RESUMEN
Assessing the individual risk of Major Adverse Cardiac Events (MACE) is of major importance as cardiovascular diseases remain the leading cause of death worldwide. Quantitative Myocardial Perfusion Imaging (MPI) parameters such as stress Myocardial Blood Flow (sMBF) or Myocardial Flow Reserve (MFR) constitutes the gold standard for prognosis assessment. We propose a systematic investigation of the value of Artificial Intelligence (AI) to leverage [ 82 Rb] Silicon PhotoMultiplier (SiPM) PET MPI for MACE prediction. We establish a general pipeline for AI model validation to assess and compare the performance of global (i.e. average of the entire MPI signal), regional (17 segments), radiomics and Convolutional Neural Network (CNN) models leveraging various MPI signals on a dataset of 234 patients. Results showed that all regional AI models significantly outperformed the global model ( p < 0.001 ), where the best AUC of 73.9% (CI 72.5-75.3) was obtained with a CNN model. A regional AI model based on MBF averages from 17 segments fed to a Logistic Regression (LR) constituted an excellent trade-off between model simplicity and performance, achieving an AUC of 73.4% (CI 72.3-74.7). A radiomics model based on intensity features revealed that the global average was the least important feature when compared to other aggregations of the MPI signal over the myocardium. We conclude that AI models can allow better personalized prognosis assessment for MACE.
Asunto(s)
Imagen de Perfusión Miocárdica , Tomografía de Emisión de Positrones , Humanos , Imagen de Perfusión Miocárdica/métodos , Femenino , Masculino , Tomografía de Emisión de Positrones/métodos , Persona de Mediana Edad , Anciano , Inteligencia Artificial , Radioisótopos de Rubidio , Pronóstico , Redes Neurales de la Computación , Enfermedades Cardiovasculares/diagnóstico por imagen , Enfermedades Cardiovasculares/diagnóstico , Circulación CoronariaRESUMEN
Angiogenesis is an essential part of the cardiac repair process after myocardial infarction, but its spatiotemporal dynamics remain to be fully deciphered.68Ga-NODAGA-Arg-Gly-Asp (RGD) is a PET tracer targeting αvß3 integrin expression, which is a marker of angiogenesis. Methods: In this prospective single-center trial, we aimed to monitor angiogenesis through myocardial integrin αvß3 expression in 20 patients with ST-segment elevation myocardial infarction (STEMI). In addition, the correlations between the expression levels of myocardial αvß3 integrin and the subsequent changes in 82Rb PET/CT parameters, including rest and stress myocardial blood flow (MBF), myocardial flow reserve (MFR), and wall motion abnormalities, were assessed. The patients underwent 68Ga-NODAGA-RGD PET/CT and rest and stress 82Rb-PET/CT at 1 wk, 1 mo, and 3 mo after STEMI. To assess 68Ga-NODAGA-RGD uptake, the summed rest 82Rb and 68Ga-NODAGA-RGD images were coregistered, and segmental SUVs were calculated (RGD SUV). Results: At 1 wk after STEMI, 19 participants (95%) presented increased 68Ga-NODAGA-RGD uptake in the infarcted myocardium. Seventeen participants completed the full imaging series. The values of the RGD SUV in the infarcted myocardium were stable 1 mo after STEMI (1 wk vs. 1 mo, 1.47 g/mL [interquartile range (IQR), 1.37-1.64 g/mL] vs. 1.47 g/mL [IQR, 1.30-1.66 g/mL]; P = 0.9), followed by a significant partial decrease at 3 mo (1.32 g/mL [IQR, 1.12-1.71 g/mL]; P = 0.011 vs. 1 wk and 0.018 vs. 1 mo). In segment-based analysis, positive correlations were found between RGD SUV at 1 wk and the subsequent changes in stress MBF (Spearman ρ: r = 0.17, P = 0.0033) and MFR (Spearman ρ: r = 0.31, P < 0.0001) at 1 mo. A negative correlation was found between RGD SUV at 1 wk and the subsequent changes in wall motion abnormalities at 3 mo (Spearman ρ: r = -0.12, P = 0.035). Conclusion: The present study found that αvß3 integrin expression is significantly increased in the infarcted myocardium 1 wk after STEMI. This expression remains stable after 1 mo and partially decreases after 3 mo. Initial αvß3 integrin expression at 1 wk is significantly weakly correlated with subsequent improvements in stress MBF, MFR, and wall motion analysis.