人工智能能否让人快乐？基于人工智能的聊天机器人对帕金森病患者微笑和言语的影响。

Can AI make people happy? The effect of AI-based chatbot on smile and speech in Parkinson's disease.

机构信息

Department of Neurology, Juntendo University Graduate School of Medicine, Tokyo, Japan; Department of Neurodegenerative and Demented Disorders, Juntendo University Graduate School of Medicine, Tokyo, Japan; Department of Home Medical Care System Based on Information and Communication Technology, Juntendo University Graduate School of Medicine, Tokyo, Japan; Department of Drug Development for Parkinson's Disease, Juntendo University Graduate School of Medicine, Tokyo, Japan; Department of Patient Reported Outcome Based Integrated Data Analysis in Neurological Disorders, Juntendo University Graduate School of Medicine, Tokyo, Japan; Department of Research and Therapeutics for Movement Disorders, Juntendo University Graduate School of Medicine, Tokyo, Japan.

出版信息

Parkinsonism Relat Disord. 2022 Jun;99:43-46. doi: 10.1016/j.parkreldis.2022.04.018. Epub 2022 May 5.

DOI:10.1016/j.parkreldis.2022.04.018

PMID:35596975

Abstract

INTRODUCTION

Approaches for objectively measuring facial expressions and speech may enhance clinical and research evaluation in telemedicine, which is widely employed for Parkinson's disease (PD). This study aimed to assess the feasibility and efficacy of using an artificial intelligence-based chatbot to improve smile and speech in PD. Further, we explored the potential predictive value of objective face and speech parameters for motor symptoms, cognition, and mood.

METHODS

In this open-label randomized study, we collected a series of face and conversational speech samples from 20 participants with PD in weekly teleconsultation sessions for 5 months. We investigated the effect of daily chatbot conversations on smile and speech features, then we investigated whether smile and speech features could predict motor, cognitive, and mood status.

RESULTS

A repeated-measures analysis of variance revealed that the chatbot conversations had a significant interaction effect on the mean and standard deviation of the smile index during smile sections (both P = .02), maximum duration of the initial rise of the smile index (P = .04), and frequency of filler words (P = .04), but no significant interaction effects were observed for clinical measurements including motor, cognition, depression, and quality of life. Explorative analysis using statistical and machine-learning models revealed that the smile indices and several speech features were associated with motor symptoms, cognition, and mood in PD.

CONCLUSION

An artificial intelligence-based chatbot may positively affect smile and speech in PD. Smile and speech features may capture the motor, cognitive, and mental status of patients with PD.

摘要

简介

客观测量面部表情和言语的方法可增强远程医疗中的临床和研究评估，而远程医疗在帕金森病（PD）中广泛应用。本研究旨在评估基于人工智能的聊天机器人改善 PD 患者微笑和言语的可行性和效果。此外，我们还探讨了客观面部和言语参数对运动症状、认知和情绪的潜在预测价值。

方法

在这项开放性、随机研究中，我们在 5 个月的每周远程会诊期间，从 20 名 PD 患者中收集了一系列面部和会话语音样本。我们研究了日常聊天机器人对话对微笑和语音特征的影响，然后研究了微笑和语音特征是否可以预测运动、认知和情绪状态。

结果

微笑部分的微笑指数平均值和标准差（均 P=0.02）、微笑指数初始上升的最大持续时间（P=0.04）和填充词频率（P=0.04）的重复测量方差分析显示，聊天机器人对话具有显著的交互效应，但在包括运动、认知、抑郁和生活质量在内的临床测量中未观察到显著的交互效应。使用统计和机器学习模型进行的探索性分析表明，微笑指数和几个语音特征与 PD 患者的运动症状、认知和情绪有关。