CohortFinder: an open-source tool for data-driven partitioning of biomedical image cohorts to yield robust machine learning models
This addresses batch effects in biomedical imaging for researchers, but it is incremental as it applies an existing partitioning approach to this domain.
The authors tackled the problem of batch effects in biomedical image data, which harm machine learning model generalizability, by releasing CohortFinder, an open-source tool for data-driven cohort partitioning that improves model performance in downstream medical image processing tasks.
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.