Posted in: Healthcare

IBM Research showcases work in healthcare informatics at AMIA 2017

Update November 13 2017:

IBM’s contributions at AMIA 2017 were recognized in a number of ways:

Distinguished Paper Nominations

IBM researchers received two best paper nominations. Fewer than one in 20 accepted papers were nominated for this award.

Design and evaluation of a web-based decision support tool for district-level disease surveillance in a low-resource setting

During the 2014 West African Ebola Virus outbreak it became apparent that the initial response to the outbreak was hampered by limitations in the collection, aggregation, analysis and use of data for intervention planning. As part of the post-Ebola recovery phase, IBM Research Africa – Meenal Pore, David Sengeh, Purity Mugambi, Nuri Purswani – partnered with the Port Loko District Health Management Team (DHMT) in Sierra Leone and GOAL Global, to design, implement and deploy a web-based decision support tool for district-level disease surveillance.

This paper discusses the design process and the functionality of the first version of the system. The paper presents qualitative assessment of the tool prior to pilot deployment,  indicating that it improves the timeliness and ease of using data for making decisions at the DHMT level.

StressHacker: Towards Practical Stress Monitoring in the Wild with Smartwatches

While the pervasive stress in our world today can lead to a broad range of health problems, there haven’t been many practical solutions to help people navigate their daily stress. This paper from IBM researchers Tian Hao, Kimberly Walter, Marion Ball, Hung-yang Chang, Si Sun, and XinXin Zhu discusses StressHacker, a smartwatch-based system designed to continuously and passively monitor stress level using bio-signals obtained from on-board sensors making it highly accessible and suited for daily use in personalized stress management.

Working Group Leadership

AMIA Working Groups (WG) provide a way for members to collaborate, network and become involved in the development of positions, issues, white papers, programs, and other activities that benefit the informatics community. Elected WG leaders from IBM:

  • Xinxin (Katie) Zhu – Chair-Elect, Global Health Informatics Working Group (GHI-WG, 203 members); Membership Chair, Consumer & Pervasive Health Informatics Working Group (CHPI-WG)
  • Pei-Yun (Sabrina) Hsueh – Chair-Elect, Consumer & Pervasive Health Informatics Working Group (CHPI-WG)
  • Fei Wang – Chair-Elect, Knowledge Discovery and Data Mining Working Group (KDDM-WG)
  • Eileen Koski – Current Chair, Knowledge Discovery and Data Mining Working Group (KDDM-WG)

It is noteworthy that the CHPI-WG and the KDDM-WG are among the largest working groups with 585 and 701 members, respectively.

——————————————

Medical informatics is a growing discipline, seen as a key to accelerating the goals of value-based healthcare and the promise of personalized and precision medicine.

The American Medical Informatics Association (AMIA) is a community committed to the vision of a world where informatics can transform people’s care. It’s the professional association for informaticians in health to gain the knowledge and resources they need to drive healthcare transformation.

At a time when artificial intelligence (AI) is poised to make major contributions to healthcare, IBM Research is deeply engaged and committed to this endeavor and to the upcoming AMIA 2017 Annual Symposium in Washington D.C. from November 4-8. There we will present or participate in more than a dozen workshops, papers, posters and panels that showcase the exciting work our teams of researchers and industry professionals are contributing to medical informatics.

Highlighted Papers

Exploring the Causal Relationships between Initial Opioid Prescriptions and Outcomes
(W07, 11/4/17, 11:30 – 11:45, International Ballroom East)
Addiction to opioid medications has become an epidemic affecting, millions of Americans.  Drug overdoses accounted for more deaths than car crashes or gun violence in 2015. In this work, the authors analyze causal relationships between factors surrounding an initial opioid prescription and the subsequent outcomes of long-term use or addiction. It’s authored by IBM researchers Jinghe Zhang, Vijay Iyengar, Dennis Wei, Bhanukiran Vinzamuri, Hamsa S. Bastani, Alexander R. Macalalad, Anne E. Fischer, Gig Yuen-Reed, Aleksandra Mojsilovic & Kush R. Varshney

A Data-Driven Method for Generating Robust Symptom Onset Indicators in Disease Registry Data
(S13, 11/6/17, 9:24-9:42 AM Fairchild)
Disease registries are collections of observational data related to specific diseases which can be powerful tools for understanding the natural course of disease.  Since they are not all created for the exact same purpose, however, biases may exist in the data.
Huntington’s Disease (HD) is a serious neurodegenerative disorder that has been the subject of several large-scale observational studies related to the natural history and pathophysiology of the disease.

This paper from IBM researchers Zhaonan Sun, Ying Li, Soumya Ghosh, Yu Cheng, Jianying Hu and their colleagues at CHDI presents a framework that leverages control participants in a disease registry to mitigate bias and generate robust clinical assessments and symptom onset indicator. The proposed framework can be generalized to other disease registries and may help researchers better understand progression pathways of a target disease.

