PreprintCC BY 4.0arXiv
Quantifying Methodological Inflation in Clinical Machine Learning: A Cross-Modality Audit of Subject-Level Leakage and Nested Feature Selection for Parkinson's Disease
Abstract
Reported accuracy in Parkinson's disease (PD) machine learning can reflect the evaluation protocol rather than real signal. We ask how much evaluation choices inflate performance by comparing naive and rigorous protocols on voice, telemonitoring and blood-transcriptomic datasets. Subject-aware splitting reduced voice AUROC from 0.892 to 0.780 and telemonitoring R^2 from 0.105 to -0.037; Random Forest recovered neither gap. Supervised probe selection before splitting inflated transcriptomic AUROC from 0.711 to 0.998. A nested 50-probe panel matched a 1,000-probe panel (0.811 versus 0.844), and recruitment-wave batch structure was detected. Future work should validate a stable panel on an independent cohort.
Related
- Quantifying and adjusting for selection biases in the Norwegian Mother, Father and Child Cohort Study using population-wide individual-level registry information
- Discussion on "Principled type I error rate inflation in two-arm clinical trial designs with external control information borrowing" by Silvia Calderazzo, Manuel Wiesenfarth, Vivienn Weru, and Annette Kopp-Schneider
- Psychometric properties of the Persian version of the self-regulated learning scale in clinical nursing practice for nursing students: A methodological study
- Healthy Participant Versus Patient Selection for Clinical Pharmacology Studies in Small-Molecule Oncology Drug Development.
- Valid Data‐Driven Subgroup Selection With Uniform Guarantees in Clinical Trials
BibTeX
@misc{datta2026quantifying,
title = {{Quantifying Methodological Inflation in Clinical Machine Learning: A Cross-Modality Audit of Subject-Level Leakage and Nested Feature Selection for Parkinson's Disease}},
author = {Namrata Datta and Victor Miranda Soberanis},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.14533},
eprint = {2611.14533},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.14533},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Quantifying Methodological Inflation in Clinical Machine Learning: A Cross-Modality Audit of Subject-Level Leakage and Nested Feature Selection for Parkinson's Disease
AU - Namrata Datta
AU - Victor Miranda Soberanis
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.14533
UR - https://arxiv.org/abs/2611.14533
AB - Reported accuracy in Parkinson's disease (PD) machine learning can reflect the evaluation protocol rather than real signal. We ask how much evaluation choices inflate performance by comparing naive and rigorous protocols on voice, telemonitoring and blood-transcriptomic datasets. Subject-aware splitting reduced voice AUROC from 0.892 to 0.780 and telemonitoring R^2 from 0.105 to -0.037; Random Forest recovered neither gap. Supervised probe selection before splitting inflated transcriptomic AUROC from 0.711 to 0.998. A nested 50-probe panel matched a 1,000-probe panel (0.811 versus 0.844), and recruitment-wave batch structure was detected. Future work should validate a stable panel on an independent cohort.
ER -