Development and testing of Machine Learning Algorithms for early diagnosis of Amyloid Transthyretin Cardiomyopathy

Transthyretin amyloid cardiomyopathy (ATTR-CM) is a rare yet fatal condition characterized by the deposition of transthyretin amyloid fibrils in the heart. This review article synthesizes the findings of a proposed study aimed at comprehensively understanding ATTR-CM in the Indian population. The diagnosis of transthyretin amyloidosis (ATTR) cardiac disease faces several constraints that complicate early and accurate detection. One major challenge is the nonspecific nature of its clinical presentation, often mimicking other more common conditions like hypertensive heart disease or hypertrophic cardiomyopathy. This overlap can delay proper identification and lead to misdiagnoses. Additionally, the gold standard diagnostic tools, such as endomyocardial biopsy and advanced imaging techniques like cardiac MRI or scintigraphy with technetium-labeled compounds, are not always readily available, especially in resource-limited settings. Genetic testing, although essential for distinguishing hereditary from wild-type ATTR, may also be limited by access and cost. Furthermore, a lack of awareness and clinical suspicion among healthcare providers can result in under diagnosis or late diagnosis, which significantly impacts patient outcomes. The study intends to identify and analyze patients diagnosed with ATTR-CM in India to estimate its prevalence and describe patient characteristics, including gender differences and mortality rates. Moreover, it seeks to investigate the significance of early symptoms ("red flags") in identifying ATTR-CM and to develop and evaluate machine learning algorithms for its early diagnosis. Patients with ATTR-CM will be identified retrospectively using diagnosis codes and diagnostic algorithms, and compared with matched non-ATTR heart failure patients. Electronic records will be utilized for algorithm development and testing. Anticipated outcomes include providing the first statewide estimates of ATTR-CM prevalence and risk factors in India, emphasizing the disease's severity, and underlining the importance of early diagnosis, particularly among female patients, to facilitate effective treatment and disease progression prevention. Additionally, the study aims to demonstrate the utility of machine learning algorithms in early disease identification and detecting missed diagnoses. This review highlights the paucity of studies examining the prevalence of ATTR cardiomyopathy in India and underscores the need for machine learning algorithms for early detection, offering valuable insights into addressing this critical healthcare challenge.