The development of a novel high-precision major depressive disorder screening system using transient autonomic responses induced by dual mental tasks
The development of a novel high-precision major depressive disorder screening system using transient autonomic responses induced by dual mental tasks
复制标题
DOI：
10.1080/03091902.2018.1435744
复制
发表时间：
2018-01-01
影响因子：
--
通讯作者：
Sun, Guanghao
中科院分区：
文献类型：
--
作者：
Matsui, Takemi;Shinba, Toshikazu;Sun, Guanghao
12.6%的重度抑郁症（MDD）患者有自杀倾向，而据报道，43%的患者没有就MDD咨询医生，自动MDD筛查令人热切期待。最近，为了实现MDD的自动化筛查，人们对多重DNA甲基化谱或使用近红外光谱(NIRS)的生理学方法等生物标志物进行了研究，但它们需要在采血后使用96孔DNA ELIZA试剂盒进行检查，或者成本高昂。使用单导联心电图 (ECG)，我们利用双重心理任务引起的短暂自主反应开发了高精度 MDD 筛查系统。我们开发了一种新型高精度MDD筛查系统，该系统由单导联心电监护仪、模数（AD）转换器和带有LabView编程语言编写的测量和分析程序的个人计算机组成。该系统通过采用HRV相关预测变量（心率（HR）、高频（HF）、低频（LF）/HF）的线性判别分析（LDA），利用由双重心理任务（即言语流利性任务和随机数生成任务）引起的热率变异性（HRV）衍生的瞬时自主反应来区分MDD患者与正常受试者。该系统针对来自静冈综合医院门诊部接受抗抑郁治疗的 12 名 MDD 患者（32 +/- 15 岁）和来自东京都立大学的 30 名正常志愿者（37 +/- 17 岁）进行了测试。所提出的系统将 42 名受试者分为 12 名 MDD 患者和 30 名正常受试者，实现了 100% 的敏感性和 100% 的特异性。所提出的系统似乎有望用于未来仅使用单导联心电图基于 HRV 的高精度和低成本 MDD 筛查。
12.6% of major depressive disorder (MDD) patients have suicide intent, while it has been reported that 43% of patients did not consult their doctors for MDD, automated MDD screening is eagerly anticipated. Recently, in order to achieve automated screening of MDD, biomarkers such as multiplex DNA methylation profiles or physiological method using near infra-red spectroscopy (NIRS) have been studied, however, they require inspection using 96-well DNA ELIZA kit after blood sampling or significant cost. Using a single-lead electrocardiography (ECG), we developed a high-precision MDD screening system using transient autonomic responses induced by dual mental tasks. We developed a novel high precision MDD screening system which is composed of a single-lead ECG monitor, analogue to digital (AD) converter and a personal computer with measurement and analysis program written by LabView programming language. The system discriminates MDD patients from normal subjects using heat rate variability (HRV)-derived transient autonomic responses induced by dual mental tasks, i.e. verbal fluency task and random number generation task, via linear discriminant analysis (LDA) adopting HRV-related predictor variables (hear rate (HR), high frequency (HF), low frequency (LF)/HF). The proposed system was tested for 12 MDD patients (32 +/- 15years) under antidepressant treatment from Shizuoka Saiseikai General Hospital outpatient unit and 30 normal volunteers (37 +/- 17years) from Tokyo Metropolitan University. The proposed system achieved 100% sensitivity and 100% specificity in classifying 42 examinees into 12 MDD patients and 30 normal subjects. The proposed system appears promising for future HRV-based high-precision and low-cost screening of MDDs using only single-lead ECG.