A hybrid ML-PBPK digital twin framework for clinically interpretable readmission risk and drug exposure stratification in diabetes.
Ahmed K N, Ibrahim Syed Masood JA. · Frontiers in digital health · 2026 · Article
10.3389/fdgth.2026.1867505
Verified evidence
Model evaluation
38c48edf-f9b1-5892-8908-5a6b7340f43dPrediction source Version of record · PDF page(s) 18
Study b95a5665-db87-5237-a379-a0e438da24eb · Prediction record (JSON)
Prediction fields
- Evaluation scope
- internal_holdout · PDF page(s) 18, 26, 33
- Measured variable
- readmission_30_days · PDF page(s) 18
- Measurement role
- classification_performance · PDF page(s) 18
- Metric
- auroc · PDF page(s) 18
- Model
- machine_learning · PDF page(s) 16, 18
- Reference type
- observed_clinical_reference · PDF page(s) 18
- Scenario
- internal_holdout · PDF page(s) 18, 26, 33
Validation Direct comparison · Same source · Unknown · Supports · Matched conditions · Not applicable
Model scenario comparison
a5f47270-158d-52cd-86e0-f2e79dfe584cPrediction source Version of record · PDF page(s) 21, 30
Study b95a5665-db87-5237-a379-a0e438da24eb · Prediction record (JSON)
Prediction fields
- Endpoint
- pk_exposure_auc_24h · PDF page(s) 21
- Measured variable
- pk_exposure_auc_24h · PDF page(s) 21
Model evaluation
c4a9fd5c-6c49-5f5b-bb82-7f34c10b42b5Prediction source Version of record · PDF page(s) 22, 26
Study b95a5665-db87-5237-a379-a0e438da24eb · Prediction record (JSON)
Prediction fields
- Endpoint
- pearson_correlation · PDF page(s) 22
- Family
- agreement · PDF page(s) 22
- Measurement role
- reference_agreement · PDF page(s) 22
- Metric
- pearson_correlation · PDF page(s) 22
- Role
- model_fit · PDF page(s) 22
- Unit
- dimensionless · PDF page(s) 22