The paper focuses on establishing a risk assessment model of femoral neck osteoporotic fracture (FNOF) in the elderly population and improving the screening efficiency and accuracy of such diseases in specific populations. In literature research, the main risk factors of femoral neck osteoporosis (FNOP) in the elderly were studied and analyzed; the femur region of interest (ROI) and the hard bone edge segmentation model were selected from the X-ray digital image by using the image depth learning method. On this basis, the femoral trabecular score and femoral neck strength (FNS) in the set region were selected as the main evaluation elements, and the quantitative analysis method was established; an X-ray image processing method was applied to the feasibility study of FNOP and compared with dual-energy X-ray absorptiometry measurements of bone mineral density; Finally, the main risk factors of FNOP were selected and the prediction model of FNOP in the elderly population was established based on medical image processing, machine learning model construction and other methods. Some FNOP health records were selected as test samples for comparative analysis with traditional manual evaluation methods. The paper shows the risk assessment model of FNOF in the elderly population, which is feasible in testing. Among them, the artificial neural network model had a better accuracy (95.83%) and recall rate (100.00%), and the support vector machine prediction model had high specificity (62.50%). With the help of a machine learning method to establish the risk assessment model of FNOF for the elderly, one can provide decision support for the fracture risk assessment of the elderly and remind the clinic to give targeted interventions for the above high-risk groups in order to reduce the fracture risk.
Keywords: X-ray; artificial intelligence; diagnosis; elderly patients; osteoporosis (OP).