervised Gleason grading

Otálora, Sebastian; Atzori, Manfredo; Khan, Amjad; Toro, Óscar Alfonso Jiménez del; Andrearczyk, Vincent; Müller, Henning

doi:10.1117/12.2548571

Cited by 14 publications

(9 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The nuclear contours are available in the form of manual annotations only for the PanNuke data. Automated contours of the nuclei in the Camelyon images are obtained by the multi-instance deep segmentation model in Otálora et al (2020). This model is a Mask R-CNN model (He et al, 2017), fine-tuned from ImageNet weights on the Kumar dataset for the nuclei segmentation task (Kumar et al, 2017).…”

Section: Configuration Of the Extra Targetsmentioning

confidence: 99%

Learning Interpretable Microscopic Features of Tumor by Multi-task Adversarial CNNs Improves Generalization

Graziani

Otálora

Marchand-Maillet

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Adopting Convolutional Neural Networks (CNNs) in the daily routine of pathological diagnosis requires not only near-perfect precision, but also sufficient generalization to data shifts and transparency. Existing CNN models act as black boxes, not ensuring to the physicians that important diagnostic features are used by the model. Building on top of successfully existing techniques such as multi-task learning, domain adversarial training and concept-based interpretability, we address the challenge of introducing diagnostic factors in the training objectives. Our architecture, by learning end-to-end an uncertainty-based weighting combination of multi-task and adversarial losses, is encouraged to focus on pathology features such as density and pleomorphism of nuclei, e.g. variations in size and appearance, while discarding misleading features such as staining differences. Our results on breast lymph node tissue show significantly improved generalization in the detection of tumorous tissue, with best average AUC 0.89 (0.01) against the baseline AUC 0.86 (0.005). By applying the interpretability technique of linearly probing intermediate representations, we also demonstrate that interpretable pathology features such as nuclei density are learned by the proposed CNN architecture, confirming the increased transparency of this model. This result is a starting point towards building interpretable multi-task architectures that are robust to data heterogeneity. Our code is available at https://bit.ly/356yQ2u.

show abstract

Section: Configuration Of the Extra Targetsmentioning

confidence: 99%

Learning Interpretable Microscopic Features of Tumor by Multi-task Adversarial CNNs Improves Generalization

Graziani

Otálora

Marchand-Maillet

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…The proposed method is comparable against works that use strong supervision via pixel-level annotations, i.e., Arvaniti et al [25] (κ = 0.75 for Gleason scoring) and Bulten at al. [31] (κ = 0.85 for Grade Group scoring), as well as works that use only global biopsy-level labels, i.e., Strom et al [30] (κ = 0.91 for Grade Group scoring) and Otlora et al [38] (κ = 0.44 for Grade Group scoring). In accordance with the observations in our work, methods based on the Gleason grades proportion in the tissue suffer a performance drop on external datasets (κ = 0.72 in Bulten et al Representative examples of the obtained results using the CAD system on the external SICAP dataset are presented in Fig.…”

Section: ) Grading Of Local Patternsmentioning

confidence: 99%

Self-Learning for Weakly Supervised Gleason Grading of Local Patterns

Silva-Rodríguez

Colomer

Dolz

et al. 2021

IEEE J. Biomed. Health Inform.

View full text Add to dashboard Cite

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, computeraided-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. Furthermore, despite the promising results achieved on global scoring the location of cancerous patterns in the tissue is only qualitatively addressed. These heatmaps of tumor regions, however, are crucial to the reliability of CAD systems as they provide explainability to the system's output and give confidence to pathologists that the model is focusing on medical relevant features. 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 patchlevel 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 (κ) score of nearly 18% compared to full-supervision for the patch-level Gleason grading task. This suggests that the absence of the annotator's bias in our approach and the capability of using large weakly labeled datasets during training leads to higher performing and more robust models. Furthermore, raw features obtained from the patch-level classifier showed to generalize better than previous approaches in the literature to the subjective global biopsy-level scoring.

show abstract

“…Deep learning models (e.g., CNNs) form a specific approach that is directly applied on images to extract, select features, and predict the class (classification) or a value (regression) in an automated fashion. Examples in the PCa literature have observed that this deep learning approach detects malignant lesions [ 125 ], predicts the GS [ 126 ], and segments the ROI [ 127 , 128 ].…”

Section: Radiomics Pipeline For Predicting Tumor Gradementioning

confidence: 99%

Magnetic Resonance Imaging Based Radiomic Models of Prostate Cancer: A Narrative Review

Chaddad

Kucharczyk

Cheddad

et al. 2021

Cancers

View full text Add to dashboard Cite

The management of prostate cancer (PCa) is dependent on biomarkers of biological aggression. This includes an invasive biopsy to facilitate a histopathological assessment of the tumor’s grade. This review explores the technical processes of applying magnetic resonance imaging based radiomic models to the evaluation of PCa. By exploring how a deep radiomics approach further optimizes the prediction of a PCa’s grade group, it will be clear how this integration of artificial intelligence mitigates existing major technological challenges faced by a traditional radiomic model: image acquisition, small data sets, image processing, labeling/segmentation, informative features, predicting molecular features and incorporating predictive models. Other potential impacts of artificial intelligence on the personalized treatment of PCa will also be discussed. The role of deep radiomics analysis-a deep texture analysis, which extracts features from convolutional neural networks layers, will be highlighted. Existing clinical work and upcoming clinical trials will be reviewed, directing investigators to pertinent future directions in the field. For future progress to result in clinical translation, the field will likely require multi-institutional collaboration in producing prospectively populated and expertly labeled imaging libraries.

show abstract

A systematic comparison of deep learning strategies for weakly supervised Gleason grading

Cited by 14 publications

References 0 publications

Learning Interpretable Microscopic Features of Tumor by Multi-task Adversarial CNNs Improves Generalization

Learning Interpretable Microscopic Features of Tumor by Multi-task Adversarial CNNs Improves Generalization

Self-Learning for Weakly Supervised Gleason Grading of Local Patterns

Magnetic Resonance Imaging Based Radiomic Models of Prostate Cancer: A Narrative Review

Contact Info

Product

Resources

About