Interpretable Auto Window Setting for Deep-Learning-Based CT Analysis

TL;DR

This paper introduces an interpretable, plug-and-play auto window setting module for deep learning-based CT analysis, enhancing interpretability and trustworthiness for clinical use while significantly improving segmentation accuracy.

Contribution

It proposes a novel, domain-invariant, interpretable module based on the Tanh function that integrates with existing deep learning models for CT window setting.

Findings

Achieved 10%-200% Dice score improvements on challenging segmentation tasks.

Demonstrated higher interpretability and clinical trust compared to existing methods.

Validated effectiveness across multiple open-source datasets.

Abstract

Whether during the early days of popularization or in the present, the window setting in Computed Tomography (CT) has always been an indispensable part of the CT analysis process. Although research has investigated the capabilities of CT multi-window fusion in enhancing neural networks, there remains a paucity of domain-invariant, intuitively interpretable methodologies for Auto Window Setting. In this work, we propose an plug-and-play module originate from Tanh activation function, which is compatible with mainstream deep learning architectures. Starting from the physical principles of CT, we adhere to the principle of interpretability to ensure the module's reliability for medical implementations. The domain-invariant design facilitates observation of the preference decisions rendered by the adaptive mechanism from a clinically intuitive perspective. This enables the proposed method to be understood not only by experts in neural networks but also garners higher trust from clinicians. We confirm the effectiveness of the proposed method in multiple open-source datasets, yielding 10%~200% Dice improvements on hard segment targets.