-
帕金森病(Parkinson’s Disease, PD)是继阿尔茨海默病(Alzheimer’s Disease, AD)之后的第二大常见的神经退行性疾病,平均发病年龄为60岁左右。PD影响着全世界700~1000万人[1],中国每年新增超过10万患者,近年来还呈现年轻化的发病趋势。到2030年,中国PD患者可能将占世界PD患者总人数的一半[2]。PD与位于人脑丘脑区的黑质损坏相关,黑质神经元不可逆的损伤将导致患者的非自主运动(异常运动)。当出现PD临床症状(如震颤、肌肉僵硬、运动迟缓和平衡失调等)时,患者已处于疾病的晚期,错过了最佳治疗时机,因此早期诊断该疾病至关重要。
目前对PD的诊断主要基于临床评分量表(如帕金森病认知功能评定量表PDCRS、蒙特利尔认知评估量表MoCA、Mattis痴呆评定量表Mattis DRS、国际运动障碍学会统一帕金森病评定量表MDS-UPDRS等)[3],缺乏客观定量的评估,没有足够灵敏度捕捉患者的细微变化,制约了PD的及时诊断和临床研究。同时,诊断PD还经常受到患者医疗数据特征的影响,包括各种指标的存在、患者数据记录的不平衡等情况,人工智能机器学习改善了PD诊断问题的现状。
人工智能是研究和开发用于模拟、延伸和扩展人类智能的理论、方法、技术及应用系统的一门技术科学,人类大脑的高级功能(包括学习、记忆、思维、推理、意识认知等)的生理基础和基本工作机制,以及人工智能模拟人类大脑高级功能的相应内容[4]。其中机器学习模拟大脑的学习和记忆的功能,是实现人工智能、使计算机具有学习能力的方法之一。机器学习通过分析和处理大量数据,对世界中发生的事件做出判断和预测的一项方法技术,是识别数据模式和关系的工具。机器学习作为一种自动化分析模型构建的数据分析方法,其常用的算法有:K-近邻算法、K-均值算法、随机森林、朴素贝叶斯算法、支持向量机、决策树、逻辑回归等,已被应用在人工智能智慧医学诊断中。
机器学习系统中学习环节的一般过程如下:在进行学习过程之前,首先需要确定学习模型,即具体采用何种方法进行机器学习;然后收集和准备训练数据,训练数据就是对事物的观察和历史经验;在获得原始数据之后对数据进行清洗和特征提取;最后运行具体的学习算法,获取相关知识[4]。
以PD患者为研究主体,联合神经影像、电生理、基因、神经认知和行为学的综合性研究是神经调控技术的重要研究方向[5],也是机器学习的重要应用方向。本文对机器学习基于行为(语音、步态、书写)、电生理、影像组学和基因组学等数据,在PD诊断中的应用进行综述,旨在对未来人工智能智慧诊断提供参考与借鉴。
Application of Machine Learning to Parkinson’s Disease Diagnosis
-
摘要: 机器学习是医学人工智能的研究热点和重点之一。针对神经退行性帕金森病(Parkinson’s Disease, PD)的早期诊断,现有的临床评分量表具有一定的主观性和局限性。该文报告了基于行为(语音、步态、书写)、电生理(脑电)、影像组学(核磁共振成像、单光子发射断层成像、正光子发射断层成像)和基因组学等数据,机器学习应用于PD诊断的研究进展,发现其较传统方法更为精准,以期为未来人工智能智慧诊断的研究与应用提供参考与借鉴。Abstract: Machine learning is one of the research hotspots and focuses of medical artificial intelligence. For the early diagnosis of neurodegenerative Parkinson’s Disease (PD), the existing clinical rating scales have certain subjectivity and limitations. This paper reports the research progress of machine learning in the diagnosis of PD based on behavioral (speech, gait, and writing), electrophysiology (Electroencephalogram, EEG), radiomics (magnetic resonance imaging, single-photon emission tomography, and positive photon emission tomography), and genomics data. The report finds that the application of machine learning is more accurate than the traditional method in the diagnosis of PD, which provides reference for the research and application of artificial intelligence intelligent diagnosis in the future.
