Novel AI model enhances diagnosis of plaque erosion in patients with acute coronary syndromes

The novel "Transformer"-based DL model integrates information from adjacent frames emulating the cardiologists who review consecutive OCT frames to make a diagnosis and compare it with the standard CNN (convolutional neural network) DL model.

The researchers used a total of 237,021 cross-sectional OCT images from 581 patients for training and internal validation, and for external validation, 65,394 images from 292 patients from another dataset were used. Model performances were evaluated using AUC (area under the receiver-operating characteristic curve).

The study led to the following findings:

• For the frame-level diagnosis of plaque erosion, the Transformer model performed better than the CNN model, with an AUC of 0.94 compared with 0.85 in the external validation.
• For the lesion-level diagnosis, the Transformer model showed improved diagnostic performance compared with the CNN model, with an AUC of 0.91 compared with 0.84 in the external validation.

The results revealed that the novel transformer-based DL model could help in the accurate diagnosis of plaque erosion in patients with acute coronary syndromes and aid clinical decision-making -- making on potential conservative management without stenting in these patients.

Identifying plaque erosion reliably is an initial step to standing-up trials testing the hypotheses developed around more conservative non-interventional therapy. How this interesting automated approach performs when deployed more broadly will be of great interest.

Reference:

Park S, Araki M, Nakajima A, Lee H, Fuster V, Ye JC, Jang IK. Enhanced Diagnosis of Plaque Erosion by Deep Learning in Patients With Acute Coronary Syndromes. JACC Cardiovasc Interv. 2022 Oct 24;15(20):2020-2031. doi: 10.1016/j.jcin.2022.08.040. PMID: 36265933.