Can GPT Be Used as an Alternative Prediction Model to Traditional Machine Learning and Neural Networks on Low-Volume Clinical Data?
- 1Institute for Population and Precision Health, University of Chicago, Department of Family Medicine, Illinois, USA
Abstract
PurposeEarly and reliable disease prediction from structured clinical data remains challenging when datasets are small, highly imbalanced, and contain limited positive disease cases. Conventional machine learning (ML) and deep learning approaches often struggle to capture clinically meaningful relationships under such low-data representation conditions due to weak statistical associations between features and prediction targets. This study proposes a clinically grounded Distil GPT2-based table-to-text framework for disease prediction using structured healthcare datasets, motivated by the contextual reasoning capability of GPT models to better capture clinically meaningful relationships when statistical learning alone becomes insufficient due to limited data availability.
MethodsStructured clinical records were transformed into physician-style textual descriptions and enriched through GPT4-generated medical paraphrasing to improve minority-class representation while preserving clinical meaning. Both the original and generated clinical texts were used to fine-tune a Distil GPT2 model across four public healthcare datasets, including heart disease, heart failure, chronic kidney disease, and thyroid cancer recurrence. Gradient-based explainable AI analysis was additionally incorporated to identify clinically important features influencing prediction outcomes.
ResultsThe proposed framework demonstrated consistently strong predictive performance across four clinical datasets, achieving average precision, specificity, sensitivity, and F1-score of 0.96, 0.97, 0.96, and 0.96, respectively. The model achieved improved sensitivity, stronger generalization, and more stable predictive behavior compared with traditional ML, deep learning, transformer-based, and GAN-augmented approaches. Importantly, the framework consistently emphasized clinically meaningful variables even under severe class imbalance, where conventional ML and neural network models often struggled to identify key clinically relevant relationships.
ConclusionThe proposed Distil GPT2-based tableto-text framework provides a practical and clinically interpretable approach for disease prediction from limited structured healthcare data. By integrating contextual clinical reasoning with explainable prediction mechanisms, the framework suggests strong potential for early risk detection, transparent clinical decision support, and reliable deployment in real-world data-scarce healthcare settings.
Related
- Traditional Machine Learning versus Deep Learning Approaches for Multi-Outcome Perioperative Risk Prediction
- Psychosocial predictors and justification patterns of traditional and alternative medicine in Peru
- Characteristics of junior residents choosing alternative career pathways without entering traditional specialty residency programs: An exploratory descriptive study
- Machine learning-based prediction of clinical pregnancy in infertile women undergoing in vitro fertilization
- Multimodal ensemble machine learning pipeline for oocyte fertilization prediction using first polar body morphology and clinical factors
BibTeX
@misc{akter2026used,
title = {{Can GPT Be Used as an Alternative Prediction Model to Traditional Machine Learning and Neural Networks on Low-Volume Clinical Data?}},
author = {S. Bin Akter and S. Akter and D. Eisenberg and C. Hill and A. Lotvola and J. Fresneda Fernandez and T. Sarkar Pias and M. Rafiqul Islam and H. Islam},
howpublished = {medRxiv},
year = {2026},
month = aug,
doi = {10.64898/2026.08.19.26360765},
url = {https://www.medrxiv.org/content/10.64898/2026.08.19.26360765v2},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Can GPT Be Used as an Alternative Prediction Model to Traditional Machine Learning and Neural Networks on Low-Volume Clinical Data?
AU - S. Bin Akter
AU - S. Akter
AU - D. Eisenberg
AU - C. Hill
AU - A. Lotvola
AU - J. Fresneda Fernandez
AU - T. Sarkar Pias
AU - M. Rafiqul Islam
AU - H. Islam
PY - 2026
DA - 2026/08/19
PB - medRxiv
DO - 10.64898/2026.08.19.26360765
UR - https://www.medrxiv.org/content/10.64898/2026.08.19.26360765v2
AB - PurposeEarly and reliable disease prediction from structured clinical data remains challenging when datasets are small, highly imbalanced, and contain limited positive disease cases. Conventional machine learning (ML) and deep learning approaches often struggle to capture clinically meaningful relationships under such low-data representation conditions due to weak statistical associations between features and prediction targets. This study proposes a clinically grounded Distil GPT2-based table-to-text framework for disease prediction using structured healthcare datasets, motivated by the contextual reasoning capability of GPT models to better capture clinically meaningful relationships when statistical learning alone becomes insufficient due to limited data availability. MethodsStructured clinical records were transformed into physician-style textual descriptions and enriched through GPT4-generated medical paraphrasing to improve minority-class representation while preserving clinical meaning. Both the original and generated clinical texts were used to fine-tune a Distil GPT2 model across four public healthcare datasets, including heart disease, heart failure, chronic kidney disease, and thyroid cancer recurrence. Gradient-based explainable AI analysis was additionally incorporated to identify clinically important features influencing prediction outcomes. ResultsThe proposed framework demonstrated consistently strong predictive performance across four clinical datasets, achieving average precision, specificity, sensitivity, and F1-score of 0.96, 0.97, 0.96, and 0.96, respectively. The model achieved improved sensitivity, stronger generalization, and more stable predictive behavior compared with traditional ML, deep learning, transformer-based, and GAN-augmented approaches. Importantly, the framework consistently emphasized clinically meaningful variables even under severe class imbalance, where conventional ML and neural network models often struggled to identify key clinically relevant relationships. ConclusionThe proposed Distil GPT2-based tableto-text framework provides a practical and clinically interpretable approach for disease prediction from limited structured healthcare data. By integrating contextual clinical reasoning with explainable prediction mechanisms, the framework suggests strong potential for early risk detection, transparent clinical decision support, and reliable deployment in real-world data-scarce healthcare settings.
ER -