TY - CHAP
T1 - Non-parametric Neighborhood Test-Time Generalization
T2 - 1st 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
AU - Ambekar, Sameer
AU - Schnabel, Julia A.
AU - Lang, Daniel M.
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
PY - 2025
Y1 - 2025
N2 - 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.
AB - 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.
KW - domain adaptation
KW - generalization
KW - parameter-free optimization
KW - unsupervised learning
UR - http://www.scopus.com/inward/record.url?scp=85219168764&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-79103-1_23
DO - 10.1007/978-3-031-79103-1_23
M3 - Conference paper
AN - SCOPUS:85219168764
SN - 9783031791024
T3 - Communications in Computer and Information Science
SP - 224
EP - 234
BT - Medical 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
A2 - Anazodo, Udunna
A2 - Akash, Naren
A2 - Fuchs, Moritz
A2 - Cintas, Celia
A2 - Crimi, Alessandro
A2 - Mutsvangwa, Tinahse
A2 - Dako, Farouk
A2 - Ogallo, Willam
PB - Springer Science and Business Media Deutschland GmbH
Y2 - 6 October 2024 through 6 October 2024
ER -