A nomogram based on clinical factors and CT radiomics for predicting anti-MDA5+ DM complicated by rapidly progressive interstitial lung disease (RP-ILD)
Authors' conclusion
Our results suggest that the combination model built with clinical and radiomics features could reliably predict the occurrence of RP-ILD in MDA5+ DM patients.
Study context
- Population
- Patients with anti-MDA5+ DM72 training, 31 testing
- Participants
- 103 patients
- Comparison
- Clinical model vs radiomics model vs combined clinical + radiomics model for predicting RP-ILD
- Outcome
- Prediction of RP-ILD vs non-RP-ILD (AUC, sensitivity, specificity, accuracy)
Methodological detailsprediction model development, random split into training and testing sets, radiomics, nomogram
- Prediction model development
- Random split into training and testing sets
- Radiomics
- Nomogram
Performance
As reported in the abstract
Predictive performance for RP-ILD, testing set
| Test or model | AUC | Sensitivity | Specificity | Accuracy |
|---|---|---|---|---|
| Clinical model | 0.890 | 0.875 | 0.800 | 0.774 |
| Radiomics model | 0.754 | — not stated in the abstract | — not stated in the abstract | — not stated in the abstract |
| Combination model | 0.890 | 0.812 | 1.000 | 0.839 |
AUC
Clinical model0.890
Radiomics model0.754
Combination model0.890
Sensitivity
Clinical model0.875
Radiomics model— not stated in the abstract
Combination model0.812
Specificity
Clinical model0.800
Radiomics model— not stated in the abstract
Combination model1.000
Accuracy
Clinical model0.774
Radiomics model— not stated in the abstract
Combination model0.839
— not stated in the abstract
Further findings
| Outcome | Clinical model | Radiomics model | Combination model |
|---|---|---|---|
| Predictive performance for RP-ILD, training set | |||
| AUC | 0.954 | 0.805 | 0.994 |
| Sensitivity | 0.931 | — not stated in the abstract | 0.966 |
| Specificity | 0.837 | — not stated in the abstract | 0.977 |
| Accuracy | 0.847 | — not stated in the abstract | 0.931 |
— not stated in the abstract
Independent predictors of RP-ILD
- Dyspnoea
- P = 0.000
- Duration of illness in months
- P = 0.001
- Three radiomics features
- P = 0.001, 0.044 and 0.008
- No imaging features significantly different between the two groups
Nomogram calibration and decision curve analyses
- Satisfactory consistency and clinical utility of the nomogram
Original publication
A nomogram based on clinical factors and CT radiomics for predicting anti-MDA5+ DM complicated by RP-ILD. Rheumatology (Oxford). Published 2 June 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →