PFI:BIC: iSee - Intelligent Mobile Behavior Monitoring and Depression Analytics Service for College Counseling Decision Support

PFI：BIC：iSee - 用于大学咨询决策支持的智能移动行为监测和抑郁分析服务

• 批准号: 1632051
• 负责人: Mi Zhang
• 金额: $ 99.5万
• 依托单位: Michigan State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2016
• 资助国家: 美国
• 起止时间: 2016-09-01 至 2022-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1632051&HistoricalAwards=false
• 关键词: PFI; BIC; iSee; Intelligent; Mobile

Depression is the leading health issue on college campuses in the U.S. Today, college students are dealing with depression at some of the highest rates in decades. Unfortunately, university counseling centers (UCCs), which are the primary access points for students to receive mental health services, are facing significant challenges in meeting the increasing demands. Specifically, clinicians at UCCs still rely on patients' inaccurate and biased self-reported symptoms for depression assessment. In addition, UCCs provide mental health services only during working hours in clinical settings. The lack of service access when needed could leave patients floundering helplessly and lead to lifelong consequences. Furthermore, with tight budgets, clinicians at UCCs have not grown and some UCCs even downsized. As a consequence, more students did not receive timely treatment. This project focuses on designing and developing iSee, a smart device based behavior monitoring and analytics platform. iSee harnesses smartphones/wristbands to extend the reach of mental health care far beyond clinical settings and to deliver timely therapies when needed. Furthermore, the continuously tracked depression symptoms allow UCCs to be more accurately informed with the severity of each patient and thus reduces unnecessary visits so that clinician time can be better utilized. If successful, iSee has the potential to enhance mental health services in thousands of colleges and universities, benefiting millions of college students. Although focusing on depression of college students, the technology can be extended to other mental health conditions such as anxiety, bipolar disorder, dementia, and schizophrenia; adapted to patients beyond college students; and deployed at other settings such as public hospitals and private clinics.iSee consists of a smartphone/wristband sensing system running on the patient side to continuously and passively track patient's daily behaviors using onboard sensors; a behavior analytics engine using machine learning and causality analysis algorithms running on the cloud side to translate behavior sensor data into meaningful analysis results for identifying the patient's depression severity and revealing behavioral causes that lead to the mitigation or the deterioration of the patient's status; and a dashboard running on the clinician side to visualize behavior information as well as analysis results to help clinicians make clinical decisions and conduct treatment. The system would allow clinicians to access an objective, quantitative, and longitudinal record of patients' daily behavior to support evidence-based clinical assessment. This project involves a multi-disciplinary and cross-organizational team of researchers from Michigan State University (lead institution) and Northwestern University (Chicago, IL). The primary industry partner is Microsoft Research (Redmond, WA), which is a large business company in U.S. Michigan State University Counseling Center (East Lansing, MI), which will be the test bed for the integration and evaluation of the iSee smart service system. Finally, the broader context partners include the MSU Office of the Vice President for Student Affairs and Services and MSU Technologies (East Lansing, MI).This award is partially supported by funds from the Directorate for Computer and Information Science and Engineering (CISE), Division of Information and Intelligent Systems (IIS).

抑郁症是当今美国大学校园的主要健康问题，大学生在几十年来的抑郁症中的一些最高比率是抑郁症。不幸的是，大学咨询中心（UCC）是学生接受心理健康服务的主要访问点，在满足日益增长的需求方面面临着巨大的挑战。具体而言，UCC的临床医生仍然依靠患者的不准确和偏见的自我报告的症状来评估抑郁症。此外，UCC仅在临床环境的工作时间内提供心理健康服务。在需要时缺乏服务访问可能会使患者无助地挣扎并导致终生后果。此外，由于预算紧张，UCC的临床医生并没有增长，一些UCC甚至缩小了尺寸。结果，越来越多的学生没有得到及时的治疗。该项目着重于设计和开发ISEE，这是一个基于智能设备的行为监控和分析平台。 ISEE利用智能手机/腕带扩展了精神保健的范围，远远超出了临床环境，并在需要时及时提供疗法。此外，连续跟踪的抑郁症状使UCC可以通过每个患者的严重程度更准确地告知UCC，从而减少不必要的就诊，从而可以更好地利用临床医生时间。 如果成功的话，ISEE有可能在数千所大学和大学中增强心理健康服务，从而使数百万的大学生受益。尽管专注于大学生的抑郁症，但该技术可以扩展到其他心理健康状况，例如焦虑，躁郁症，痴呆和精神分裂症。适合大学生以外的患者；并在其他设置（例如公立医院和私人诊所）部署。ISEE由智能手机/腕带传感系统在患者方面运行，可连续，被动地使用船上传感器跟踪患者的日常行为；行为分析引擎使用机器学习和因果关系分析算法在云方面运行，以将行为传感器数据转化为有意义的分析结果，以识别患者的抑郁严重程度并揭示导致缓解或恶化患者状态的行为原因；以及在临床医生方面运行的仪表板，以可视化行为信息以及分析结果，以帮助临床医生做出临床决策并进行治疗。该系统将允许临床医生获得患者日常行为的客观，定量和纵向记录，以支持基于证据的临床评估。该项目涉及密歇根州立大学（LEAD机构）和西北大学（伊利诺伊州芝加哥）的研究人员的多学科和跨组织团队。主要行业合作伙伴是Microsoft Research（华盛顿州雷德蒙德），该研究是美国密歇根州立大学咨询中心（密歇根州东兰辛）的一家大型商业公司，该公司将是ISEE智能服务系统集成和评估的测试床。最后，更广泛的背景合作伙伴包括学生事务和服务副总裁和MSU Technologies（MI East Lansing）的MSU办公室。该奖项得到了计算机和信息科学与工程局（CISE）的资金（CISE），信息和情报系统（IIS）的部分支持。

项目成果

Deep AutoAugment

• DOI: 10.48550/arxiv.2203.06172
• 发表时间: 2022-03
• 期刊: ArXiv
• 作者: Yu Zheng;Z. Zhang;Shen Yan;Mi Zhang

FedMask: Joint Computation and Communication-Efficient Personalized Federated Learning via Heterogeneous Masking

FedMask：通过异构掩码进行联合计算和高效通信的个性化联合学习

• DOI: 10.1145/3485730.3485929
• 发表时间: 2021
• 期刊: ACM Conference on Embedded Networked Sensor Systems (SenSys'21
• 作者: Li, Ang;Sun, Jingwei;Zeng, Xiao;Zhang, Mi;Li, Hai;Chen, Yiran
• 通讯作者: Chen, Yiran

Mercury: Efficient On-Device Distributed DNN Training via Stochastic Importance Sampling

• DOI: 10.1145/3485730.3485930
• 发表时间: 2021-11
• 期刊: Proceedings of the 19th ACM Conference on Embedded Networked Sensor Systems
• 作者: Xiao Zeng;Ming Yan;Mi Zhang

Does Unsupervised Architecture Representation Learning Help Neural Architecture Search?

• 发表时间: 2020-06
• 期刊: ArXiv
• 作者: Shen Yan;Yu Zheng;Wei Ao;Xiao Zeng;Mi Zhang

PyramidFL: a fine-grained client selection framework for efficient federated learning

• DOI: 10.1145/3495243.3517017
• 发表时间: 2022-10
• 期刊: Proceedings of the 28th Annual International Conference on Mobile Computing And Networking
• 作者: Chenning Li;Xiao Zeng;Mi Zhang;Zhichao Cao