目的 构建并对比维持性血液透析患者发生血管钙化的3种机器学习预测模型。 方法 选取300例维持性血液透析患者为研究对象，按照7:3的比例随机分为训练集(210例)和验证集(90例)，根据患者是否发生血管钙化将训练集分为有钙化组(124例)和无钙化组(86例)。分别采用Logistic回归、随机森林、支持向量机构建维持性血液透析患者发生血管钙化的预测模型，并使用验证集数据评价这3种机器学习预测模型的预测能力。 结果 在训练集中，Logistic回归模型、随机森林模型、支持向量机的AUC分别为0.835、0.886、0.872；在验证集中Logistic回归模型、随机森林模型、支持向量机的AUC分别为0.823、0.879、0.866。Delong检验显示3种机器学习预测模型的AUC具有差异(Z=2.663、2.751， P=0.003、0.001)。Logistic回归模型、随机森林模型、支持向量机模型均具有较好的一致性(χ2=4.018、4.661、3.892，P=0.642、0.887、0.739)。 结论 基于机器学习的维持性血液透析患者发生血管钙化的Logistic回归、随机森林、支持向量机模型均显示出较好的预测效果，其中随机森林模型的表现最好。
Objective To develop and compare three machine learning prediction models for predicting vascular calcification in patients with maintenance hemodialysis (MHD). Methods A total of 300 MHD patients were enrolled and randomly divided into a training set (n=210) and a validation set (n=90) in a 7:3 ratio. Based on the presence or absence of vascular calcification, the training set was further categorized into a calcification group (n=124) and a non-calcification group (n=86). Logistic Regression (LR), Random Forest (RF), and Support Vector Machine (SVM) models were developed to predict vascular calcification. The predictive performance of these models was evaluated using the validation set. Results In the training set, the area under curve (AUC) values for the LR, RF, and SVM models were 0.835, 0.886, and 0.872, respectively. In the validation set, the AUC values were 0.823, 0.879, and 0.866, respectively. DeLong's test showed significant differences in the AUC values among the three models (Z=2.663, 2.751; P=0.003, 0.001). All three models demonstrated good goodness-of-fit (χ²=4.018, 4.661, 3.892; P=0.642, 0.887, 0.739). Conclusion The LR, RF, and SVM machine learning models demonstrated good predictive performance for vascular calcification in MHD patients, with the Random Forest model showing superior performance.
[1]游奴佳,熊于勤,刘瑶,等.透析充分性对维持性血液透析患者血压变异性和血管钙化的影响[J].中国血液净化,2024,23(6):416-420.
[2]王新梓,邱溢博,邓雨,等.维持性血液透析患者血管钙化治疗的研究进展[J].实用临床医药杂志,2022,26(22):144-148.
[3]Li Q, Li P, Xu Z, et al. Association of diabetes with cardiovascular calcification and all-cause mortality in end-stage renal disease in the early stages of hemodialysis: a retrospective cohort study[J]. Cardiovasc Diabetol, 2024, 23(1): 259.
[4]Lee WT, Fang YW, Chang WS, et al. Data-driven, two-stage machine learning algorithm-based prediction scheme for assessing 1-year and 3-year mortality risk in chronic hemodialysis patients[J]. Sci Rep, 2023, 13(1): 21453.
[5]Hoang AT, Nguyen PA, Phan TP, et al. Personalised prediction of maintenance dialysis initiation in patients with chronic kidney disease stages 3-5: a multicentre study using the machine learning approach[J]. BMJ Health Care Inform, 2024, 31(1): e100893.
[6]Delrue C, De Bruyne S, Speeckaert MM. Application of machine learning in chronic kidney disease: current status and future prospects[J]. Biomedicines, 2024, 12(3): 568.
[7]Kato T, Torii S, Nakamura N, et al. Pathological analysis of medial and intimal calcification in lower extremity artery disease: impact of hemodialysis[J]. JACC Adv, 2023, 2(9): 100656.
[8]Nishibori N, Okazaki M, Miura Y, et al. Association of calciprotein particles with serum phosphorus among patients undergoing conventional and extended-hours haemodialysis[J]. Clin Kidney J, 2024, 17(6): 121.
[9]Cao Q, Shi Y, Liu X, et al. Analysis of factors influencing vascular calcification in peritoneal dialysis patients and their impact on long-term prognosis[J]. BMC Nephrol, 2024, 25(1): 157.
[10]田英,丁弘,曹婧媛,等.老年维持性血液透析病人心脏瓣膜钙化的相关因素分析[J].实用老年医学,2024,38(7):649-653.
[11]Petrovi? M, Brkovi? V, Barali? M, et al. Comparative analysis of vascular calcification risk factors in pre-hemodialysis and prevalent hemodialysis adult patients: insights into calcification biomarker associations and implications for intervention strategies in chronic kidney disease[J]. Diagnostics (Basel), 2024, 14(8): 824.
[12]Ha J, Jeong JC, Ryu JH, et al. Impact of arterial calcification on cardiovascular and renal outcomes in kidney transplant patients[J]. Kidney Dis (Basel), 2024, 10(4): 249-261.
[13]Jin J, Cheng M, Wu X, et al. Circulating miR-129-3p in combination with clinical factors predicts vascular calcification in hemodialysis patients[J]. Clin Kidney J, 2024, 17(3): 38.
[14]Wu J, Li X, Zhang H, et al. Development and validation of a prediction model for all-cause mortality in maintenance dialysis patients: a multicenter retrospective cohort study[J]. Ren Fail, 2024, 46(1): 2322039.
[15]Nakamura K, Isoyama N, Nakayama Y, et al. Association between amorphous calcium-phosphate ratios in circulating calciprotein particles and prognostic biomarkers in hemodialysis patients[J]. Sci Rep, 2022, 12(1): 13030.
[16]范慧,任直亲,杨美荣.老年糖尿病肾病患者行维持性血液透析发生血管钙化及其影响因素分析[J].老年医学与保健,2023,29(6):1175-1179.
[17]王颖,何华妮,杨海龙,等.高低通量血液透析对维持性血液透析患者微炎症、碱性磷酸酶水平和冠状动脉钙化的影响[J].临床与病理杂志,2023,43(1):130-136.
[18]Wang Y, Shen Q, Wang J, et al. The risk factors and predictive model for cardiac valve calcification in patients on maintenance peritoneal dialysis: a single-center retrospective study[J]. Ren Fail, 2023, 45(2): 2271069.
[19]Xiong L, Chen QQ, Cheng Y, et al. The relationship between coronary artery calcification and bone metabolic markers in maintenance hemodialysis patients[J]. BMC Nephrol, 2023, 24(1): 238.
[20]Xu Y, Li W, Yang Y, et al. Deep learning-based prediction of coronary artery calcium scoring in hemodialysis patients using radial artery calcification[J]. Semin Dial, 2024, 37(3): 234-241.
