Derivation and external validation of a deep learning model to predict changes in coronary plaque burden

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: Accuracy 0.84 in external validation

• Primary Outcome: Prediction of coronary plaque progression vs regression using ΔPAV

• Nnt Or Sensitivity: Matthews correlation coefficient 0.68, F1 score 0.84 for both progression and regression classes

• Confidence Interval: Not reported for key performance metrics

Clinical Application

Limited by requirement for high-quality IVUS imaging and computational infrastructure for deep learning model deployment Relevant to Australian interventional cardiology centres with IVUS capability, though model would require TGA approval for clinical use and integration with existing imaging systems Patients with coronary artery disease undergoing IVUS imaging in research or clinical settings

Abstract

BACKGROUND: Predicting the progression/regression of coronary plaque burden is challenging. AIMS: We aimed to develop a deep learning model to forecast changes in percent atheroma volume (ΔPAV) using intravascular ultrasound (IVUS). METHODS: We analysed data from IBIS-4 and PACMAN-AMI. Core lab measurements of plaque burden were available from IVUS pullbacks. Each model consists of a bidirectional Long Short-Term Memory (biLSTM) layer followed by two fully connected layers with one neuron each, resulting in both a classification for input progression/regression and an estimation of the ΔPAV. RESULTS: For the derivation and validation, a total of 1,960 regions of interest (ROIs) from the IBIS-4 dataset were used. The mean±standard deviation of the model accuracy was 0.85±0.02, the Matthews correlation coefficient was 0.70±0.04, and the F1 score was 0.85±0.02 for both progression and regression classes. In the testing (external validation) process with the PACMAN-AMI dataset, 5,283 ROIs were utilised. The mean ΔPAV was -0.31±5.63, for which 2,665 featured regression with a mean ΔPAV of -4.57±3.73, and 2,618 presented progression with a mean ΔPAV of 4.02±3.55, representing 49.6% of plaque progression prevalence. The predictive performance across the 100 trained models in the testing dataset showed an accuracy of 0.84, a Matthews correlation coefficient of 0.68, and an F1 score for the progression and regression classes of 0.84. CONCLUSIONS: This is the first deep learning model capable of detecting changes in plaque progression by analysing the rate of plaque burden change between adjacent frames.

References

1. 1.García-García, H. M., Bulant, C. A., Boroni, G. A., Clausse, A., Engstrøm, T., Lemos, P. A., Lecaros Yap, N. A., Cap, M., Iglesias, J. F., Geuns, R. V., Lang, I. M., Spirk, D., Häner, J. D., Koskinas, K. C., Kakizaki, R., Ueki, Y., Siontis, G. C. M., Bourantas, C. V., Blanco, P. J., & Räber, L. (2026). Derivation and external validation of a deep learning model to predict changes in coronary plaque burden. EuroIntervention, Article EIJ-D-25-01352. https://doi.org/10.4244/EIJ-D-25-01352