Comprehensive Summary
This research is primarily focused on the use of an artificial intelligence-based tool to accurately test for and evaluate anomalous aortic origin of the coronary artery (AAOCA). To test the efficacy of the artificial intelligence tools, the models were given access to multiple databases with coronary CT angiography (CCTA) images, which are used to diagnose AAOCA. The models were given the parameters and conditions to diagnose the patients with the condition, and the subsequent results were then analyzed. All of the models scored an area-under-curve (AUC) of greater than or equal to 0.99, which is very strong, with varying sensitivities between the models between 0.95 and 0.99. These models are fully automated and extremely accurate in diagnosing cases of AAOCA from CCTA imaging, and can be easily integrated into healthcare systems.
Outcomes and Implications
AAOCA is a condition that is extremely difficult to diagnose, and is often misdiagnosed or overlooked by clinicians. The disease itself is extremely dangerous if left untreated, leading to ischemia or sudden cardiac death, emphasizing the need for precise diagnostic measures, which these artificial intelligence tools can provide. As a result, through the implementation of this practical tool, the number of patient deaths due to AAOCA can potentially decrease steeply through the early diagnosis and treatment of the condition.