IMPACTS OF HYPERPARATHYROIDISM ON ATHLETIC PERFORMANCE: A DEEP LEARNING APPROACH TO ENHANCING DIAGNOSIS THROUGH ULTRASOUND IMAGING

Abstract

Objective: Hyperparathyroidism is a metabolic condition characterized by the excessive secretion of parathyroid hormone, which can significantly impact bone density and muscular function—key factors in athletic performance. This study aims to leverage deep learning models for enhancing the diagnostic accuracy of hyperparathyroidism through ultrasound imaging, thereby helping in the management of athletes' health. Methods: This research develops a deep learning algorithm that processes ultrasound images to detect hyperparathyroidism. The approach begins with segmenting the ultrasound images into smaller blocks using an advanced image segmentation algorithm. Subsequently, critical features such as texture, gray-level co-occurrence matrix, and gradient features are extracted for analysis. Results: The implementation of the deep learning model has shown promising results, with a diagnostic accuracy rate of up to 99%. These findings suggest a significant enhancement in identifying hyperparathyroidism, which is crucial for managing the health of athletes. Conclusion: The study confirms that deep learning models are highly effective in improving the diagnostic accuracy of hyperparathyroidism using ultrasound imaging. This advancement could aid sports medicine professionals in better understanding and managing the impact of this condition on athletes, ultimately leading to optimized performance and reduced risk of injury related to poor bone health and muscle weakness.