Me-NDT: Neural-backed Decision Tree for visual Explainability of deep Medical models - École polytechnique Access content directly
Conference Papers Year : 2021

Me-NDT: Neural-backed Decision Tree for visual Explainability of deep Medical models

Abstract

Despite the progress of deep learning on medical imaging, there is still not a true understanding of what networks learn and of how decisions are reached. Here, we address this by proposing a Visualized Neural-backed Decision Tree for Medical image analysis, Me-NDT. It is a CNN with a tree-based structure template that allows for both classification and visualization of firing neurons, thus offering interpretability. We also introduce node and path losses that allow Me-NDT to consider the entire path instead of isolated nodes. Our experiments on brain CT and chest radiographs outperform all baselines. Overall, Me-NDT is a lighter, comprehensively explanatory model, of great value for clinical practice.
Fichier principal
Vignette du fichier
MIDL_2021_MeNDT_NeuralBackedDecisionTree.pdf (1.06 Mo) Télécharger le fichier
Origin : Publisher files allowed on an open archive

Dates and versions

hal-03353559 , version 1 (24-09-2021)

Identifiers

  • HAL Id : hal-03353559 , version 1

Cite

Guanghui Fu, Ruiqian Wang, Jianqiang Li, Maria Vakalopoulou, Vicky Kalogeiton. Me-NDT: Neural-backed Decision Tree for visual Explainability of deep Medical models. Medical Imaging with Deep Learning 2021, Jul 2021, Lübeck (virtual), Germany. ⟨hal-03353559⟩
130 View
232 Download

Share

Gmail Facebook X LinkedIn More