-
Key words:
- Parkinson’s disease /
- machine learning /
- artificial intelligence /
- diagnosis
-
[1] MAITIN A M, ROMERO M, JUAN P, et al. Survey of machine learning techniques in the analysis of EEG signals for Parkinson’s disease: A systematic review[J]. Applied Sciences-Basel, 2022, 12(14): 39-55. [2] TONELLI L C, WOHR M, SCHWARTING R, et al. Paradoxical kinesia induced by appetitive 50-kHz ultrasonic vocalizations in rats depends on glutamatergic mechanisms in the inferior colliculus[J]. Neuropharmacology, 2018, 135(2): 172-179. [3] ISELLA V, MAPELLI C, MORIELLI N, et al. Diagnosis of possible mild cognitive impairment in Parkinson’s disease: Validity of the SCOPA-Cog[J]. Parkinsonism & Related Disorders, 2013, 19(12): 1160-1163. [4] 唐子惠. 医学人工智能导论[M]. 上海: 上海科学技术出版社, 2020. TANG Z H. Introduction to medical artificial intelligence[M]. Shanghai: Shanghai Science and Technology Press, 2020. [5] 李文娟, 张巧俊. 非侵入性神经调控技术在帕金森病治疗中的应用现状及展望[J]. 中华物理医学与康复杂志, 2022, 44(1): 83-87. LI W J, ZHANG Q J. Application status and prospect of non-invasive neuromodulation technology in the treatment of Parkinson’s disease[J]. Chinese Journal of Physical Medicine and Rehabilitation, 2022, 44(1): 83-87. [6] KARAN B, SAHU S S, MAHTO K. Parkinson disease prediction using intrinsic mode function based features from speech signal[J]. Biocybernetics and Biomedical Engineering, 2020, 40(1): 249-264. doi: 10.1016/j.bbe.2019.05.005 [7] SECHIDIS K, FUSAROLI R, OROZCO-ARROYAVE J R, et al. A machine learning perspective on the emotional content of Parkinsonian speech[J]. Artificial Intelligence in Medicine, 2021, 115(3): 102-111. [8] SUPPA A, COSTANTINI G, ASCI F, et al. Voice in Parkinson’s disease: A machine learning study[J]. Frontiers in Neurology, 2022, 13(2): 831-850. [9] FERRUCC R, MAMELI F, RUGGIERO F, et al. Alternate fluency in Parkinson’s disease: A machine learning analysis[J]. Heterocycles, 2022, 17(3): 265-280. [10] AVULU E, ELEN A. Evaluation of train and test performance of machine learning algorithms and Parkinson diagnosis with statistical measurements[J]. Medical & Biological Engineering & Computing, 2020, 58(11): 2775-2788. [11] SEHGAL S, AGARWAL M, GUPTA D, et al. Optimized grass hopper algorithm for diagnosis of Parkinson’s disease[J]. SN Applied Sciences, 2020, 2(6): 1-18. [12] ELEN A. A Comparison of classification methods for diagnosis of Parkinson’s[J]. International Journal of Intelligent Systems and Applications in Engineering, 2020, 8(4): 164-170. doi: 10.18201/ijisae.2020466309 [13] 闫红娇, 方伯言. 帕金森病患者步态适应性的研究进展[J]. 中华物理医学与康复杂志, 2022, 44(4): 373-376. YAN H J, FANG B Y. Research progress on gait adaptability in patients with Parkinson’s disease[J]. Chinese Journal of Physical Medicine and Rehabilitation, 2022, 44(4): 373-376. [14] 许之珩, 张梦翰, 王坚. 语音声学分析在帕金森病中的诊断价值[J]. 