Bourla A, Ferreri F, Ogorzelec L, Guinchard C, Mouchabac S
UPMC, service de psychiatrie et de psychologie médicale des adultes, hôpital Saint-Antoine, AP-HP, 184, rue du Faubourg-Saint-Antoine, 75012 Paris, France.
UPMC, service de psychiatrie et de psychologie médicale des adultes, hôpital Saint-Antoine, AP-HP, 184, rue du Faubourg-Saint-Antoine, 75012 Paris, France.
Encephale. 2018 Apr;44(2):168-175. doi: 10.1016/j.encep.2017.07.007. Epub 2017 Oct 31.
The search for objective clinical signs is a constant practitioners' and researchers' concern in psychiatry. New technologies (embedded sensors, artificial intelligence) give an easier access to untapped information such as passive data (i.e. that do not require patient intervention). The concept of "digital phenotype" is emerging in psychiatry: a psychomotor alteration translated by accelerometer's modifications contrasting with the usual functioning of the subject, or the graphorrhea of patients presenting a manic episode which is replaced by an increase of SMS sent. Our main objective is to highlight the digital phenotype of mood disorders by means of a selective review of the literature.
We conducted a selective review of the literature by querying the PubMed database until February 2017 with the terms [Computer] [Computerized] [Machine] [Automatic] [Automated] [Heart rate variability] [HRV] [actigraphy] [actimetry] [digital] [motion] [temperature] [Mood] [Bipolar] [Depression] [Depressive]. Eight hundred and forty-nine articles were submitted for evaluation, 37 articles were included.
For unipolar disorders, smartphones can diagnose depression with excellent accuracy by combining GPS and call log data. Actigraphic measurements showing daytime alteration in basal function while ECG sensors assessing variation in heart rate variability (HRV) and body temperature appear to be useful tools to diagnose a depressive episode. For bipolar disorders, systems which combine several sensors are described: MONARCA, PRIORI, SIMBA and PSYCHE. All these systems combine passive and active data on smartphones. From a synthesis of these data, a digital phenotype of the disorders is proposed based on the accelerometer and the GPS, the ECG, the body temperature, the use of the smartphone and the voice. This digital phenotype thus brings into question certain clinical paradigms in which psychiatrists evolve.
All these systems can be used to computerize the clinical characteristics of the various mental states studied, sometimes with greater precision than a clinician could do. Most authors recommend the use of passive data rather than active data in the context of bipolar disorders because automatically generated data reduce biases and limit the feeling of intrusion that self-questionnaires may cause. The impact of these technologies questions the psychiatrist's professional culture, defined as a specific language and a set of common values. We address issues related to these changes. Impact on psychiatrists could be important because their unity seems to be questioned due to technologies that profoundly modify the collect and process of clinical data.
寻找客观的临床体征一直是精神病学领域从业者和研究人员所关注的问题。新技术（嵌入式传感器、人工智能）使人们更容易获取未开发的信息，如被动数据（即不需要患者干预的数据）。“数字表型”的概念正在精神病学领域兴起：通过加速度计的变化反映出的精神运动改变，与受试者的正常功能形成对比，或者躁狂发作患者的书写狂被短信发送量的增加所取代。我们的主要目标是通过对文献的选择性回顾来突出情绪障碍的数字表型。
我们通过查询PubMed数据库对文献进行选择性回顾，截至2017年2月，使用的检索词为[计算机][计算机化][机器][自动][自动化][心率变异性][HRV][活动记录仪][活动测量][数字][运动][温度][情绪][双相情感障碍][抑郁症][抑郁]。共提交849篇文章进行评估，纳入37篇文章。
对于单相情感障碍，智能手机通过结合GPS和通话记录数据能够以极高的准确率诊断抑郁症。活动记录仪测量显示基础功能在白天出现改变，而心电图传感器评估心率变异性（HRV）和体温的变化似乎是诊断抑郁发作的有用工具。对于双相情感障碍，描述了几种结合多种传感器的系统：MONARCA、PRIORI、SIMBA和PSYCHE。所有这些系统都结合了智能手机上的被动和主动数据。综合这些数据，基于加速度计、GPS、心电图、体温、智能手机的使用情况和语音，提出了这些障碍的数字表型。因此，这种数字表型对精神病医生所处的某些临床范式提出了质疑。
所有这些系统都可用于将所研究的各种精神状态的临床特征计算机化，有时其精度比临床医生更高。大多数作者建议在双相情感障碍的背景下使用被动数据而非主动数据，因为自动生成的数据减少了偏差，并限制了自我问卷可能引起的侵扰感。这些技术的影响对精神病医生的专业文化提出了质疑，专业文化被定义为一种特定的语言和一套共同的价值观。我们讨论了与这些变化相关的问题。对精神病医生的影响可能很大，因为由于深刻改变临床数据收集和处理方式的技术，他们的统一性似乎受到了质疑。
