基于机器学习预测肥厚型心肌病患者的基因型阳性率。
Prediction of Genotype Positivity in Patients With Hypertrophic Cardiomyopathy Using Machine Learning.
机构信息
Division of Cardiology, Department of Medicine (L.W.L., M.S.M., M.P.R., Y.J.S.), Columbia University Irving Medical Center, New York, NY.
Cardiology Division, Department of Medicine (M.A.F.), Massachusetts General Hospital, Boston.
出版信息
Circ Genom Precis Med. 2021 Jun;14(3):e003259. doi: 10.1161/CIRCGEN.120.003259. Epub 2021 Apr 23.
BACKGROUND
Genetic testing can determine family screening strategies and has prognostic and diagnostic value in hypertrophic cardiomyopathy (HCM). However, it can also pose a significant psychosocial burden. Conventional scoring systems offer modest ability to predict genotype positivity. The aim of our study was to develop a novel prediction model for genotype positivity in patients with HCM by applying machine learning (ML) algorithms.
METHODS
We constructed 3 ML models using readily available clinical and cardiac imaging data of 102 patients from Columbia University with HCM who had undergone genetic testing (the training set). We validated model performance on 76 patients with HCM from Massachusetts General Hospital (the test set). Within the test set, we compared the area under the receiver operating characteristic curves (AUROCs) for the ML models against the AUROCs generated by the Toronto HCM Genotype Score (the Toronto score) and Mayo HCM Genotype Predictor (the Mayo score) using the Delong test and net reclassification improvement.
RESULTS
Overall, 63 of the 178 patients (35%) were genotype positive. The random forest ML model developed in the training set demonstrated an AUROC of 0.92 (95% CI, 0.85-0.99) in predicting genotype positivity in the test set, significantly outperforming the Toronto score (AUROC, 0.77 [95% CI, 0.65-0.90], =0.004, net reclassification improvement: <0.001) and the Mayo score (AUROC, 0.79 [95% CI, 0.67-0.92], =0.01, net reclassification improvement: =0.001). The gradient boosted decision tree ML model also achieved significant net reclassification improvement over the Toronto score (<0.001) and the Mayo score (=0.03), with an AUROC of 0.87 (95% CI, 0.75-0.99). Compared with the Toronto and Mayo scores, all 3 ML models had higher sensitivity, positive predictive value, and negative predictive value.
CONCLUSIONS
Our ML models demonstrated a superior ability to predict genotype positivity in patients with HCM compared with conventional scoring systems in an external validation test set.
背景
基因检测可确定家族筛查策略，并在肥厚型心肌病（HCM）中具有预后和诊断价值。然而，它也会带来重大的社会心理负担。传统的评分系统对预测基因型阳性的能力有限。我们的研究目的是通过应用机器学习（ML）算法，为 HCM 患者的基因型阳性建立一个新的预测模型。
方法
我们使用来自哥伦比亚大学的 102 名 HCM 患者的临床和心脏成像数据构建了 3 个 ML 模型，这些患者已经接受了基因检测（训练集）。我们使用来自马萨诸塞州综合医院的 76 名 HCM 患者对模型性能进行了验证（测试集）。在测试集中，我们比较了 ML 模型的接收者操作特征曲线（AUROC）与多伦多 HCM 基因型评分（多伦多评分）和梅奥 HCM 基因型预测器（梅奥评分）生成的 AUROC，使用 Delong 检验和净重新分类改善（net reclassification improvement）。
结果
总体而言，178 名患者中有 63 名（35%）为基因型阳性。在测试集中，训练集中开发的随机森林 ML 模型在预测基因型阳性方面的 AUROC 为 0.92（95%CI，0.85-0.99），明显优于多伦多评分（AUROC，0.77 [95%CI，0.65-0.90]，=0.004，净重新分类改善：<0.001）和梅奥评分（AUROC，0.79 [95%CI，0.67-0.92]，=0.01，净重新分类改善：=0.001）。梯度提升决策树 ML 模型也明显优于多伦多评分（<0.001）和梅奥评分（=0.03），其 AUROC 为 0.87（95%CI，0.75-0.99）。与多伦多评分和梅奥评分相比，所有 3 个 ML 模型的敏感性、阳性预测值和阴性预测值都更高。
结论
与外部验证测试集中的传统评分系统相比，我们的 ML 模型在预测 HCM 患者的基因型阳性方面表现出优越的能力。
Zotero 插件