TriadNet: Sampling-free predictive intervals for lesional volume in 3D brain MR images - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2023

TriadNet: Sampling-free predictive intervals for lesional volume in 3D brain MR images

Abstract

The volume of a brain lesion (e.g. infarct or tumor) is a powerful indicator of patient prognosis and can be used to guide the therapeutic strategy. Lesional volume estimation is usually performed by segmentation with deep convolutional neural networks (CNN), currently the state-of-the-art approach. However, to date, few work has been done to equip volume segmentation tools with adequate quantitative predictive intervals, which can hinder their usefulness and acceptation in clinical practice. In this work, we propose TriadNet, a segmentation approach relying on a multi-head CNN architecture, which provides both the lesion volumes and the associated predictive intervals simultaneously, in less than a second. We demonstrate its superiority over other solutions on BraTS 2021, a large-scale MRI glioblastoma image database.
Fichier principal
Vignette du fichier
triadnet.pdf (4.47 Mo) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-04436218 , version 1 (05-02-2024)

Licence

Identifiers

Cite

Benjamin Lambert, Florence Forbes, Senan Doyle, Michel Dojat. TriadNet: Sampling-free predictive intervals for lesional volume in 3D brain MR images. UNSURE 2023 - 5th International Workshop on Uncertainty for Safe Utilization of Machine Learning in Medical Imaging, Oct 2023, Vancouver (BC), Canada. ⟨10.48550/arXiv.2307.15638⟩. ⟨hal-04436218⟩
58 View
25 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More