Heart failure & cardiomyopathies
Predicting prognosis of light-chain cardiac amyloidosis by magnetic resonance imaging and deep learning
Authors' conclusion
This fully automated DL model can predict individualized prognosis of AL-CA through LGE images, which complements the presently used Mayo staging method.
Study context
- Population
- Patients with AL-CA who underwent standardized chemotherapy and had at least 1 year of follow-uptesting set 79
- Participants
- 394 patients
- Comparison
- Deep learning (DL) model vs Mayo staging
- Outcome
- Prediction of survival (C-index, AUC)
- Follow-up
- 2.6 years
Methodological detailsdeep learning, Transformer-based, contrastive pretraining, ensemble learning
- Deep learning
- Transformer-based
- Contrastive pretraining
- Ensemble learning
Performance
As reported in the abstract
Prediction of 2.6-year survival (testing set)
| Test or model | Concordance index (C-index) | AUC | HR |
|---|---|---|---|
| DL model | 0.91 | 0.95 | 2.67 |
| Mayo model | 0.65 | 0.71 |
Concordance index (C-index)
DL model0.91
Mayo model0.65
AUC
DL model0.95
Mayo model0.71
HR
DL model2.67
Mayo modelNot reported in the abstract
Further findings
Discrimination within Mayo stage
- DL model effectively distinguished patients with the same Mayo stage but different prognoses
Original publication
Predicting prognosis of light-chain cardiac amyloidosis by magnetic resonance imaging and deep learning. Eur Heart J Cardiovasc Imaging. Published 2 September 2025. doi:10.
Summarised from the published abstract; every number is checked against it