Integration of Biomedical Engineering and Computer Vision for Morphological Alteration Identification in Diabetic Cardiomyopathy: An Analysis and Evaluation Strategy
Diabetic cardiomyopathy (DCM) is a type of heart disease that affects people with diabetes. It is characterized by changes in the structure and function of the heart, including thickening and stiffening of the heart muscle, impaired relaxation of the heart, and reduced pumping function. DCM is considered an ailment of the heart muscle and an increased risk factor in people with type 2 diabetes mellitus (T2DM). Although the exact mechanisms behind DCM are not fully understood, high blood glucose levels are known to contribute to the development and progression of the disease. While many options to manage and diagnose DCM have been established in the late stages, it remains challenging to identify the early stages of disease development and effectively prevent its progression. Therefore, this research combines biomedical engineering principles and computer vision applications to study the early-stage development of diabetic cardiomyopathy. By employing new programmatic biomedical imaging techniques and utilizing computer vision, the goal is to better diagnose and understand the disease by algorithmically evaluating a larger sample of images with automation and sophisticated image and geometry analysis. Comprehensive open source image analysis libraries are maintained and provided by Intel Corporation and have been used on a diversity of applications including the first autonomous crossing of California in a driverless car in 2011 and countless other applications in the biomedical, consumer (clothes fitting, augmented reality, etc), manufacturing, agriculture, and automotive fields. The central objective of this research is to utilize computer vision to identify and characterize the pathophysiology of DCM. By rapidly identifying and quantifying subtle geometric, color, and pixel features as well as spatial frequency information in biomedical images, computer vision enables a better understanding of the early stages of DCM, leading to improved diagnosis and treatment. Through computer vision analysis, valuable insights into the microscopic features of cardiac tissue affected by DCM, including cellular structure, fibrosis, inflammation, and other pathological changes, can be gained. This allows researchers to establish correlations, identify key biomarkers, and deepen their understanding of the disease mechanisms underlying DCM. Ultimately, the integration of computer vision with microscopic tissue analysis aims to enhance knowledge, enable early detection, and advance the development of more effective diagnostic and therapeutic approaches for DCM. Preliminary investigations of diabetic rodent hearts have revealed hyper-contracted and degenerated myofibers, disrupted collagen fibrils, and fragmented sarcoplasm, providing evidence of fibrosis. Myocardial fibrosis is a hallmark of hypertrophic cardiomyopathy and is proposed as a mechanism for arrhythmias and heart failure. To accomplish the overarching goal, three specific aims have been proposed. The first aim is to study the histopathology of diabetic rat myocardium to identify and establish specific biomarkers that may play a role in DCM. The second aim involves investigating the effects of Glycyrrhizin (GLC), anti-inflammatory agent on the identified biomarkers and evaluating its potential therapeutic impact on DCM. The third aim focuses on determining whether computer vision software can be utilized to develop a potential disease model for studying the progression of DCM in vitro under diabetogenic conditions, including GLC treatment. Successful completion of this study may lead to the establishment of parameters for identifying DCM, provide insights into the therapeutic effects of GLC, and offer a basis for studying the early development of DCM. Therefore, the integration of biomedical engineering, computer vision technology, and GLC treatment plays a crucial role in elucidating the etiology, initiation, and potential therapeutic interventions for Diabetic Cardiomyopathy (DCM). By leveraging these technologies, researchers can enhance their capability to simultaneously examine a large number of histopathological images and other structures, enabling the identification of complex morphology and multicellular environment associated with the early stages of cardiac tissue disease and evaluating the effects of GLC treatment on DCM progression.
Delgado, Monica, "Integration of Biomedical Engineering and Computer Vision for Morphological Alteration Identification in Diabetic Cardiomyopathy: An Analysis and Evaluation Strategy" (2023). ETD Collection for University of Texas, El Paso. AAI30632779.