Machine learning models for diabetic neuropathy diagnosis using microcirculatory parameters in type 2 diabetes patients.

IF 1.5 4区医学 Q3 PERIPHERAL VASCULAR DISEASE

International Angiology Pub Date : 2023-06-01 DOI:10.23736/S0392-9590.23.05008-3

Xiaoyu Zhang, Yining Sun, Zuchang Ma, Liang Lu, Mengyuan Li, Xueya Ma

{"title":"Machine learning models for diabetic neuropathy diagnosis using microcirculatory parameters in type 2 diabetes patients.","authors":"Xiaoyu Zhang, Yining Sun, Zuchang Ma, Liang Lu, Mengyuan Li, Xueya Ma","doi":"10.23736/S0392-9590.23.05008-3","DOIUrl":null,"url":null,"abstract":"Background: Diabetic peripheral neuropathy (DPN) is a primary cause of diabetic foot, early detection of DPN is essential. This study aimed to construct a machine learning model for DPN diagnosis based on microcirculatory parameters, and identify the most predictive parameters for DPN.Methods: Our study involved 261 subjects, including 102 diabetics with neuropathy (DMN), 73 diabetics without neuropathy (DM), and 86 healthy controls (HC). DPN was confirmed by nerve conduction velocity and clinical sensory tests. Microvascular function was measured by postocclusion reactive hyperemia (PORH), local thermal hyperemia (LTH), and transcutaneous oxygen pressure (TcPO<inf>2</inf>). Other physiological information was also investigated. Logistic regression (LR) and other machine learning (ML) algorithms were used to develop the model for DPN diagnosis. Kruskal-Wallis Test (non-parametric) were performed for multiple comparisons. Several performance measures, such as accuracy, sensitivity and specificity, were used to access the efficacy of the developed model. All the features were ranked based on the importance score to find features with higher DPN predictions.Results: There was an overall decrease in microcirculatory parameters in response to PORH and LTH, as well as TcPO<inf>2</inf>, in DMN group compared to DM group and HC group. Random forest (RF) was found to be the best model, and achieved 84.6% accuracy along with 90.2% sensitivity and 76.7% specificity. RF_PF% of PORH was the main predictor of DPN. In addition, diabetic duration was also an important risk factor.Conclusions: PORH Test is a reliable screening tool for DPN, which can accurately distinguish DPN from diabetics using RF.","PeriodicalId":13709,"journal":{"name":"International Angiology","volume":"42 3","pages":"191-200"},"PeriodicalIF":1.5000,"publicationDate":"2023-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Angiology","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.23736/S0392-9590.23.05008-3","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"PERIPHERAL VASCULAR DISEASE","Score":null,"Total":0}

引用次数: 0

Abstract

Background: Diabetic peripheral neuropathy (DPN) is a primary cause of diabetic foot, early detection of DPN is essential. This study aimed to construct a machine learning model for DPN diagnosis based on microcirculatory parameters, and identify the most predictive parameters for DPN.

Methods: Our study involved 261 subjects, including 102 diabetics with neuropathy (DMN), 73 diabetics without neuropathy (DM), and 86 healthy controls (HC). DPN was confirmed by nerve conduction velocity and clinical sensory tests. Microvascular function was measured by postocclusion reactive hyperemia (PORH), local thermal hyperemia (LTH), and transcutaneous oxygen pressure (TcPO2). Other physiological information was also investigated. Logistic regression (LR) and other machine learning (ML) algorithms were used to develop the model for DPN diagnosis. Kruskal-Wallis Test (non-parametric) were performed for multiple comparisons. Several performance measures, such as accuracy, sensitivity and specificity, were used to access the efficacy of the developed model. All the features were ranked based on the importance score to find features with higher DPN predictions.

Results: There was an overall decrease in microcirculatory parameters in response to PORH and LTH, as well as TcPO2, in DMN group compared to DM group and HC group. Random forest (RF) was found to be the best model, and achieved 84.6% accuracy along with 90.2% sensitivity and 76.7% specificity. RF_PF% of PORH was the main predictor of DPN. In addition, diabetic duration was also an important risk factor.

Conclusions: PORH Test is a reliable screening tool for DPN, which can accurately distinguish DPN from diabetics using RF.

查看原文本刊更多论文

2型糖尿病患者使用微循环参数诊断糖尿病神经病变的机器学习模型。

背景:糖尿病周围神经病变(DPN)是糖尿病足的主要病因，早期发现DPN至关重要。本研究旨在构建基于微循环参数的DPN诊断机器学习模型，并识别最具预测性的DPN诊断参数。方法:本研究共纳入261例受试者，其中伴有神经病变的糖尿病患者102例，无神经病变的糖尿病患者73例，健康对照86例。通过神经传导速度和临床感觉检查证实DPN。通过术后反应性充血(PORH)、局部热充血(LTH)和经皮氧压(TcPO2)测量微血管功能。其他生理信息也被调查。采用逻辑回归(LR)和其他机器学习(ML)算法建立DPN诊断模型。多重比较采用Kruskal-Wallis检验(非参数检验)。几个性能指标，如准确性，敏感性和特异性，被用来访问所开发的模型的有效性。所有的特征都是根据重要性评分排序的，以找到具有更高DPN预测的特征。结果:与DM组和HC组相比，DMN组的poh、LTH及TcPO2对微循环参数的影响均有所降低。随机森林(Random forest, RF)是最佳模型，准确率为84.6%，灵敏度为90.2%，特异性为76.7%。PORH的RF_PF%是DPN的主要预测因子。此外，糖尿病病程也是一个重要的危险因素。结论:PORH试验是一种可靠的DPN筛查工具，可准确区分DPN与糖尿病。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

International Angiology 医学-外周血管病

CiteScore

2.80

自引率

28.60%

发文量

审稿时长

6-12 weeks

期刊介绍： International Angiology publishes scientific papers on angiology. Manuscripts may be submitted in the form of editorials, original articles, review articles, special articles, letters to the Editor and guidelines. The journal aims to provide its readers with papers of the highest quality and impact through a process of careful peer review and editorial work. Duties and responsibilities of all the subjects involved in the editorial process are summarized at Publication ethics. Manuscripts are expected to comply with the instructions to authors which conform to the Uniform Requirements for Manuscripts Submitted to Biomedical Editors by the International Committee of Medical Journal Editors (ICMJE).