四川大学学报(医学版), 2022, 53(4): 726-731. XU Z H, ZHANG M H, WANG J. The diagnostic value of acoustic speech analysis in Parkinson’s disease[J]. Journal of Sichuan University (Medical Science), 2022, 53(4): 726-731. [15] KUBOTA K J, CHEN J A, LITTLE M A. Machine learning for large-scale wearable sensor data in Parkinson’s disease: Concepts, promises, pitfalls, and futures[J]. Movement Disorders, 2016, 31(9): 1314-1326. doi: 10.1002/mds.26693 [16] MUNOZ-OSPINA B, ALVAREZ-GARCIA D, CLAVIJO-MORAN H J C, et al. Machine learning classifiers to evaluate data from gait analysis with depth cameras in patients with Parkinson’s disease[J]. Frontiers in Human Neuroscience, 2022, 16(1): 826-840. [17] TRABASSI D, SERRAO M, VARRECCHIA T, et al. Machine learning approach to support the detection of Parkinson’s disease in IMU-based gait analysis[J]. Sensors, 2022, 22(10): 32-45. [18] LIN S N, GAO C, LI H X, et al. Wearable sensor-based gait analysis to discriminate early Parkinson’s disease from essential tremor[J]. Journal of Neurology, 2023, 270(4): 2283-2301. [19] MA Y W, CHEN J L, CHEN Y J, et al. Explainable deep learning architecture for early diagnosis of Parkinson’s disease[J]. Soft Computing, 2023, 27(5): 2729-2738. doi: 10.1007/s00500-021-06170-w [20] 孙玉波, 刘培培, 杨宇辰, 等. 一种基于二维视频的运动障碍量化评估方法与临床应用研究[J]. 生物医学工程学杂志, 2023, 40(3): 499-507. SUN Y B, LIU P P, YANG Y C, et al. A quantitative assessment method of movement disorders based on two-dimensional video and its clinical application[J]. Journal of Biomedical Engineering, 2023, 40(3): 499-507. [21] KHERA P, KUMAR N. Novel machine learning-based hybrid strategy for severity assessment of Parkinson’s disorders[J]. Medical & Biological Engineering & Computing, 2022, 60(3): 811-828. [22] BALAJI E, BRINDHA D, BALAKRISHNAN R. Supervised machine learning based gait classification system for early detection and stage classification of Parkinson’s disease[J]. Applied Soft Computing, 2020, 94(1): 1-15. [23] 李文丹, 陈绣君, 李蒙燕, 等. 基于手机视频的帕金森病患者冻结步态的自动识别[J]. 中华神经医学杂志, 2022, 21(4): 348-353. LI W D, CHEN X J, LI M Y, et al. Automatic recognition of frozen gait in patients with Parkinson’s disease based on mobile phone video[J]. Chinese Journal of Neuromedicine, 2022, 21(4): 348-353. [24] KOVALENKO E, TALITCKII A, ANIKINA A, et al. Distinguishing between Parkinson’s disease and essential tremor through video analytics using machine learning: A pilot study[J]. IEEE Sensors Journal, 2020, 21(10): 11916-11925. [25] VALLA E, NOMM S, MEDIJAINEN K, et al. Tremor-related feature engineering for machine learning based Parkinson’s disease diagnostics[J]. Biomemedical Signal Processing and Control, 2022, 75(1): 1-9. [26] GUPTA U, BANSAL H, JOSHI D. An improved sex specific and age dependent classification model for Parkinson’s diagnosis using handwriting measurement[J]. Computer Methods and Programs in Biomedicine, 2020, 189(3): 105-118. [27] VJGA B, MK C, MFCA D, et al. Machine learning for automated EEG-based biomarkers of cognitive impairment during deep brain stimulation screening in patients with Parkinson’s disease[J]. Clinical Neurophysiology, 2021, 132(5): 1041-1048. doi: 10.1016/j.clinph.2021.01.021 [28] CHAWLA P, RANA S B, KAUR H, et al. A decision support system for automated diagnosis of Parkinson’s disease from EEG using FAWT and entropy features[J]. Biomedical Signal Processing and Control, 2023, 79(1): 112-135. [29] YANG C Y, HUANG Y Z. Parkinson’s disease classification using machine learning approaches and resting-state EEG[J]. Journal of Medical and Biological Engineering, 2022, 42(2): 263-270. doi: 10.1007/s40846-022-00695-7 [30] LU J, SOROOSHYARI S K. Machine learning identifies a rat model of Parkinson’s disease via sleep-wake electroencephalogram[J]. Neuroscience, 2023, 510(2): 1-8. [31] 亚洋, 王二磊, 伋立荣, 等. 