Regression-based Deep-Learning predicts molecular biomarkers from pathology slides

Abstract

Deep Learning (DL) can predict biomarkers from cancer histopathology. Several clinically approved applications use this technology. Most approaches, however, predict categorical labels, whereas biomarkers are often continuous measurements. We hypothesize that regression-based DL outperforms classification-based DL. Therefore, we develop and evaluate a self-supervised attention-based weakly supervised regression method that predicts continuous biomarkers directly from 11,671 images of patients across nine cancer types. We test our method for multiple clinically and biologically relevant biomarkers: homologous recombination deficiency score, a clinically used pan-cancer biomarker, as well as markers of key biological processes in the tumor microenvironment. Using regression significantly enhances the accuracy of biomarker prediction, while also improving the predictions’ correspondence to regions of known clinical relevance over classification. In a large cohort of colorectal cancer patients, regression-based prediction scores provide a higher prognostic value than classification-based scores. Our open-source regression approach offers a promising alternative for continuous biomarker analysis in computational pathology.

Predicting the Tumor Microenvironment Composition and Immunotherapy Response in Non-Small Cell Lung Cancer from Digital Histopathology Images

“…However, its dynamic nature, with multiple interacting components, makes it challenging to identify robust predictive biomarkers 32,33 . The integration of deep learning methods with digital pathology presents a promising and potentially cost-effective approach to interrogate the TME and alleviate some of these issues 41 .…”

Surrogate Biomarker Prediction from Whole-Slide Images for Evaluating Overall Survival in Lung Adenocarcinoma

“…AMIL models were trained on the extracted patch features to predict selected targets [ 7 , 33 ]. The AMIL model was selected due to its ability to operate effectively in a weakly-supervised manner (i.e., using WSI-level labels rather than region-level annotations) and relatively efficient training time compared to other deep learning models [ 34 ].…”

“…Computational pathology aims to make use of readily available data to open new avenues for cost-effective biomarker evaluation. In particular, machine learning applied to haematoxylin and eosin (H&E) stained whole-slide images (WSIs) has shown promise in predicting various molecular biomarkers [ 4 , 5 , 6 , 7 ]. However, end-to-end strategies, in which machine learning is used to directly predict clinical outcomes, remain a challenge [ 8 ].…”

Practical Application of Deep Learning in Diagnostic Neuropathology—Reimagining a Histological Asset in the Era of Precision Medicine

“…Beyond efficient morphologic assessment, DL algorithms have been shown to identify patterns in WSIs, which may, themselves, be used to predict molecular alterations both within neuropathology [ 16 , 17 ] and other pathology disciplines [ 18 , 19 ]. This capability may be harnessed to assist with subtype classification and prognosis in resource-limited settings, as the availability and accessibility of costly advanced molecular diagnostic technologies varies greatly within and between countries [ 20 ].…”