1 Ontology-enabled Healthcare Applications Exploiting Physical-Cyber-Social Big Data Ontology Summit for the Health Care Track on Seman8c Integra8on , 7 April 2016 Amit Sheth Kno.e.sis – Ohio Center of Excellence in Knowledgeenabled Compu8ng: An Ohio COE on BioHealth Innova8on Wright State University Special thanks: Sujan Parera
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Ontology-enabled Healthcare Applications Exploiting Physical-Cyber-Social Big Data
Ontology Summit for the Health Care Track on Seman8c Integra8on, 7 April 2016
Amit Sheth Kno.e.sis – Ohio Center of Excellence in Knowledge-‐enabled Compu8ng:
An Ohio COE on BioHealth Innova8on Wright State University
Special thanks: Sujan Parera
33%
35%
32%
Kno.e.sis: Ohio Center of Excellence in Knowledge-enabled Computing
DoD & Industry • Metabolomics & Proteomics • Medical Info Decisions • Human Detec8on on Synthe8c FMV
NIH: Na;onal Inst. of Health • kHealth -‐ Asthma • eDrug Trends • Depression on Social Media • Drug Abuse Early Warning
NSF: Na;onal Science Founda;on • Harassment on Social Media • Ci8zen & Physical Sensing • Twitris -‐ Collec8ve Intelligence • Aerial Surveillance • Visual Experience • Web Robot Traffic
Kno.e.sis’ research in World Wide Web ranks Wright State University among the top 10 organiza8ons in the world based on 10-‐yr impact. Its total budget for currently ac8ve projects is $11,443,751, with $5,912,162 for new projects star8ng aaer July 2015. The significant majority of funds are highly compe88ve federal grants. World-‐class research is complemented by excep8onal student outcomes and commercializa8on with local economic impact.
As an Ohio COE on Bio Health Innova8on, Kno.e.sis conducts research leading to building intelligent systems for clinical, biomedical, policy, and epidemiological applica8ons.
Example clinical/healthcare applica8ons include major diseases such as asthma, depression, cardiology, demen8a and GI.
This is complemented by social and development challenges such as marijuana legaliza8on policy, harassment on social media, gender-‐based violence, and disaster coordina8on.
eDrugTrends is social media data analy8cs plagorm to monitor the cannabis and synthe8c cannabinoids usage. It uses social media and Web forums data to: 1) Iden8fy and compare trends in knowledge, aitudes, and behaviors related to cannabis and synthe8c cannabinoid, and 2) Iden8fy key influencers in cannabis and synthe8c cannabinoid-‐related discussions on Twieer.
Iden8fying combina8ons of online socio-‐behavioral factors and neighborhood environmental condi8ons that can enable detec8on of depressive behavior in communi8es and studying access and u8liza8on of healthcare services
Depression Behavior
Data Sources
Electronic Medical Records Public Surveys
Project Wiki
Projects @ Kno.e.sis
This project seeks to understand and sa8sfy users’ need for keeping track of new informa8on in healthcare and well-‐being. The project harvest collec8ve intelligence to iden8fy high quality, reliable and informa8ve healthcare content shared over social media based on following analysis: Text Analysis, Seman;c analysis, Reliability analysis, Popularity Analysis.
kHeath analyzes both ac8ve and passive observa8ons of the pa8ents to generate the alarms that helps to improve health, fitness, and wellbeing of the pa8ent. It uses Seman8c Sensor Web technology, Seman8c Percep8on, and Intelligence at the Edge to enable sophis8cated analysis of personal health observa8ons.
Monitor the health status of the military personnel in training period through self-‐reported fitness notes and con8nuous monitoring with body sensors. The collected data is used to assess the health status of the person and suggest exercise regimen change or treatment plans if needed.
PREDOSE developed techniques to facilitate prescrip8on drug abuse epidemiology, related to the illicit use of pharmaceu8cal opioids. PREDOSE is designed to capture the knowledge, aitudes and behaviors of prescrip8on drug abusers through the automa8c extrac8on of seman8c informa8on from social media.
The scien8fic analysis of the parasite Trypanosoma cruzi (T. cruzi), the principal causa8ve agent of human Chagas disease, is the driving biological applica8on of this project. We developed and deployed a novel ontology-‐driven seman8c problem-‐solving environment (SPSE) for T.cruzi
Ontologies Developed at Kno.e.sis • Drug Abuse Ontology – 83 classes, 37 proper8es • Depression Insight Ontology – ongoing work • Healthcare Ontology/ezDI Knowledge Graph – proprietary • Human Performance and Cogni8on “Ontology” – 2 million en88es, 3 million facts (HPCO)
• Ontology for Parasite Lifecycle – 360 classes, 12 proper8es (BioPortal) • Parasite Experiment Ontology – 142 classes, 40 proper8es (BioPortal) • Provenir Ontology -‐ 88 classes, 23 proper8es (Provenir) – a key input to W3C provenance work
Explana8on Module
Explained?
