Revolutionizing Healthcare Decision-Making via Social Networks and Informat
Social networks and information systems offer new opportunities for understanding and improving the management of complex health condition. Patients, providers and researchers differentially rely upon these systems to meet their goals. In this workshop investigators will describe their research in the areas of decision-support, chronic care management and social computing and the variety of tools they utilize. These include, but are not limited to, mobile phone interventions, crowdsourcing, machine learning prediction models and web data analyses. Researchers will also discuss the value of studying and using these interactive systems for guiding practice and policy. The research described will deal with conditions that run the gamut between physical and mental health such as cerebral palsy, epilepsy, asthma and autism.
- Talks will be in English
- The number of participants for the workshop is limited
Talk Titles, Abstracts, and Bios
Developing Interventions for Chronic Care Management via mHealth Systems and Crowd Sourcing
Computing holds the promise of alleviating some of the negative impact of both chronic disease and developmental disorders by scaling human effort over time and space. In this talk I will discuss two technology interventions for personalized care. One was designed to impact two major gaps in chronic care management (a) the lack of awareness of the patient’s symptoms between scheduled visits and (b) the scarcity of tools to promote communication and support decision-making between patients and medical care providers. This intervention was a text-message based system that was designed for pediatric asthma patients. The other was a web-based system that used crowdsourcing to help build social stories, a common therapy used for individuals with autism. This tool provided a basic set of steps to a routine (e.g., going to a fast food restaurant) and allowed the user (therapist or parent) to individualize it for the needs of a specific child with autism. I will also discuss how the underlying design principles that power these two systems can also be employed for other chronic diseases and developmental disorders.
Dr. Arriaga is a developmental psychologist in the School of Interactive Computing. Her emphasis is on using psychological concepts, theories and methods to address fundamental topics of human computer interaction. She is currently a Senior Research Scientist in Interactive Computing at Georgia Tech. She received her Ph.D. in Developmental Psychology from Harvard University.
The MobiGuide Project: Guiding Patients Any Time, Everywhere in a Personalized Way
During the past 15 years, clinical decision-support systems (DSS) that are based on evidence-based clinical guidelines have been used to deliver patient-specific recommendations to care providers during patient encounters. In the European project "MobiGuide: Guiding Patients Any Time Everywhere" (http://www.mobiguide-project.eu/) we are addressing several novelties: (a) delivery of decision-support to patients (via Smartphones) and not only to care providers, (b) personalization of guidelines to the patients' preferences and their personal context as well as the technological state of the MobiGuide system, (c) distribution of decision-support between a backend DSS Server and a mobile DSS operating on a Smartphone, (d) semantically-integrated Personal Health Record (PHR) that integrates data from hospital EMRs, wearable monitoring devices, DSS events, and temporal data abstractions, and (e) intelligent data and process mining algorithms that learn from past care process execution and suggests ways in which clinical guidelines could be improved.
In this talk I will explain the objectives of MobiGuide and how we customize computer-interpretable guidelines to personal context and personalize decision support to individual patients.
Mor Peleg is Assoc. Prof at the Dept. of Information Systems, University of Haifa, Israel, since 2003, and has been Department Head in 2009-2012. Her BSc and MSc in Biology and PhD in Information Systems are from the Technion, Israel. She spent 6 years at Stanford Medical Research during her post-doctoral studies and Sabbatical. She was awarded the New Investigator Award by the American Medical Informatics Association (AMIA) and was elected Fellow of the American College of Medical Informatics in 2013. Her research concerns knowledge representation, decision support systems, and process-aware information systems in healthcare, and appeared in journals such as JAMIA, International Journal of Medical Informatics, Journal of Biomedical Informatics, IEEE Transactions on Software Eng, TKDE, Bioinformatics. She is the coordinator of the FP7-ICT large-scale project MobiGuide (http://www.mobiguide-project.eu/). She has edited a number of special issues related to process support in healthcare and artificial intelligence in medicine. Mor has served in program committees of numerous conferences, including, among others, AI in Medicine (were she chaired the scientific PC in 2011), Medinfo, ER. She has been co-chair of the BPM ProHealth Workshop six times and an organizing committee member of Knowledge Representation for Healthcare Workshop five times. She is a member of the editorial board of Journal of BioMedical Informatics and Methods of Information in Medicine. http://mis.hevra.haifa.ac.il/~morpeleg/
OSCAR: A Clinical Decision Support System for Prescription of Assistive Technology to People with Disabilities
Alexandra Danial-Saad , Tsvika Kuflik, Patrice L. (Tamar) Weiss, Naomi Schreuer
This presentation will focus on a novel, knowledge-driven approach for pointing device prescription that aims to enhance the prescription process via OSCAR, a Clinical Decision Support System. OSCAR was constructed in a three-stage process: developing a pointing-device ontology; formulating IF-THEN matching rules, and constructing a web-based user interface. Evaluating OSCAR's usability through the System Usability Scale reflected a high satisfaction rate with a mean score = 80.9 (SD = 10.6). OSCAR's effectiveness was evaluated through the prescription results which showed no significant difference between experts using the traditional approach and novices using OSCAR. These results indicate that OSCAR is able to support novice clinicians in the AT prescription process to a level that is comparable to experts. The main contribution of this research is in the development and evaluation of a CDSS which bridges the gap between existing general models for pointing device prescription and a process that can support expert and novice clinicians. Although the system is currently designed for adaptation of a specific device (physically controllable pointing device) it can be expanded to include other AT devices. The results will also serve as an example to demonstrate the utility of using a CDSS in other clinical areas.
