Anatomical Positional Embeddings

TL;DR

This paper introduces a self-supervised model that generates 3D anatomical positional embeddings for medical images, enabling efficient voxel-wise mapping and improving tasks like organ localization and image cropping.

Contribution

It presents a novel method for producing comprehensive voxel-wise anatomical embeddings for entire volumetric images, outperforming existing models in accuracy and efficiency.

Findings

Superior performance in anatomical landmark retrieval

Effective weakly-supervised organ localization

High recall in anatomical region cropping

Abstract

We propose a self-supervised model producing 3D anatomical positional embeddings (APE) of individual medical image voxels. APE encodes voxels' anatomical closeness, i.e., voxels of the same organ or nearby organs always have closer positional embeddings than the voxels of more distant body parts. In contrast to the existing models of anatomical positional embeddings, our method is able to efficiently produce a map of voxel-wise embeddings for a whole volumetric input image, which makes it an optimal choice for different downstream applications. We train our APE model on 8400 publicly available CT images of abdomen and chest regions. We demonstrate its superior performance compared with the existing models on anatomical landmark retrieval and weakly-supervised few-shot localization of 13 abdominal organs. As a practical application, we show how to cheaply train APE to crop raw CT images to different anatomical regions of interest with 0.99 recall, while reducing the image volume by 10-100 times. The code and the pre-trained APE model are available at https://github.com/mishgon/ape .