The prevalence of thoracic vertebral fractures identified with the support of Artificial Intelligence-assisted diagnostic methods in patients hospitalized due to internal medicine diseases

Abstract

Objectives The number of internal medicine inpatients is increasing, and elderly individuals account for most hospitalizations. Vertebral fractures should be considered in patients hospitalized for internal medicine. This study aimed to investigate thoracic vertebral fractures in patients hospitalized for internal medicine diseases. Methods This study included 2,977 patients who underwent chest computed tomography (CT). Chest CT images were reconstructed to visualize the thoracic spine. The anterior, middle, and posterior vertebral heights were measured using artificial intelligence (AI). A vertebral fracture was defined as a height loss of ≥20% at a vertebra. Results The prevalence rates of thoracic vertebral fractures in all patients, males, and females were 31.6%, 25.2%, and 38.7%, respectively, and the prevalence rates of severe thoracic vertebral fractures were 12.4%, 9.2%, and 15.9%, respectively. The prevalence rates of multiple thoracic vertebral fractures in all patients, males, and females were 9.6%, 5.0%, and 14.6%, respectively. Conclusion The high prevalence of thoracic vertebral fractures among older internal medicine inpatients underscores the need for routine vertebral assessment. Incorporating AI-assisted CT evaluation may enhance the early detection and management of osteoporosis-related fractures in this vulnerable population.

Ähnliche Arbeiten