Comp2Comp: Open-Source Body Composition Assessment on Computed Tomography

Louis Blankemeier, Malte Jensen, Eduardo Pontes Reis, Juan Manuel Zambrano Chaves, Adrit Rao, Sally Yao, Pauline Margaret Berens, Andrew Wentland, Bhanushree Bahl, Kushboo Arora, Oliver Oppers Aalami, Bhavik Patel, Leon Lenchik, Marc H. Willis, Robert D. Boutin, Arjun D Desai, Akshay S Chaudhari

Keywords: computed tomography, segmentation, body composition, abdominal CT

TL;DR: We present Comp2Comp, an open-source Python package for rapid and automated body composition analysis of CT scans.

Abstract: Computed tomography (CT) can provide quantitative body composition metrics of tissue volume, morphology, and quality which are valuable for disease prediction and prognostication. However, manually extracting these measures is a cumbersome and time-consuming task. Proprietary software to automate this process exist, but these software are closed-source, impeding large-scale access to and usage of these tools. To address this, we have built Comp2Comp, an open-source Python package for rapid and automated body composition analysis of CT scans. The primary advantages of Comp2Comp are its open-source nature, the inclusion of multiple tissue analysis capabilities within a single package, and its extensible design. We discuss the architecture of Comp2Comp and report initial validation results. Comp2Comp can be found at https://github.com/StanfordMIMI/Comp2Comp.

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:2302.06568/code)