Bone mineral density response prediction following osteoporosis treatment using machine learning to aid personalized therapy
Problem
Doctors lack a reliable way to select the best osteoporosis treatment regimen for individual patients to improve bone mineral density and prevent fractures.
Why it matters
Osteoporosis is a global health problem in ageing populations, and wrong treatment choices lead to poor outcomes and fractures.
What the AI did
Seven machine learning models were trained on 13,562 treatment instances from 10 years of electronic medical records to predict BMD response to 11 therapeutic regimens; Random Forests was selected as best.
Results
Best model: AUC 0.70, accuracy 0.69, precision 0.70, recall 0.89.
Impact
Provides a data-driven aid for personalizing osteoporosis therapy, though predictive accuracy is modest.
Record
| Venue | Scientific Reports |
|---|---|
| Published | 2021 |
| Work type | article |
| Authors | Thiraphat Tanphiriyakun, Sattaya Rojanasthien, Piyapong Khumrin |
| Citations | 39 (OpenAlex, at import time) |
| References | 55 |
| Model or method | Random Forests |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1038/s41598-021-93152-5 |
| Open access | gold |
| Keywords | osteoporosis, bone mineral density, machine learning, Random Forest, personalized medicine |
Importance rating
significant — thousands to millions affected
Abstract states osteoporosis is a global health problem for ageing populations; wrong treatment choices cause poor outcomes and fractures, affecting millions, though the model itself is a modest-accuracy decision aid.