人工智能三维术前规划在全髋关节置换术中的临床应用

Clinical Application of Artificial Intelligence Three-Dimensional Preoperative Planning in Total Hip Arthroplasty

目的探讨人工智能三维规划软件在全髋关节置换术中假体测量的准确性,并分析影响准确性的因素。方法选取2021年8月至2022年6月行单侧全髋关节置换术的57例患者,其中男20例,女37例;年龄48~84岁,平均(69.20±8.16)岁。术前收集患侧髋关节X线及CT三维重建等影像资料,由1名住院医师分别使用胶片模板和人工智能三维术前规划软件进行术前规划,对比术前预测假体型号与手术实际使用假体的符合程度,同时评估患者身体质量指数(body mass index,BMI),下肢长度差值是否与人工智能软件测量准确度相关。结果人工智能规划髋臼侧测量在±0、±1、±2假体型号的符合率分别为82.46%、10.53%、7.01%。传统模板测量的髋臼侧符合率分别为42.11%、43.86%、14.03%。两者比较差异有统计学意义(P<0.05)。人工智能股骨侧测量在±0、±1、±2型号符合率为77.19%、19.30%、3.51%。传统模板测量符合率为54.39%、36.84%、8.77%。两者比较差异有统计学意义(P<0.05)。人工智能三维规划系统预测髋臼杯型号同实际应用型号存在高度相关性(r=0.886,P<0.05),人工智能三维规划系统预测股骨假体型号同实际应用型号存在高度相关性(r=0.874,P<0.05)。BMI和人工智能测量臼杯及股骨柄假体准确度之间无显著相关性,差异无统计学意义(P>0.05)。下肢长度差和人工智能测量臼杯及股骨柄假体准确度之间无显著相关性,差异无统计学意义(P>0.05)。结论人工智能三维术前规划较胶片模板测量对全髋关节置换术中假体型号选择具有更好的预测作用,BMI及下肢长度差和测量准确度之间没有明显相关性。

Objective To investigate the accuracy of artificial intelligence three-dimensional planning software in the selection of prosthetic size in total hip arthroplasty and analyze the factors affecting the accuracy.Methods A total of 57 patients(20 males and 37 females)who underwent unilateral primary total hip arthroplasty were selected.The X-ray and CT 3D reconstruction image data of the affected hip joint were collected from August 2021 to June 2022 before operation.A resident make a preoperative planning for each total hip arthroplasty using both the traditional templating and artificial intelligence three-dimensional preoperative planning software.We analyzed the accuracy of implant size prediction using two methods.The coincidence rate of the two methods was compared.At the same time,whether the BMI and leg length discrepancy of patients were the factors affecting the accuracy of measurement.Results The coincidence rates of 0,±1 and±2 prosthesis size measurements of artificial intelligence planning on the acetabular side were 82.46%,10.53%and 7.01%,respectively.The acetabular concordance rate of traditional template was 42.11%,43.86%and 14.03%.The difference between the two groups was statistically significant(P<0.05).The concordance rate of artificial intelligence femoral side measurement in 0,±1,±2 model was 77.19%,19.30%,3.51%.The coincidence rate of traditional template was 54.39%,36.84%and 8.77%.The difference between the two groups was statistically significant(P<0.05).The prediction of acetabular cup model by artificial intelligence 3D programming system was highly correlated with the actual model(r=0.886,P<0.05),and a high correlation between prediction of femur prosthesis size and actual application type(r=0.874,P<0.05).There was no significant correlation between BMI and the accuracy of artificial intelligence measurement of mortar cup and femoral stem prosthesis(P>0.05).And no significant correlation between the lower limb length difference and the accuracy of artificial intelligence measurement of socket and femoral stem prosthesis(P>0.05).Conclusion Artificial intelligence three-dimensional planning has a better predictive effect on the selection of prosthesis size in total hip arthroplasty than film template measurement.There is no significant correlation between BMI and lower limb length difference and measurement accuracy.