论著

基于机器学习的功能性近红外光谱信号识别早期帕金森病患者的研究

  • 于津 ,
  • 汪杰 ,
  • 王虎军 ,
  • 王丛笑 ,
  • 李瑛琦 ,
  • 方伯言 ,
  • 王颖鹏
展开
  • 首都医科大学附属北京康复医院,北京 100144
王颖鹏 E-mail: ypwang@ccmu.edu.cn

收稿日期: 2023-12-04

  录用日期: 2024-06-07

  网络出版日期: 2025-02-25

基金资助

北京康复医院院内课题(2022-029)

Study on the recognition of early-stage Parkinson’s disease patients using functional near-infrared spectroscopy signals based on machine learning

  • YU Jin ,
  • WANG Jie ,
  • WANG Hujun ,
  • WANG Congxiao ,
  • LI Yingqi ,
  • FANG Boyan ,
  • WANG Yingpeng
Expand
  • Beijing Rehabilitation Hospital, Capital Medical University, Beijing 100144, China

Received date: 2023-12-04

  Accepted date: 2024-06-07

  Online published: 2025-02-25

摘要

目的:探索应用功能性近红外光谱(functional near-infrared spectroscopy, fNIRS)信号结合机器学习算法对早期PD患者进行诊断的可行性。方法:研究连续纳入自2021年12月至2023年8月期间在首都医科大学附属北京康复医院确诊的60例PD患者和60名健康对照者,使用22个通道(channel, CH)的ETG-4000型近红外脑功能成像仪采集受试者前额叶氧合血红蛋白和脱氧血红蛋白浓度变化值,使用一般线性模型计算每个通道激活程度β值。构建4种机器学习诊断模型,即支持向量机(support vector machine, SVM)、反向传播(back-propagation, BP)神经网络、随机森林和逻辑回归模型。采用准确率、灵敏度、特异度、受试者操作特征(receiver operating characteristic, ROC)曲线下面积对4种诊断学模型的效果进行评价。此外,使用SHAP(SHapley Additive exPlanations)技术来提高最优模型的可解释性,计算每个通道的SHAP值,将不同通道SHAP值进行加权平均汇总后,结合脑区分布,得到不同脑区对于模型分类任务的贡献比例。结果:4种诊断模型的准确率范围为81%~90%,灵敏度为69%~89%,特异度为93%~100%,ROC曲线下面积为0.90~0.98。其中,SVM模型表现最佳, ROC曲线下面积为0.96,准确率为90%,灵敏度为89%,特异度为93%。SHAP分析显示对于SVM模型贡献最大的4个通道为:CH08、CH05、CH01和CH13,其中右侧前额极皮层(frontopolar cortex,FPC)区域占比最大占总贡献36.5%。结论:基于fNIRS信号和SVM算法构建的模型在诊断早期PD患者中表现出诊断优势,其灵敏度(89%)和特异度(93%)均优于大多数现有方法。未来的研究应重点关注右侧前额极皮层区域和背外侧前额叶皮层区域的fNIRS信号特征,以进一步提高诊断模型的效能。

本文引用格式

于津 , 汪杰 , 王虎军 , 王丛笑 , 李瑛琦 , 方伯言 , 王颖鹏 . 基于机器学习的功能性近红外光谱信号识别早期帕金森病患者的研究[J]. 诊断学理论与实践, 2024 , 23(05) : 484 -493 . DOI: 10.16150/j.1671-2870.2024.05.004

