Deep Learning for Bone Mineral Density and T-Score Prediction from Chest X-rays: A Multicenter Study.

Although the number of patients with osteoporosis is increasing worldwide, diagnosis and treatment are presently inadequate. In this study, we developed a deep learning model to predict bone mineral density (BMD) and T-score from chest X-rays, which are one of the most common, easily accessible, and low-cost medical imaging examination methods. The dataset used in this study contained patients who underwent dual-energy X-ray absorptiometry (DXA) and chest radiography at six hospitals between 2010 and 2021. We trained the deep learning model through ensemble learning of chest X-rays, age, and sex to predict BMD using regression and T-score for multiclass classification. We assessed the following two metrics to evaluate the performance of the deep learning model: (1) correlation between the predicted and true BMDs and (2) consistency in the T-score between the predicted class and true class. The correlation coefficients for BMD prediction were hip = 0.75 and lumbar spine = 0.63. The areas under the curves for the T-score predictions of normal, osteopenia, and osteoporosis diagnoses were 0.89, 0.70, and 0.84, respectively. These results suggest that the proposed deep learning model may be suitable for screening patients with osteoporosis by predicting BMD and T-score from chest X-rays.

Femoral Valgus Correction Angle for the Intramedullary Alignment Rod Is Strongly Associated with Femoral Lateral Bowing in Japanese Patients with Varus Knee Osteoarthritis Undergoing Total Knee Arthroplasty.

Background: This study aimed to investigate factors, such as differences in femoral shape, that could affect the femoral valgus correction angle (VCA) for the intramedullary alignment rod (IM rod) by using a three-dimensional (3D) measurement system in patients with varus knee osteoarthritis undergoing total knee arthroplasty (TKA). Methods: A total of 305 knees in 233 Japanese patients with varus knee osteoarthritis who underwent primary TKA by using Jig Engaged 3D Pre-Operative Planning Software for the TKA operation support system was examined. We retrospectively analysed factors, such as the shape of the proximal, middle, and distal femur in the coronal plane, all of which could affect the VCA for the IM rod, by multiple linear regression analyses. Results: The VCA for the IM rod was 5.9° ± 1.6° (range: 1.7° to 10.7°), and the femoral lateral bowing angle (FBA) was 3.5° ± 3.2°. Major factors independently associated with the VCA for the IM rod were the FBA (ß: 0.75), femoral offset (ß: 0.38), and the medial angle between the mechanical femoral axis and the line that connects the distal margins of the medial and lateral femoral condyles (ß: -0.16). The model was created by stepwise multiple linear regression (F = 266.6, p < 0.001, and estimated effect size = 4.4) explained 85% of the variance in the VCA for the IM rod (R 2 = 0.85). Conclusions: The VCA for the IM rod was most strongly associated with femoral lateral bowing in patients with varus knee osteoarthritis undergoing TKA. Our findings suggest that preoperatively measuring the VCA for the IM rod in patients with femoral lateral bowing by using a 3D measurement system could be useful for accurate coronal alignment of the femoral component in TKA.