Faithful learning with sure data for lung nodule diagnosis

TL;DR

This paper introduces a novel deep learning framework for lung nodule classification that leverages pathologically-confirmed data and integrates unsure data, improving model reliability, interpretability, and clinical applicability.

Contribution

It proposes a collaborative learning approach with a new loss function and interpretability constraints to enhance faithful lung nodule diagnosis using sure and unsure data.

Findings

Improved classification accuracy with pathologically-confirmed data.

Enhanced model interpretability through segmentation and score regression.

Effective retrieval of similar nodules for diagnosis support.

Abstract

Recent evolution in deep learning has proven its value for CT-based lung nodule classification. Most current techniques are intrinsically black-box systems, suffering from two generalizability issues in clinical practice. First, benign-malignant discrimination is often assessed by human observers without pathologic diagnoses at the nodule level. We termed these data as "unsure data". Second, a classifier does not necessarily acquire reliable nodule features for stable learning and robust prediction with patch-level labels during learning. In this study, we construct a sure dataset with pathologically-confirmed labels and propose a collaborative learning framework to facilitate sure nodule classification by integrating unsure data knowledge through nodule segmentation and malignancy score regression. A loss function is designed to learn reliable features by introducing interpretability constraints regulated with nodule segmentation maps. Furthermore, based on model inference results that reflect the understanding from both machine and experts, we explore a new nodule analysis method for similar historical nodule retrieval and interpretable diagnosis. Detailed experimental results demonstrate that our approach is beneficial for achieving improved performance coupled with faithful model reasoning for lung cancer prediction. Extensive cross-evaluation results further illustrate the effect of unsure data for deep-learning-based methods in lung nodule classification.