Artificial intelligence (AI) is transforming medical diagnostics, especially through Computer-Aided Diagnosis (CAD) systems in imaging exams. Deep learning techniques offer significant advantages, including centralized knowledge aggregation, bias mitigation, and faster workflows. These benefits result in earlier, more accurate diagnoses and can provide valuable second opinions in complex cases. YOLO (You Only Look Once) has emerged as a leading computer vision tool in medicine due to its single-stage neural network, which enables faster detection and classification with reduced processing time and computational cost. This paper focuses on the effectiveness of YOLOv8 for classifying and detecting pathology and clinical findings in X-ray scans, particularly spinal abnormalities. Key findings show impressive performance in classification, with an accuracy of 97.3\%, recall of 83.8\%, and F1-score of 98\%. Detection results demonstrate a mean Average Precision at 50\% (mAP50) of 90.2\%, indicating promising outcomes with the chosen dataset. These findings underscore the potential of YOLOv8x for accurate and efficient detection of spinal pathology in X-ray imaging, leading to earlier diagnoses and improved patient outcomes. Furthermore, the use of AI in medical diagnostics can reduce healthcare costs by streamlining workflows and democratize access to expert-level care, particularly in under-resourced areas.
