Ontology-enabled Healthcare Applications exploiting Physical-Cyber-Social Big Data
Download as PPTX, PDF10 likes1,553 views
Kno.e.sis at Wright State University is a leader in research applying semantic integration in healthcare, with a focus on using big data to address clinical and social challenges. The center manages a budget of over $11 million, primarily funded by competitive federal grants, and works on multiple applications, including monitoring drug abuse trends and analyzing health behavior through social media. Key projects include eDrugTrends, KHealth, and Midas, which collectively aim to enhance understanding of public health issues and improve patient outcomes.
Ontology-enabled Healthcare Applications exploiting Physical-Cyber-Social Big Data
- 1. 1 Ontology-enabled Healthcare Applications Exploiting Physical-Cyber-Social Big Data Ontology Summit for the Health Care Track on Semantic Integration, 7 April 2016 Amit Sheth Kno.e.sis – Ohio Center of Excellence in Knowledge-enabled Computing: An Ohio COE on BioHealth Innovation Wright State University Special thanks: Sujan Parera
- 2. 33% 35% 32% Kno.e.sis: Ohio Center of Excellence in Knowledge-enabled Computing DoD & Industry • Metabolomics & Proteomics • Medical Info Decisions • Human Detection on Synthetic FMV • Sensor & Information • Material Genomics • Cardiology Semantic Analysis NIH: National Inst. of Health • kHealth - Asthma • eDrug Trends • Depression on Social Media • Drug Abuse Early Warning NSF: National Science Foundation • Harassment on Social Media • Citizen & Physical Sensing • Twitris - Collective Intelligence • Aerial Surveillance • Visual Experience • Web Robot Traffic Kno.e.sis’ research in World Wide Web ranks Wright State University among the top 10 organizations in the world based on 10-yr impact. Its total budget for currently active projects is $11,443,751, with $5,912,162 for new projects starting after July 2015. The significant majority of funds are highly competitive federal grants. World-class research is complemented by exceptional student outcomes and commercialization with local economic impact. As an Ohio COE on Bio Health Innovation, Kno.e.sis conducts research leading to building intelligent systems for clinical, biomedical, policy, and epidemiological applications. Example clinical/healthcare applications include major diseases such as asthma, depression, cardiology, dementia and GI. This is complemented by social and development challenges such as marijuana legalization policy, harassment on social media, gender- based violence, and disaster coordination. 60+ Funded Students • 40 PhD • 16 MS • 5 BS
- 4. Projects @ Kno.e.sis Image Credits: https://www.boxchilli.com/wp-content/uploads/2014/07/socialmediamarketing.png , http://rotwnews.com/wp- content/uploads/2014/04/DRUG-TRENDS-Talk.jpg, http://slapheadmarketing.com/wp-content/uploads/2012/05/Forum-Marketing.jpg eDrugTrends is social media data analytics platform to monitor the cannabis and synthetic cannabinoids usage. It uses social media and Web forums data to: 1) Identify and compare trends in knowledge, attitudes, and behaviors related to cannabis and synthetic cannabinoid, and 2) Identify key influencers in cannabis and synthetic cannabinoid-related discussions on Twitter. eDrugTrends Data Sources Project Wiki Daily average content: Tweets: 135,553 Forum Posts*: 8,899 Total: 144,452 * Bluelight, Drugs-forum, and Reddit
- 5. Projects @ Kno.e.sis Image Credits: https://i.ytimg.com/vi/GOK1tKFFIQI/maxresdefault.jpg, http://slapheadmarketing.com/wp-content/uploads/2012/05/Forum- Marketing.jpg, https://www.boxchilli.com/wp-content/uploads/2014/07/socialmediamarketing.png, https://encrypted- tbn1.gstatic.com/images?q=tbn:ANd9GcSucmCuyWvX4dFHv5XvS3KzjvD11hC8HwK9N4004LnBZOGLOgf6, http://www.crmchealth.org/sites/default/files/images/medical-records/Medical_Records_0.png?1314713869 Identifying combinations of online socio-behavioral factors and neighborhood environmental conditions that can enable detection of depressive behavior in communities and studying access and utilization of healthcare services Depression Behavior Data Sources Electronic Medical Records Public Surveys Project Wiki Depending on collection method, We get 7-17K tweets per day, and Have 800K to 18M total tweets in several months.