Abstract

Objective This study aims to investigate the feasibility of diagnosing early-stage Parkinson’s disease (PD) patients by combining functional near-infrared spectroscopy (fNIRS) signals with machine learning algorithms. Methods Sixty PD patients as well as 60 healthy controls, diagnosed between December 2021 and August 2023 at Beijing Rehabilitation Hospital, Capital Medical University, were consecutively enrolled in this study. The ETG-4000 near-infrared brain imaging system with 22 channels (CH) was used to record changes in oxyhemoglobin and deoxyhemoglobin concentrations in the prefrontal cortex of the subjects. A general linear model was applied to calculate the activation degree (β value) for each channel. Four machine learning diagnostic models were developed: support vector machine (SVM), back-propagation (BP) neural network, random forest, and logistic regression models. The performance of the four diagnostic models was evaluated based on accuracy, sensitivity, specificity, and the area under the Receiver Operating Characteristic (ROC) curve. Additionally, SHapley Additive exPlanations (SHAP) analysis was applied to improve the interpretability of the optimal model. SHAP values for each channel were calculated, and the weighted average of the SHAP values from different channels was summarized. By combining this with the brain region distribution, the contribution of different brain regions to the model’s classification task was obtained. Results The accuracy of the four diagnostic models ranged from 81% to 90%, sensitivity from 69% to 89%, specificity from 93% to 100%, and the area under the ROC curve from 0.90 to 0.98. The SVM model outperformed the others, achieving an area under the ROC curve of 0.96, accuracy of 90%, sensitivity of 89%, and specificity of 93%. SHAP analysis revealed that the four channels contributing most to the SVM model were CH08, CH05, CH01, and CH13, with the right frontopolar cortex (FPC) region contributing the largest share (36.5% of the total). Conclusions The model based on fNIRS signals and the SVM algorithm shows great diagnostic advantages in diagnosing early-stage PD patients, with sensitivity (89%) and specificity (93%) exceeding those of most existing methods. Future research should focus on the fNIRS signal characteristics of the right frontopolar cortex and dorsolateral prefrontal cortex regions to further improve the performance of the diagnostic model.

