Application of logistic regression and machine learning methods for idiopathic inflammatory myopathies malignancy prediction
Authors' conclusion
The LR algorithm appears to be a good predictor of malignancy
Study context
- Population
- Patients diagnosed with idiopathic inflammatory myopathies (IIM), Shantou Central hospital, 2013 to 2021
- Participants
- 168 patients
- Exposure
- Logistic regression (LR) vs five other machine learning (ML) algorithm models for predicting malignancy in IIM
- Outcome
- Prediction of malignancy in IIM (AUC of ROC)
Methodological detailsprediction model development and validation, single-centre, random split into training (70%) and validation (30%) sets, six ML algorithms compared, multivariable regression
- Prediction model development and validation
- Single-centre
- Random split into training (70%) and validation (30%) sets
- Six ML algorithms compared
- Multivariable regression
Main finding
As reported in the abstract
Malignancy prediction, LR model
AUC of ROC, LR, training set
0.900
AUC of ROC, LR, validation set
0.784
Observational data: associations, not effects.
Further findings
Predictors of malignancy (multivariable regression)
- Risk factors: age, ALT<80U/L, anti-TIF1-γ
- ILD a protective factor
- Nomogram constructed from the four factors; web version built
Original publication
Application of logistic regression and machine learning methods for idiopathic inflammatory myopathies malignancy prediction. Clin Exp Rheumatol. Published 1 March 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →