Exploiting Intermediate Reconstructions in Optical Coherence Tomography for Test-Time Adaption of Medical Image Segmentation

Thomas Pinetz, Veit Hucke, Hrvoje Bogunović

Keywords: Diffusion, Test-Time Adaptation, Uncertainty Estimation, Optical Coherence Tomography

TL;DR: We exploit reconstruction trajectories by adapting a frozen segmentation network via a modulator network to improve downstream performance.

Abstract: Primary health care frequently relies on lower–quality imaging devices, which are commonly used for screening purposes. To ensure accurate diagnosis, these systems depend on advanced reconstruction algorithms designed to approximate the performance of high-quality counterparts. Such algorithms typically employ iterative reconstruction methods that incorporate domain-specific prior knowledge. However, downstream task performance is generally assessed using only the final reconstructed image, thereby disregarding the informative intermediate representations generated throughout the reconstruction process. In this work, we propose to exploit these intermediate representations by adapting the normalization-layer parameters of a frozen downstream network via a modulator network that conditions on the current reconstruction timescale. The modulator network is learned during test-time using an averaged entropy loss across all individual timesteps. Variation among the timestep-wise segmentations additionally provides uncertainty estimates at no extra cost. This approach enhances segmentation performance and enables semantically meaningful uncertainty estimation, all without modifying either the reconstruction process or the downstream model. Code is available.

Primary Subject Area: Transfer Learning and Domain Adaptation

Secondary Subject Area: Application: Ophthalmology

Registration Requirement: Yes

Reproducibility: https://github.com/tpinetz/domain_adaption_by_iterative_reconstruction

Visa & Travel: No

Read CFP & Author Instructions: Yes

Originality Policy: Yes

Single-blind & Not Under Review Elsewhere: Yes

LLM Policy: Yes

Submission Number: 86