Prof. Patrice L. (Tamar) Weiss is Assoc. Prof at the Dept. of Occupational Therapy. She established the Laboratory for Innovations in Rehabilitation Technology (LIRT) at the University of Haifa in 2001. LIRT's focus is on the development and evaluation of novel virtual environments, computer interfaces, co-located technologies and remote care interventions to explore their effect on body functions (e.g. motor and cognitive abilities), activities (e.g. meal preparation) and participation in community life. Rehabilitation populations of interest include spinal cord injury, stroke, cerebral palsy, developmental coordination disorder, autism and head trauma.
Danial-Saad Alexandra is an occupational therapist in the field of Assistive Technology. She is post submitting her PhD degree in Assistive Technology, and a lecturer at the Occupational Therapy Department in Haifa University, Israel. In addition, she is also the General Practicum coordinator at the Academic Arab College for Education in Haifa, Israel.
Leveraging Advanced Analytics to Transform Data into Healthcare Solutions
Vast amount of healthcare related information is gathered every day by healthcare systems through various stakeholders such as payers, providers, pharmaceutical companies and more. While this information is used today for care management it can be leveraged much more with the help of advanced analytics. I will present our approach to addressing this real world evidence data through the use of machine learning and statistical analytics, including specific examples where we transformed the richnessof data into solutions such as decision support for HIV patients, predictive analytics and patient similarity for epilepsy and more.
Ya'ara Goldschmidt, PhD. is a Research Staff Member at IBM Research, Haifa. Dr. Goldschmidt received her B.Sc. degree in bioinformatics from Ben Gurion University in 2001, and her M.S. and Ph.D. degrees in bioinformatics from the Weizmann Institute of Science in 2002 and 2007, respectively. She subsequently joined IBM at the Haifa Research Lab and from 2010 served as manager of the Machine Learning Technologies group working on projects involving text mining, anomaly detection and fraud detection. Since 2012 she serves as manager of the Machine Learning for Healthcare and Life Sciences group, a team focused on research projects in machine learning for clinical genomic and healthcare applications.
Some Insights as to Why People Seek Medical Information on the Internet, and How this Can Help in Learning about their Predicament
Surveys show that the vast majority of Internet users turn to the Web when they have a medical concern. However, since medical information is a sensitive topic, users find some Internet forums are better suited for medical inquiries than others. In this talk I’ll compare several such forums and show that identifiability is a major deciding factor for the types of medical concerns posted on a forum, and that the more identifiable the forum, the less likely it is that medical concerns will be discussed on it. I’ll argue that these insights should be a guiding principle when trying to learn about medical conditions from the Internet, and provide examples of how non-identifiable media can help in guiding the decision process of clinicians through deeper understanding of the concerns of patients and caregivers.
Elad Yom-Tov is a Senior Researcher at Microsoft Research Israel. Before joining Microsoft he was with Yahoo Research, IBM Research, and Rafael. Dr. Yom-Tov studied at Tel-Aviv University and the Technion, Israel. He has published two books, over 80 papers (of which 3 were awarded prizes), and filed more than 30 patents (16 of which have been granted so far). His primary research interests are in applying large-scale Machine Learning and Information Retrieval methods to medicine. The results of his work have flown at four times the speed of sound, enabled people to communicate with computers using only their brain-waves, and analyzed the cellphone records of a significant portion of the worlds’ population. He is a Senior Member of IEEE and held the title of Master Inventor while at IBM.
What Can We Learn About Autism from the Web?
Ayelet Ben Sasson
Autism spectrum disorder (ASD) is a complex neuro-developmental disorder, and as such it is not clear what are the core behavioral, environmental, physiological, medical, and genetic parameters of this disorder and how they are inter-related. Individuals with ASD face different sets of symptoms, severity, co-morbidities, and/or etiology. ASD rates continue to rise and despite notable progress in genetic and molecular biology we are far from understanding the blue print of this disorder. Individuals with autism and their families are increasingly relying on the web for informational and social-emotional support. This opens a window into immense amounts of natural data describing the daily expression of autism and its co-morbidities. In this talk I will outline mixed traditional and modern computational methods we have implemented towards conducting basic science using internet data. I will describe ways to deal with the ‘noise’ associated with internet data and discuss validity testing mechanisms. Finally I will present work in progress with the Technion on a cognitive clustering interface for crowdsourcing information related to autism in an interactive algorithm mediated manner.
Ayelet Ben-Sasson is an Assistant Professor at the Department of Occupational Therapy, University of Haifa. She has graduated with a Sc.D in Rehabilitation Sciences from Boston University and conducted her doctoral studies at the Boston University STAART center. Her post-doctoral training was at the Psychology Department of UMASS Boston and at the Center for the Study of Child Development of the University of Haifa. Her research interests relate to autism spectrum disorders, sensory processing disorders, early developmental screening, psychopathology, and online clinical data. She has conducted the first longitudinal community early screening study for ASD in Israel, which has been published in various interdisciplinary platforms. Among her awards and fellowships are the LEND and the Alon Fellowships. Currently she is working on multiple research projects that analyze autism data from various web-based systems.