Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-1323
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-10-11T06:55:17.857825Z
status: published
lens: g_space
sector: health
headline: Development and Validation of an Interpretable Machine Learning Model for Osteoporosis Prediction Using Multimodal Musculoskeletal Ultrasound and Clinical Features
dek: This study aimed to develop and validate an interpretable machine learning model for osteoporosis prediction based on multimodal ultrasound and clinical features. Participants with suspected osteoporosis were enrolled from 2 medical centers. Clinical variables and multimodal musculoskeletal ultrasound parameters, including muscle thickness, cross-sectional area, and shear wave elastography measurements, were collected. LASSO regression was used for feature selection. Eight machine learning algorithms, including…
gain_title: An interpretable XGBoost model combining clinical variables and multimodal musculoskeletal ultrasound achieved high accuracy for osteoporosis prediction, offering a practical tool for early risk stratification.
problem_title: (none)
trace_subject: (none)
gain_reading: An interpretable XGBoost model combining clinical variables and multimodal musculoskeletal ultrasound achieved high accuracy for osteoporosis prediction, offering a practical tool for early risk stratification.
gain_evidence: An interpretable machine learning model based on multimodal musculoskeletal ultrasound and clinical features demonstrated promising performance for osteoporosis prediction. | The XGBoost model demonstrated accuracy of 0.889, 0.825, and 0.773 in the training, validating, and testing sets. | That may provide a practical and explainable tool for early osteoporosis risk stratification and individualized assessment.
problem_reading: (none)
problem_evidence: (none)
quick_read: Researchers developed and validated an interpretable machine learning model for osteoporosis prediction using clinical variables and multimodal musculoskeletal ultrasound from participants enrolled at two medical centers. After LASSO feature selection, eight algorithms were compared, and the XGBoost model achieved accuracies of 0.889 in training, 0.825 in internal validation, and 0.773 in external testing, with SHAP used to explain feature contributions.

The work matters because it points toward a non-invasive, explainable screening approach that could support early osteoporosis risk stratification and individualized assessment in clinical care. As of the October 2026 publication date, results are reported as model performance in research cohorts, not as deployed clinical impact, and generalizability beyond the two centers and integration into care pathways remain to be established.
limitation: 
tag: Evidence-backed gain
key_points: Participants with suspected osteoporosis were enrolled from 2 medical centers with clinical and ultrasound parameters collected. | LASSO regression selected age, GmhMT, RFCSA, gender, BMI, and TibialisSWE as key features. | Eight algorithms were compared including logistic regression, decision tree, random forest, k-nearest neighbors, support vector machine, naïve Bayes, neural network, and XGBoost. | Model interpretability was assessed using SHapley Additive exPlanations (SHAP) with personalized prediction explanations.
rundown: Researchers collected muscle thickness, cross-sectional area, and shear wave elastography measurements alongside age, gender, and BMI, then applied LASSO regression for feature selection.

Performance was evaluated in training, internal validation, and external test cohorts using AUC, accuracy, and F1 score, with SHAP analysis ranking importance as age, GmhMT, RFCSA, gender, BMI, and TibialisSWE.
sources:
- peer_reviewed | Journal of Ultrasound in Medicine | https://doi.org/10.1002/jum.70508 | 2026-10-04
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

52f48dbe75352b1ddc114c96cd19c47843b3760c00d0e53b3ac267bbdbb28679

previous

0000000000000000000000000000000000000000000000000000000000000000