Machine Learning–Based Survival Prediction Tool for Adrenocortical Carcinoma
Authors' conclusion
S-GRAS parameters can efficiently predict outcome in patients with ACC, even using a robust ML model approach. Our web app instantly estimates the mortality and disease progression for patients with ACC, representing an accessible tool to drive personalized management decisions in clinical practice.
Study context
- Population
- Patients with adrenocortical carcinoma (ACC); S-GRAS variables (tumor stage, grade, resection status, age, symptoms)
- Participants
- 942 vs 152
- Comparison
- Training vs Validation
- Trainingn = 942
- Validationn = 152
- Outcome
- Prediction of 5-year overall mortality (OM) and 1-year and 3-year disease progression (DP)
Methodological detailsmachine learning, independent validation cohort, 16 machine learning (ML) models
- Machine learning
- Independent validation cohort
- 16 ML models
Performance
As reported in the abstract
F1 scores of best models (QDA, LightGBM, AdaBoost) for 5-year OM, 1-year DP and 3-year DP
| Test or model | Sensitivity 5-year overall mortality (OM) | Specificity 5-year OM | F1 5-year OM | F1 1-year disease progression (DP) | F1 3-year DP |
|---|---|---|---|---|---|
| Training | 77% | 77% | 0.79 | 0.63 | 0.83 |
| Validation | 65% | 81% | 0.72 | 0.60 | 0.83 |
Sensitivity 5-year overall mortality (OM)
Training77%
Validation65%
Specificity 5-year OM
Training77%
Validation81%
F1 5-year OM
Training0.79
Validation0.72
F1 1-year disease progression (DP)
Training0.63
Validation0.60
F1 3-year DP
Training0.83
Validation0.83
Further findings
- Web-based tool developed for individualized risk prediction of mortality and DP (https://acc-survival.streamlit.app)
Original publication
Machine Learning–Based Survival Prediction Tool for Adrenocortical Carcinoma. J Clin Endocrinol Metab. Published 14 February 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →