CohortFinder: an open-source tool for data-driven partitioning of biomedical image cohorts to yield robust machine learning models

TL;DR

CohortFinder is an open-source tool designed to mitigate batch effects in biomedical imaging data by data-driven cohort partitioning, thereby improving machine learning model robustness and generalizability.

Contribution

The paper introduces CohortFinder, a novel open-source tool that effectively reduces batch effects in biomedical image datasets to enhance ML model performance.

Findings

CohortFinder improves ML model accuracy on medical imaging tasks.

The tool effectively mitigates batch effects in biomedical datasets.

Open-source availability facilitates widespread adoption.

Abstract

Batch effects (BEs) refer to systematic technical differences in data collection unrelated to biological variations whose noise is shown to negatively impact machine learning (ML) model generalizability. Here we release CohortFinder, an open-source tool aimed at mitigating BEs via data-driven cohort partitioning. We demonstrate CohortFinder improves ML model performance in downstream medical image processing tasks. CohortFinder is freely available for download at cohortfinder.com.