Non-parametric Neighborhood Test-Time Generalization: Application to Medical Image Classification

Sameer Ambekar*, Julia A. Schnabel, Daniel M. Lang

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference paperpeer-review

Abstract

Reliable and stable performance is crucial for the application of computer-aided medical image systems in clinical settings. However, approaches based on deep learning often fail to generalize well under distribution shifts. In medical imaging, such distribution shifts can, for example, be introduced by changes in scanner types or imaging protocols. To counter this, test-time generalization aims to optimize a model that has been trained on single or multiple source domains to an unseen target domain. Common test-time adaptation methods fine-tune model weights utilizing losses with gradient-based optimization, a time-consuming and computationally demanding procedure. In contrast, our approach adopts a non-parametric method that is entirely feedforward and utilizes information from target samples to extract neighborhood information with dynamic voting. By doing so, we avoid fine-tuning or optimization procedures, enabling our method to be more efficient and achieve stable adaptation. We demonstrate the effectiveness of our approach by benchmarking it against different state-of-the-art methods with three backbones on two publicly available medical imaging datasets, consisting of fetal ultrasound and retinal images, and achieve classification accuracy improvements by up to 3.4% and 1.1%, respectively. Moreover, we also demonstrate the utility of our method in practical scenarios, proving efficiency in terms of computational runtime and handling of uncertainty. Our code is publicly available at: https://github.com/compai-lab/2024-miccai-emerge-ambekar.

Original languageEnglish
Title of host publicationMedical Information Computing - First MICCAI Meets Africa Workshop, MImA 2024, and First MICCAI Student Board Workshop on Empowering Medical Information Computing and Research through Early-Career Expertise, EMERGE 2024, Held in Conjunction with MICCAI 2024, Revised Selected Papers
EditorsUdunna Anazodo, Naren Akash, Moritz Fuchs, Celia Cintas, Alessandro Crimi, Tinahse Mutsvangwa, Farouk Dako, Willam Ogallo
PublisherSpringer Science and Business Media Deutschland GmbH
Pages224-234
Number of pages11
ISBN (Print)9783031791024
DOIs
Publication statusPublished - 2025
Event1st MICCAI Meets Africa Workshop, MImA 2024 and 1st MICCAI Student Board Workshop on Empowering Medical Information Computing and Research through Early-Career Expertise, EMERGE 2024, Held in Conjunction with MICCAI 2024 - Marrakesh, Morocco
Duration: 6 Oct 20246 Oct 2024

Publication series

NameCommunications in Computer and Information Science
Volume2240
ISSN (Print)1865-0929
ISSN (Electronic)1865-0937

Conference

Conference1st MICCAI Meets Africa Workshop, MImA 2024 and 1st MICCAI Student Board Workshop on Empowering Medical Information Computing and Research through Early-Career Expertise, EMERGE 2024, Held in Conjunction with MICCAI 2024
Country/TerritoryMorocco
CityMarrakesh
Period6/10/20246/10/2024

Keywords

  • domain adaptation
  • generalization
  • parameter-free optimization
  • unsupervised learning

Fingerprint

Dive into the research topics of 'Non-parametric Neighborhood Test-Time Generalization: Application to Medical Image Classification'. Together they form a unique fingerprint.

Cite this