Objective To explore the accuracy of artificial intelligence 3D programming software in prosthesis measurement during total hip arthroplasty,analyze the factors affecting the accuracy,and preliminarily verify its clinical application value.Methods A total of 80 patients who underwent unilateral total hip arthroplasty from August 2021 to August 2022 were selected,including 40 patients who underwent preoperative planning using artificial intelligence 3D programming software and 40 patients who used film template for preoperative planning,to predict the size of acetabulum cup and femoral stalk.The degree of consistency between preoperative prediction of prosthesis size and the actual use of prosthesis,preoperative and postoperative difference of bilateral femoral eccentricity,bilateral combined eccentricity difference,length difference of both lower limbs,planning time,operation time and blood loss were compared between the two groups.We analyzed the accuracy of implant size prediction using two methods.Meanwhile,the effects of body mass index(BMI)and lower limb length difference on the measurement accuracy of artificial intelligence software were evaluated.Results Artificial intelligence planning acetabular side measurement at 0,±1,±2pseudobody size coincidence rates of 82.5%,15%,2.5%.The coincidence rates of acetabular side measured by traditional template were 42.5%,40% and 17.5%.The AI femur side measurement at 0,±1,±2 model coincidence rate was 72.5%,22.5%,5%.The coincidence rates of traditional template measurement were 47.5%,40% and 12.5%.The difference between them was statistically significant,P < 0.05.The kappa values of correlation between acetabular cup and femoral prosthesis size predicted by artificial intelligence 3D programming system and actual application model were 0.778 and0.747,P < 0.05.The kappa values of correlation analysis between acetabular cup and femoral prosthesis size predicted by film template and actual application model were0.259 and 0.359,P < 0.05.The preoperative lower limb length difference,bilateral femoral eccentric distance difference and bilateral combined eccentric distance difference of the experimental group were 11.7±11.62 mm,4.3±3.07 mm and8.83±5.58 mm,respectively,while those of the control group were 11.75±9.85 mm,3.4±2.71 mm and 7.45±6.02 mm,respectively.There was no statistical significance between the two groups,P > 0.05.The length difference of lower extremities,bilateral femoral eccentricity difference and bilateral combined eccentricity difference of experimental group measured immediately after surgery were 4.45±3.39 mm,3.2±3.36 mm and 6.85±5.74 mm.While those of control group were 5.65±3.96 mm,4.3±2.55 mm and 7.01±5.89 mm,respectively.There was no statistical significance between the two groups,P > 0.05.The preoperative planning time of experimental group was 5.07±1.86 min and that of control group was 3.42±1.37 min,and the difference was statistically significant,P<0.05.The operation time of the experimental group was 114.7±39.41 min,and the intraoperative blood loss was 246.25±202.98 ml.The operation time of the control group was 118.45±38.86 min,and the intraoperative blood loss was 258.75.67±141.82 ml.There was no statistical significance between the groups,P > 0.05.During the follow-up period,the Harris scores of the experimental group were75.69±4.12 points,82.95±3.63 points and 85.18±4.42 points at 1 month,3months and 6 months.Respectively,while those of the control group were 74.34±3.41 points,80.29±4.84 points and 84.42±3.94 points,and there was no statistical significance between the two groups,P > 0.05.The correlation coefficient between BMI and fitting degree of prosthesis was rs=-0.121,P > 0.05,and the correlation coefficient of femoral stalk was rs=-0.037,P > 0.05.The correlation coefficient between lower limb length difference and prosthesis fit was rs= 0.032,P > 0.05,and the correlation coefficient of femur stalk was rs= 0.006,P > 0.05.Conclusion 3D programming of artificial intelligence can be applied to preoperative planning of THA.Compared with film template,it has higher accuracy and reproducibility of prosthesis measurement,which provides a reliable basis for surgeons to select appropriate joint prosthesis and has good clinical application value. |