参考文献

[1] MCDONALD C, GORDON G, HAND A, et al. 200 Years of Parkinson's disease: what have we learnt from James Parkinson?[J]. Age Ageing, 2018, 47(2):209-214.
[2] TITOVA N, MARTINEZ-MARTIN P, KATUNINA E, et al. Advanced Parkinson's or "complex phase" Parkinson's disease? Re-evaluation is needed[J]. J Neural Transm (Vienna), 2017, 124(12):1529-1537.
[3] MARINO B L B, DE SOUZA L R, SOUSA K P A, et al. Parkinson's disease: a review from pathophysiology to treatment[J]. Mini Rev Med Chem, 2020, 20(9):754-767.
[4] BARKHUIZEN M, ANDERSON D G, GROBLER A F. Advances in GBA-associated Parkinson's disease--Pathology, presentation and therapies[J]. Neurochem Int, 2016, 93:6-25.
[5] 刘浩宇, 朋文佳, 芈静, 等. 1990-2019年全球帕金森病疾病负担的APC分析[J]. 中华全科医学, 2024, 22(1):154-157.
  LIU H Y, PENG W J, MI J, et al. APC analysis of the global disease burden of Parkinson's disease from 1990 to 2019[J]. Chin J Gen Pract, 2024, 22(1):154-157.
[6] CHHOR V, KARACHI C, BONNET A M, et al. Anaesthesia and Parkinson's disease[J]. Ann Fr Anesth Reanim, 2011, 30(7-8): 559-68.
[7] KHAN H, NASEER N, YAZIDI A, et al. Analysis of human gait using hybrid EEG-fNIRS-based BCI system: A review[J]. Front Hum Neurosci, 2021, 14:613254.
[8] 李进, 艾芳, 刘媛, 等. 震颤分析在原发性震颤与帕金森病鉴别诊断中的准确性及价值[J]. 中国临床研究, 2021, 34(10):1354-1357.
  LI J, AI F, LIU Y, et al. Tremor analysis in differential diagnosis of essential tremor and Parkinson's disease[J]. Chin J Clin Res, 2021, 34(10):1354-1357.
[9] DOVONOU A, BOLDUC C, SOTO LINAN V, et al. Animal models of Parkinson's disease: bridging the gap between disease hallmarks and research questions[J]. Transl Neurodegener, 2023, 12(1):36.
[10] WILCOX T, BIONDI M. fNIRS in the developmental scie-nces[J]. Wiley Interdiscip Rev Cogn Sci, 2015, 6(3):263-283.
[11] COCKX H, OOSTENVELD R, TABOR M, et al. fNIRS is sensitive to leg activity in the primary motor cortex after systemic artifact correction[J]. Neuroimage, 2023, 269:119880.
[12] WEIBLEY H, DI FILIPPO M, LIU X, et al. fNIRS monitoring of infant prefrontal cortex during crawling and an executive functioning task[J]. Front Behav Neurosci, 2021, 15:675366.
[13] GUNASEKARA N, GAETA G, LEVY A, et al. fNIRS neuroimaging in olfactory research: A systematic literature review[J]. Front Behav Neurosci, 2022, 16:1040719.
[14] ZIMEO MORAIS G A, BALARDIN J B, SATO J R. fNIRS Optodes' Location Decider (fOLD): a toolbox for probe arrangement guided by brain regions-of-interest[J]. Sci Rep, 2018, 8(1):3341.
[15] VITORIO R, STUART S, ROCHESTER L, et al. fNIRS response during walking - Artefact or cortical activity? A systematic review[J]. Neurosci Biobehav Rev, 2017, 83:160-172.
[16] NASEER N, HONG K S. fNIRS-based brain-computer interfaces: a review[J]. Front Hum Neurosci, 2015, 9:3.
[17] LU J, WANG Y, SHU Z, et al. fNIRS-based brain state transition features to signify functional degeneration after Parkinson's disease[J]. J Neural Eng, 2022, 19(4):10.1088/1741-2552/ac861e.
[18] NIEUWHOF F, REELICK M F, MAIDAN I, et al. Measuring prefrontal cortical activity during dual task walki-ng in patients with Parkinson's disease: feasibility of using a new portable fNIRS device[J]. Pilot Feasibility Stud, 2016, 2:59.
[19] KAUSHIK C, MCRAE A D, DAVENPORT M, et al. New equivalences between interpolation and SVMs: Kernels and Structured Features[J]. ArXiv, 2023, abs/2305.02304.
[20] JANG R. Learning representations by forward-propagating errors[J]. ArXiv, 2023, abs/2308.09728.
[21] CHEN C, HUANG T S, HUANG J C, et al. Design of music style classification teaching system based on BP neural network[C]. International Conference on Information System, 2022.
[22] LIMAM H, ZOUHAIR A, OUESLATI W. A new hybrid multiclass approach based on KNN and SVM[J]. J Inf Knowl Manag, 2022(21): 2250061:1-2250061:16.
[23] DEDJA K, NAKANO F K, PLIAKOS K, et al. Explaining random forest prediction through diverse rulesets[J]. ArXiv, 2022, abs/2203.15511.
[24] ZAIDI A, LUHAYB ASM A L. Two statistical approaches to justify the use of the logistic function in binary logistic regression[J]. Math probl Eng, 2023, 5525675, 11pages.
[25] NAGASAWA T, SATO T, NAMBU I, et al. fNIRS-GANs: data augmentation using generative adversarial networks for classifying motor tasks from functional near-infrared spectroscopy[J]. J Neural Eng, 2020, 17(1):016068.
