Quality or Quantity: Toward a Unified Approach for Multi-organ Segmentation in Body CT

TL;DR

This paper investigates the tradeoffs between data quality and quantity in multi-organ segmentation of body CT scans, proposing a unified approach that leverages pseudo-labels to improve segmentation accuracy.

Contribution

It compares segmentation architectures, introduces a pseudo-labeling method, and demonstrates that high-quality data significantly enhances model performance.

Findings

3D-Unet outperformed DenseVNet in initial tests.

Including pseudo-labeled data improved segmentation for organs with true labels.

High-quality labeled data is more crucial than larger quantity for model accuracy.

Abstract

Organ segmentation of medical images is a key step in virtual imaging trials. However, organ segmentation datasets are limited in terms of quality (because labels cover only a few organs) and quantity (since case numbers are limited). In this study, we explored the tradeoffs between quality and quantity. Our goal is to create a unified approach for multi-organ segmentation of body CT, which will facilitate the creation of large numbers of accurate virtual phantoms. Initially, we compared two segmentation architectures, 3D-Unet and DenseVNet, which were trained using XCAT data that is fully labeled with 22 organs, and chose the 3D-Unet as the better performing model. We used the XCAT-trained model to generate pseudo-labels for the CT-ORG dataset that has only 7 organs segmented. We performed two experiments: First, we trained 3D-UNet model on the XCAT dataset, representing quality data, and tested it on both XCAT and CT-ORG datasets. Second, we trained 3D-UNet after including the CT-ORG dataset into the training set to have more quantity. Performance improved for segmentation in the organs where we have true labels in both datasets and degraded when relying on pseudo-labels. When organs were labeled in both datasets, Exp-2 improved Average DSC in XCAT and CT-ORG by 1. This demonstrates that quality data is the key to improving the model's performance.