基于结构MRI机器学习模型诊断帕金森病的价值[J]. 中华放射学杂志, 2023, 57(4): 370-377. YA Y, WANG E L, LING L R, et al. The value of structural MRI machine learning model in diagnosing Parkinson’s disease[J]. Chinese Journal of Radiology, 2023, 57(4): 370-377. [32] 杨一风, 胡颖, 聂生东. 基于磁共振图像的帕金森病计算机诊断研究进展[J]. 中国生物医学工程学报, 2020, 39(5): 603-610. YANG Y F, HU Y, NIE S D. Research progress on computer diagnosis of Parkinson’s disease based on magnetic resonance images[J]. Chinese Journal of Biomedical Engineering, 2020, 39(5): 603-610. [33] HUANG G, LIN C, CAI Y, et al. Multiclass machine learning classification of functional brain images for Parkinson’s disease stage prediction[J]. Statistical Analysis and Data Mining, 2020, 13(5): 508-523. doi: 10.1002/sam.11480 [34] SALMANPOUR M R, SHAMSAEI M, SABERI A, et al. Robust identification of Parkinson’s disease subtypes using radiomics and hybrid machine learning[J]. Computers in Biology and Medicine, 2021, 129(2): 104-142. [35] ABOS A, BAGGIO H C, SEGURA B, et al. Discriminating cognitive status in Parkinson’s disease through functional connectomics and machine learning[J]. Scientific Reports, 2017, 7(1): 45-63. doi: 10.1038/s41598-017-00079-x [36] YA Y, JI L, JIA Y Y, et al. Machine learning models for diagnosis of Parkinson’s disease using multiple structural magnetic resonance imaging features[J]. Frontiers in Aging Neuroscience, 2022, 14(3): 808-820. [37] SOLANA-LAVALLE G, ROSAS-ROMERO R. Classification of PPMI MRI scans with voxel-based morphometry and machine learning to assist in the diagnosis of Parkinson’s disease[J]. Computer Methods and Programs in Biomedicine, 2021, 198(2): 105-131. [38] ZHANG J. Mining imaging and clinical data with machine learning approaches for the diagnosis and early detection of Parkinson’s disease[J]. NPJ Parkinsons Disease, 2022, 8(1): 1-15. doi: 10.1038/s41531-021-00272-w [39] TERMINE A, FABRIZIO C, STRAFELLA C, et al. A hybrid machine learning and network analysis approach reveals two Parkinson’s disease subtypes from 115 RNA-seq post-mortem brain samples[J]. International Journal of Molecular Sciences, 2022, 23(5): 109-123. [40] JUCAITE A, CSELENY Z, KREISL W C, et al. Glia imaging differentiates multiple system atrophy from Parkinson’s disease: A positron emission tomography study with [C-11]PBR28 and machine learning analysis[J]. Movement Disorders, 2022, 37(1): 119-129. doi: 10.1002/mds.28814 [41] XU N, ZHOU Y X, PATEL A, et al. Parkinson’s disease diagnosis beyond clinical features: A bio-marker using topological machine learning of resting-state functional magnetic resonance imaging[J]. Neuroscience, 2023, 509(2): 43-50. [42] ARORA P, MISHRA A, MALHI A. Machine learning ensemble for the Parkinson’s disease using protein sequences[J]. Multimedia Tools and Applicaitons, 2022, 81(22): 32215-32242. doi: 10.1007/s11042-022-12960-7 [43] PANTALEO E, MONACO A, AMOROSO N, et al. A machine learning approach to Parkinson’s disease blood transcriptomics[J]. Genes, 2022, 13(5): 164-178.
计量
- 文章访问数: 5066
- HTML全文浏览量: 2253
- PDF下载量: 28
- 被引次数: 0