[26] HAN Y, HUANG J, YIN Y, et al. From brain to worksite: the role of fNIRS in cognitive studies and worker safety[J]. Front Public Health, 2023, 11:1256895.
[27] KUMAR V, SHIVAKUMAR V, CHHABRA H, et al. Functional near infra-red spectroscopy (fNIRS) in schizophrenia: A review[J]. Asian J Psychiatr, 2017, 27:18-31.
[28] BEHBOODI B, LIM S-H, LUNA M, et al. Artificial and convolutional neural networks for assessing functional connectivity in resting-state functional near infrared spectroscopy[J]. J Near Infrared Spectrosc, 2019, 27(3):191-205.
[29] FERRARI M, QUARESIMA V. A brief review on the history of human functional near-infrared spectroscopy (fNIRS) development and fields of application[J]. Neuroimage, 2012, 63(2):921-935.
[30] WANG K, TIAN J, ZHENG C, et al. Interpretable prediction of 3-year all-cause mortality in patients with heart failure caused by coronary heart disease based on machine learning and SHAP[J]. Comput Biol Med, 2021, 137:104813.
[31] ARREDONDO M M. Shining a light on cultural neuros-cience: Recommendations on the use of fNIRS to study how sociocultural contexts shape the brain[J]. Cultur Dive-rs Ethnic Minor Psychol, 2023, 29(1):106-117.
[32] LI Y, ZHANG X, MING D. Early-stage fusion of EEG and fNIRS improves classification of motor imagery[J]. Front Neurosci, 2023, 16:1062889.
[33] BLASI A, LLOYD-FOX S, KATUS L, et al. fNIRS for tracking brain development in the context of global health projects[J]. Photonics, 2019, 6(3):89.
[34] ALEXANDER R E, GAGE T W. Parkinson's disease: an update for dentists[J]. Gen Dent, 2000, 48(5):572-582.
[35] 杜静, 吴铁妤, 严孙宏, 等. 脑白质病变与帕金森病患者临床症状的相关性研究[J]. 重庆医科大学学报, 2024, 49(5):558-562.
  DU J, WU T Y, YAN S H, et al. Association between white matter lesion and clinical symptoms in patients with Parkinson's disease[J]. J Chongqing Med Univ, 2024, 49(5):558-562.
[36] VIRAMETEEKUL S, REVESZ T, JAUNMUKTANE Z, et al. Clinical diagnostic accuracy of Parkinson's disease: where do we stand?[J]. Mov Disord, 2023, 38(4):558-566.
[37] FILIPPI M, ELISABETTA S, PIRAMIDE N, et al. Functional MRI in idiopathic Parkinson's disease[J]. Int Rev Neurobiol, 2018, 141:439-467.
[38] CORDES D, ZHUANG X, KALEEM M, et al. Advances in functional magnetic resonance imaging data analysis methods using Empirical Mode Decomposition to investigate temporal changes in early Parkinson's disease[J]. Alzheimers Dement (N Y), 2018, 4:372-386.
[39] SYED NASSER N, IBRAHIM B, SHARIFAT H, et al. Incremental benefits of EEG informed fMRI in the study of disorders related to meso-corticolimbic dopamine pathway dysfunction: A systematic review of recent literature[J]. J Clin Neurosci, 2019, 65:87-99.
[40] LU J, WANG Y, SHU Z, et al. fNIRS-based brain state transition features to signify functional degeneration after Parkinson's disease[J]. J Neural Eng, 2022, 19(4):10.1088/1741-2552/ac861e.
[41] ABTAHI M, BAHRAM BORGHEAI S, JAFARI R, et al. Merging fNIRS-EEG brain monitoring and body motion capture to distinguish parkinson’s disease[J]. IEEE Trans Neural Syst Rehabil Eng, 2020, 28(6):1246-1253.
[42] HOLPER L, TEN BRINCKE R H, WOLF M, et al. fNIRS derived hemodynamic signals and electrodermal responses in a sequential risk-taking task[J]. Brain Res, 2014, 1557:141-154.
[43] CHEN Z, LI G, LIU J. Autonomic dysfunction in Parkinson's disease: Implications for pathophysiology, diagnosis, and treatment[J]. Neurobiol Dis, 2020, 134:104700.
[44] GHOUSE A, NARDELLI M, VALENZA G. fNIRS Complexity analysis for the assessment of motor imagery and Mental Arithmetic Tasks[J]. Entropy (Basel), 2020, 22(7):761.
[45] LIU W Y, TUNG T H, ZHANG C, et al. Systematic review for the prevention and management of falls and fear of falling in patients with Parkinson’s disease[J]. Brain Behav, 2022, 12(8):e2690.
[46] 张玉玲, 陈安安, 张海涵, 等. 中西医结合治疗帕金森病伴发睡眠障碍的临床研究进展[J]. 神经病学与神经康复学杂志, 2023, 19(4):127-134.
  ZHANG Y L, CHEN A A, ZHANG H H, et al. Progress of clinical research on the treatment of Parkinson’s di-sease accompanied by sleep disorder with integrative medicine[J]. J Neurol Neurorehabil, 2023, 19(4):127-134.
[47] 李杨夏, 张克忠. 帕金森病睡眠障碍研究进展[J]. 神经病学与神经康复学杂志, 2022, 18(1):22-28.
  LI Y X, ZHANG K Z. Advances in Parkinson's disease sleep disorder[J]. J Neurol Neurorehabil, 2022, 18(1):22-28.
文章导航

/