Comprehensive risk assessment of patients following type-A aortic dissection surgery: A shape analysis and machine learning approach
Please login to view abstract download link
Background: After acute type-A aortic dissection (ATAAD) surgery, we observed true-lumen narrowing during follow-up examinations of patients, collected medical data at 14 days post-surgery, and assessed patient risk levels over 2.8 years. Early assessment is crucial for effective treatment, underscoring the need for a framework to identify the post-surgery risk level of aortic dissection cases. Purpose: Develop an implementable framework using mathematical techniques to predict the patients' risk due to true-lumen narrowing after ATAAD surgery. Materials and Methods: The retrospective study analyzed CT data from 21 ATAAD patients. True-, false-, and full-lumens were extracted from the dissected aorta, and 40 cross-sectional shapes (CSSs) were derived from each lumen to account for the progressive shape change. The Form Factor (FF) was introduced to assess CSS morphology. Linear discriminant analysis (LDA) with leave-one-patient-out cross-validation (LOPO-CV) was utilized for risk prediction. Results: Investigated 21 ATAAD patients (mean age 54.1 ± 9.1 years, 71.4% male) were categorized into high-risk (4 cases), medium-risk (4 cases), and low-risk (13 cases) based on clinical observations of the range of true-lumen narrowing. The risk classification machine-learning model uses four parameters, thus preserving its generalizability. It reliably identified low-risk patients and accurately predicted risk for high-risk patients, preserving generalizability. Conclusion: The proposed method anticipates aortic enlargement risk in patients with true-lumen narrowing just 14 days after ATAAD surgery, aiding cardiologists in enhancing patient care. REFERENCES [1] T. Watanabe, T. Ito, H. Sato, T. Mikami, et al. Morphological predictor of remodelling of the descending thoracic aortic false lumen that remains patent after repair of acute type-A dissection. Interdisciplinary Cardiovascular and Thoracic Surgery, 28(4), 629-634, 2019. [2] T. Ohara, H. Ikeda, Y. Sugitani, H. Suito, VQH Huynh, et al. Artificial intelligence supported anemia control system (AISACS) to prevent anemia in maintenance hemodialysis patients. International Journal of Medical Sciences,18(8), 1831-1839, 2021. [3] H. Sato, T. Ito, Y. Kuroda, H. Uchiyama, T. Watanabe, N. Kawaharada, et al. New predictor of aortic enlargement in uncomplicated type B aortic dissection based on elliptic Fourier analysis. European Journal of Cardio-Thoracic Surgery, 52, 1118-1124, 2017.