Yes
No Hypothesis Filtering
Hypothesis Genera8on
Hypothesis with High Confidence
D
D D
DD
D
Pa8ent Notes
UMLS
Knowledgebase Enrichment
Sujan Perera, Cory Henson, Krishnaprasad Thirunarayan, Amit Sheth, and Sahas Nair. "Seman8cs driven approach for knowledge acquisi8on from EMRs." Biomedical and Health Informa2cs, IEEE Journal of 18, no. 2 (2014): 515-‐524.
Knowledgebase Enrichment
● Knowledge in a given knowledge base may not always sufficient ● Acquiring required knowledge in some domains is a tedious task ● Data available for a par8cular domain may contain required knowledge ● Par8al knowledge about the domain can be used to efficiently acquire
domain knowledge from data that can fill exis8ng gaps in a knowledge base
Data
● Qualita8ve studies such as telephonic survey which suffer from limited popula;on coverage and large temporal gaps.
● To address limita8ons of the qualita8ve studies, researchers have used various data sources such as social media (e.g. Twieer), web search logs, and neighborhood factors.....but in silos
Depression
Social Media
Web Search log
Neighborhood factors
EHR data
Depression Behavior
Depression Behavior
PREDOSE
Popula'on Level
Personal
Wheeze – Yes Do you have 2ghtness of chest? –Yes
Observations Physical-Cyber-Social System Health Signal Extraction Health Signal Understanding
<Wheezing=Yes, time, location> <ChectTightness=Yes, time, location>
<PollenLevel=Medium, time, location>
<Pollution=Yes, time, location> <Activity=High, time, location>
tweet reporting pollution level and asthma attacks
Acceleration readings from on-phone sensors
Sensor and personal observations Signals from personal,
personal spaces, and community spaces
Risk Category assigned by doctors
Qualify
Quantify
Enrich
Outdoor pollen and pollution
Public Health
Well Controlled -‐ con8nue Not Well Controlled – contact nurse Poor Controlled – contact doctor
kHealth
Dealing with Heterogeneity
He showed shortness of breath in last visit
Dyspnea was observed in his last visit
It is observed that pa8ent has labored breathing
The pa8ent was breathing comfortably in room air
He showed short of breath in last visit C0013404
shortness of breath dyspnea
Labored or difficult breathing associated with a variety of disorders, indica8ng inadequate ven8la8on or low blood
oxygen.
rdfs:label rdfs:label
is_defined_as
Expressing the Shortness of Breath
I was sent home with 5 x 2 mg Suboxones. I also got a bunch of phenobarbital (I took all 180 mg and it didn't do shit except make me a walking zombie for 2 days). I waited 24 hours after my last 2 mg dose of Suboxone and tried injecting 4 mg of the bupe. It gave me a bad headache, for hours, and I almost vomited. I could feel the bupe working but overall the experience sucked. Of course, junkie that I am, I decided to repeat the experiment. Today, after waiting 48 hours after my last bunk 4 mg injection, I injected 2 mg. There wasn't really any rush to speak of, but after 5 minutes I started to feel pretty damn good. So I injected another 1 mg. That was about half an hour ago. I feel great now.
Codes Triples (subject-predicate-object) Suboxone used by injection, negative experience Suboxone injection-causes-Cephalalgia
Suboxone used by injection, amount Suboxone injection-dosage amount-2mg
Suboxone used by injection, positive experience Suboxone injection-has_side_effect-Euphoria
experience sucked
feel preZy damn good
didn’t do shit
feel great
Sen8ment Extrac8on
bad headache
+ve
-‐ve
Triples
DOSAGE PRONOUN
INTERVAL Route of Admin.
RELATIONSHIPS SENTIMENTS
DIVERSE DATA TYPES
ENTITIES
I was sent home with 5 x 2 mg Suboxones. I also got a bunch of phenobarbital (I took all 180 mg and it didn't do shit except make me a walking zombie for 2 days). I waited 24 hours after my last 2 mg dose of Suboxone and tried injecting 4 mg of the bupe. It gave me a bad headache, for hours, and I almost vomited. I could feel the bupe working but overall the experience sucked. Of course, junkie that I am, I decided to repeat the experiment. Today, after waiting 48 hours after my last bunk 4 mg injection, I injected 2 mg. There wasn't really any rush to speak of, but after 5 minutes I started to feel pretty damn good. So I injected another 1 mg. That was about half an hour ago. I feel great now.
I was sent home with 5 x 2 mg Suboxones. I also got a bunch of phenobarbital (I took all 180 mg and it didn't do shit except make me a walking zombie for 2 days). I waited 24 hours after my last 2 mg dose of Suboxone and tried injecting 4 mg of the bupe. It gave me a bad headache, for hours, and I almost vomited. I could feel the bupe working but overall the experience sucked. Of course, junkie that I am, I decided to repeat the experiment. Today, after waiting 48 hours after my last bunk 4 mg injection, I injected 2 mg. There wasn't really any rush to speak of, but after 5 minutes I started to feel pretty damn good. So I injected another 1 mg. That was about half an hour ago. I feel great now.