Feature Robustness and Sex Differences in Medical Imaging: A Case Study in MRI-Based Alzheimer’s Disease Detection

Alzheimer’s Disease Neuroimaging Initiative

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

98 Downloads (Pure)

Abstract

Convolutional neural networks have enabled significant improvements in medical image-based diagnosis. It is, however, increasingly clear that these models are susceptible to performance degradation when facing spurious correlations and dataset shift, leading, e.g., to underperformance on underrepresented patient groups. In this paper, we compare two classification schemes on the ADNI MRI dataset: a simple logistic regression model using manually selected volumetric features, and a convolutional neural network trained on 3D MRI data. We assess the robustness of the trained models in the face of varying dataset splits, training set sex composition, and stage of disease. In contrast to earlier work in other imaging modalities, we do not observe a clear pattern of improved model performance for the majority group in the training dataset. Instead, while logistic regression is fully robust to dataset composition, we find that CNN performance is generally improved for both male and female subjects when including more female subjects in the training dataset. We hypothesize that this might be due to inherent differences in the pathology of the two sexes. Moreover, in our analysis, the logistic regression model outperforms the 3D CNN, emphasizing the utility of manual feature specification based on prior knowledge, and the need for more robust automatic feature selection.

Original languageEnglish
Title of host publicationMedical Image Computing and Computer Assisted Intervention
EditorsLinwei Wang, Qi Dou, P. Thomas Fletcher, Stefanie Speidel, Shuo Li
PublisherSpringer
Publication date2022
Pages88-98
ISBN (Print)9783031164309
DOIs
Publication statusPublished - 2022
Event25th International Conference on Medical Image Computing and Computer-Assisted Intervention - Resorts World Convention Centre Singapore, Singapore, Singapore
Duration: 18 Sept 202222 Sept 2022
Conference number: 25
https://conferences.miccai.org/2022/en/

Conference

Conference25th International Conference on Medical Image Computing and Computer-Assisted Intervention
Number25
LocationResorts World Convention Centre Singapore
Country/TerritorySingapore
CitySingapore
Period18/09/202222/09/2022
Internet address
SeriesLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13431 LNCS
ISSN0302-9743

Keywords

  • Alzheimer’s disease
  • Deep learning
  • MRI
  • Robustness

Fingerprint

Dive into the research topics of 'Feature Robustness and Sex Differences in Medical Imaging: A Case Study in MRI-Based Alzheimer’s Disease Detection'. Together they form a unique fingerprint.

Cite this