Dashboards for Clinician Monitoring of Patients Through a Mobile Sensing Platform

描述(由申请人提供)：精神健康障碍在我国和世界各地造成了非常高的个人、社会和经济代价，构成了一个重大的公共卫生挑战。Cogito已经开发和部署了能够倾听自然对话的技术，并对抑郁症等各种结果产生70%-90%的预测准确率(Azarbayejani，Feast，Zhang，&Massey，2011；Azarbayejani，Feast，&Pentland，2010；Chu，2009)。这一软件工具最近已得到扩充，并扩展到移动电话。Cogito的新软件套件Mobile Sensing Platform客观地测量行为指标(包括运动、交流、位置和语音)，使患者能够更轻松地监控一段时间内的抑郁、焦虑和创伤后应激障碍的症状。此第一阶段应用程序的主要重点是开发一种新工具-仪表板-将移动平台扩展到研究人员和临床医生，从而提供一种易于可视化和分析这些复杂的患者行为数据模式的方法。将Cogito的移动传感平台扩展到包括新的临床医生仪表盘，将极大地提高通过我们的移动传感软件收集的丰富数据流的可用性和价值。研究人员将能够准实时地监测和量化不同患者在个人和群体层面上的治疗效果。这项第一阶段提案的具体目标是 设计和开发仪表板可视化工具，并确定通过仪表板呈现给临床医生和研究人员的数据是有效的、有价值的和有用的。临床研究人员目前面临着通过自我报告和历史行为在就诊之间推断患者行为和治疗依从性的艰巨任务。通过智能手机应用程序使用被动收集的行为信号，可以根据实际行为而不是报告的行为构建症状评估模型。准实时监测、分析和可视化变化为临床医生提供了一个了解患者行为的新窗口。Cogito移动传感平台将提供这样的服务，通过测量治疗效果、调整药物和快速评估行为变化来改善患者护理。Cogito目前正在使用我们的平台参与临床试验，以监测美国军事人员的精神疾病。我们将使用此数据集来说明 典型的症状群。这些纵向症状集将构成构建仪表板可视化工具的基础。与临床研究人员组成的焦点小组将使数据呈现和可用性与执业医疗专业人员的需求保持一致。

DESCRIPTION (provided by applicant): Mental health disorders exact very high personal, social and economic costs in our country and around the world, presenting a significant public health challenge. Cogito has developed and deployed technology that listens to natural conversation and produces predictive accuracies of 70-90% with respect to various outcomes such as depression (Azarbayejani, Feast, Zhang, & Massey, 2011; Azarbayejani, Feast, & Pentland, 2010; Chu, 2009). This software tool has been recently augmented and extended to mobile telephones. Cogito's new software suite, the Mobile Sensing Platform, objectively measures behavioral indicators (including motion, communication, location, and voice) and allows patients to more easily monitor their symptoms over time for depression, anxiety, and PTSD. The primary focus of this Phase I application is to develop a new tool - a dashboard - that will extend the mobile platform to researchers and clinicians, thus providing a means of readily visualizing and analyzing these complex patient behavioral data patterns. Extending Cogito's Mobile Sensing Platform to include a new clinician dashboard will enhance greatly the usability and value of the rich data streams collected via our mobile sensing software. Researchers will be able to monitor and quantify the effect of treatments on different patients in quasi-real time, at both the individual and group level. Specific aims of this Phase I proposal are to design and develop a dashboard visualization tool and to establish that the data presented to clinicians and researchers via the dashboard are valid, valuable and useful. Clinical researchers are currently faced with the difficult task of inferring patient behavior and treatment compliance between visits through self-report and historical behavior. Using passively gathered behavioral signals through a smart phone application allows for symptom assessment models to be built from actual behavior instead of reported behavior. Monitoring, analyzing, and visualizing changes in quasi-real time allow clinicians a new window into understanding patient behavior. The Cogito Mobile Sensing Platform would provide such a service, allowing for improvements in patient care through measurement of treatment efficacy, adjustments of medication, and rapid assessments of behavioral change. Cogito is currently involved in clinical trials using our platform to monitor mental illness in US military personnel. We will use this dataset to illustrate prototypical symptom clusters. These longitudinal symptom sets will form the basis for building the dashboard visualization tool. Focus groups with clinical researchers will align the data presentation and usability with the needs of practicing medical professionals.