StressHacker: Towards Practical Stress Monitoring in the Wild with Smartwatches
(S81, 11/7/17, 2:57-3:15 PM, Jefferson East)
While the pervasive stress in our world today can lead to a broad range of health problems, there haven’t been many practical solutions to help people navigate their daily stress. This paper from IBM researchers Tian Hao, Kimberly Walter, Marion Ball, Hung-yang Chang, Si Sun, and XinXin Zhu discusses StressHacker, a smartwatch-based system designed to continuously and passively monitor stress level using bio-signals obtained from on-board sensors making it highly accessible and suited for daily use in personalized stress management.

Other highlighted activities IBM is participating in:

  • Working Group Pre-Symposia Workshop on Causal Inference – KDDM WG. Workshop Organizing Committee Lead: IBM Research healthcare analytics manager Kenney Ng. (W07, 11/4/17, 8:30 AM – 4:30 PM, International Ballroom East)
  • Interactive Panel: And Now for Something Completely Different: Successful Career Transformations in Biomedical and Health Informatics. With IBM Chief Nursing Officer Judy Murphy. (S83, 11/7/17, 1:45 -3:15 PM, International Ballroom East)

Here’s a list of all of IBM’s other activities at AMIA 17:

Papers

Interpretable Clustering for Prototypical Patient Understanding: A Case Study of Hypertension and Depression Subgroup Behavioral Profiling in National Health and Nutrition Examination Survey Data
(S24, 11/6/17, 10:48-11:06 AM, Gunston)
Authors:  Pei-Yun (Sabrina) Hsueh (IBM), Subhro Das (IBM)

Design and evaluation of a web-based decision support tool for district-level disease surveillance in a low-resource setting
(S26, 11/6/17, 11:24-11:42 AM,Fairchild )
Authors:  Meenal Pore (IBM), David Sengeh (IBM), Purity Mugambi (IBM), Nuri Purswani (IBM), Tom Sesay, Anna Lena Arnold, Ralph Myers, Anh-Minh Tran

Detection of adverse drug reactions using medical named entities on Twitter
(S47, 11/6/17, 3:48-4:06 PM, Gunston)
Authors:  Andrew MacKinlay (IBM), Hafsah Aamer (IBM), Antonio Jimeno Yepes (IBM)

Interactive Machine Learning for Medical Research: A Framework to Enhance the Engagement of Clinical Researchers
(S45, 11/6/17, 4:42-5:00 PM, Lincoln East/Monroe
Authors:  Bibo Hao (IBM), Yiqin Yu (IBM), Wen Sun (IBM), Yingxue Li (IBM), Guotong Xie (IBM)

A First Step Towards Behavioral Coaching for Managing Stress: A Case Study on Optimal Policy Estimation with Multi-stage Threshold Q-learning
(S57, 11/7/17, 8:48–9:06 AM , Fairchild)
Authors:  Xinyu Hu (IBM), Pei-Yun (Sabrina) Hsueh (IBM), Ching-Hua Chen (IBM), Keith Diaz, Ying-Kuen Cheung, Min Qian

Designing decision-support technologies for patient-generated data in type 1 diabetes
(S59, 11/7/17, 9:06-9:24 AM, Lincoln West)
Authors:  Si Sun (IBM), Kaitlin Costello

Bootstrap-based Feature Selection to Balance Model Discrimination and Predictor Significance: A Study of Stroke Prediction in Atrial Fibrillation
(S104, 11/8/17, 9:36-9:58 AM, Lincoln East)
Xiang Li (IBM), Zhaonan Sun (IBM), Haifeng Liu (IBM), Xin Du (IBM), Gang Hu (IBM), Guotong Xie (IBM)

Learning Doctors’ Medicine Prescription Pattern for Chronic Disease Treatment by Mining Electronic Health Records: A Multi-Task Learning Approach
(S113, 11/8/17, 10:30-10:48 AM, Jefferson West)
Authors:  Eryu Xia (IBM), Jing Mei (IBM), Guotong Xi (IBM), Xuejun Li, Zhibin Li, Meilin Xu

Data-driven Risk Characterization and Prediction of Renal Failure among Diabetic Type 2 Patients using Electronic Medical Records
(S117, 11/8/17, 10:30-10:48 AM, Monroe)
Authors:  Prithwish Chakraborty (IBM), Vishrawas Gopalakrishnan, Sharon Hensley Alford (IBM), Faisal Farooq (IBM)

Posters

Multi-Agent System Architecture to Measure Patient Engagement Based on Automatic Patient Authored Text Analysis
(Poster Session 2, Board 098, 11/7/17, 5:00-6:30 PM, Columbia Hall)
Fábio de Souza (IBM), Márcia Ito (IBM), Maira de Bayser (IBM)

Self-Controlled Structural Nested Mean Models
(W07, 11/4/17, 10:00 – 10:15 AM, International Ballroom East)
Authors:  Zach Shan (IBM), James Robin

Comments

Add Comment

Your email address will not be published. Required fields are marked *

informatics

Eileen Koski

Program Director, Health Data & Insights, Center for Computational Health