HyperSpace: Hypernetworks for spacing-adaptive image segmentation

TL;DR

HyperSpace introduces hypernetworks to condition image segmentation models on voxel spacing, enabling flexible resolution processing at inference time without sacrificing performance, thus improving adaptability and efficiency in medical imaging workflows.

Contribution

The paper presents a novel hypernetwork-based method to adapt segmentation models to different voxel spacings, reducing the need for multiple models and simplifying deployment.

Findings

Achieves competitive accuracy across various resolutions.

Provides greater flexibility for end-user resolution choices.

Simplifies model development and deployment processes.

Abstract

Medical images are often acquired in different settings, requiring harmonization to adapt to the operating point of algorithms. Specifically, to standardize the physical spacing of imaging voxels in heterogeneous inference settings, images are typically resampled before being processed by deep learning models. However, down-sampling results in loss of information, whereas upsampling introduces redundant information leading to inefficient resource utilization. To overcome these issues, we propose to condition segmentation models on the voxel spacing using hypernetworks. Our approach allows processing images at their native resolutions or at resolutions adjusted to the hardware and time constraints at inference time. Our experiments across multiple datasets demonstrate that our approach achieves competitive performance compared to resolution-specific models, while offering greater flexibility for the end user. This also simplifies model development, deployment and maintenance. Our code is available at https://github.com/ImFusionGmbH/HyperSpace.