Authors:

Phuong Anh Le

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Background: Takotsubo Cardiomyopathy (TCM), also known as "Broken Heart Syndrome," is characterized by chest pain, ST-segment elevation, elevated troponin levels, apical ballooning, reduced ejection fraction (EF), and angiographically normal coronary arteries. While emotional stress is recognized as a trigger, the underlying mechanism of chest pain in TCM remains unclear. This study hypothesized that abnormal coronary flow - specifically, prolonged arterial phase (AP) -may contribute to chest pain in TCM. Methods: Fifteen patients with TCM were evaluated using dynamic coronary angiography and artificial intelligence (AI) analysis. Inclusion criteria included chest pain, ST-segment changes, elevated troponin, apical ballooning, and non-obstructive coronary arteries. Ten patients controls with minimal CAD lesion and no left ventricular dysfunction served as a comparison group. Coronary flow was recorded at 15 frames per second. An AI model was trained by combining U-net and DenseNet-121. The model identified AP onset and completion based on contrast washout. AP duration was calculated from the first frame of blood inflow to complete contrast clearance from the distal coronary tree. All TCM patients were followed for one year. Results: The mean AP duration was significantly prolonged in TCM patients (3.17 ± 0.53 s) compared with controls (1.62 ± 0.26 s; p < 0.001). Mean EF improved from 33.9 ± 7.6% to 58.8 ± 12.6% at two weeks. At one-year follow-up, all patients were asymptomatic, except one who died from sepsis unrelated to cardiac disease. Conclusion: † Presenting Undergrad Author; ‡ Contributing Undergrad Author; * Undergrad Acknowledgment Prolonged coronary arterial phase in TCM may underlie chest pain and transient myocardial dysfunction. Despite initial severity, patient outcomes were favorable. Keywords: Takotsubo Cardiomyopathy; Artificial Intelligence; Arterial Phase; Ejection Fraction; Chest Pain

Source:

Purdue University / 2025

Topics:

No topics listed

Co-authors:

Phuong Anh Le