- 6. Projects @ Kno.e.sis This project seeks to understand and satisfy users’ need for keeping track of new information in healthcare and well-being. The project harvest collective intelligence to identify high quality, reliable and informative healthcare content shared over social media based on following analysis: Text Analysis, Semantic analysis, Reliability analysis, Popularity Analysis. Social Health Signals Data Sources Image Credits: https://www.boxchilli.com/wp-content/uploads/2014/07/socialmediamarketing.png, https://encrypted- tbn2.gstatic.com/images?q=tbn:ANd9GcS6qI3Z_Y0Uh0sPNCgy0J_0d66-5NsCwK3VqWsIkAKRmqjTSXK0uA Project Wiki
- 7. kHeath analyzes both active and passive observations of the patients to generate the alarms that helps to improve health, fitness, and wellbeing of the patient. It uses Semantic Sensor Web technology, Semantic Perception, and Intelligence at the Edge to enable sophisticated analysis of personal health observations. kHealth Projects @ Kno.e.sis Data Sources image credits: https://www.boxchilli.com/wp-content/uploads/2014/07/socialmediamarketing.png, https://www.cooking- hacks.com/media/cooking/images/documentation/e_health_v2/e_health_sensors_small.png, http://www.co.freeborn.mn.us/ImageRepository/Document?documentID=483 Public Health APIs Project Wiki
- 8. Projects @ Kno.e.sis Monitor the health status of the military personnel in training period through self- reported fitness notes and continuous 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. MIDAS Data Sources image credits: https://www.cooking-hacks.com/media/cooking/images/documentation/e_health_v2/e_health_sensors_small.png, https://encrypted-tbn3.gstatic.com/images?q=tbn:ANd9GcROFoUIaLPDWcvcCmoi1-sl8Bl3CPUtZooX5HHPuDiQKGI7oFZfuQ Self-reported Data Project Wiki
- 9. Projects @ Kno.e.sis PREDOSE developed techniques to facilitate prescription drug abuse epidemiology, related to the illicit use of pharmaceutical opioids. PREDOSE is designed to capture the knowledge, attitudes and behaviors of prescription drug abusers through the automatic extraction of semantic information from social media. PREDOSE Data Sources image credits: https://www.boxchilli.com/wp-content/uploads/2014/07/socialmediamarketing.png, http://slapheadmarketing.com/wp- content/uploads/2012/05/Forum-Marketing.jpg, https://encrypted- tbn3.gstatic.com/images?q=tbn:ANd9GcTrMsVTVc6RJrWZtst5ZTILWoD83HO0DPbj3I89YSqMiNRdwI7S Project Wiki
- 10. Projects @ Kno.e.sis The scientific analysis of the parasite Trypanosoma cruzi (T. cruzi), the principal causative agent of human Chagas disease, is the driving biological application of this project. We developed and deployed a novel ontology-driven semantic problem-solving environment (SPSE) for T.cruzi SPSE – T.cruzi Data Sources image credits: https://encrypted-tbn2.gstatic.com/images?q=tbn:ANd9GcRB-YQT3LWMXm9vfv3IdclcMDjP-_ChizcFMw53OAkptnHdaUAn6w, http://www.clfs.umd.edu/biology/machadolab/images/trypanosoma.jpg, http://web.eecs.umich.edu/~dkoutra/courses/W16_484/ Public & Private Databases (Uniprot, GO, KEGG, TriTrypDB Project Wiki experimental data from mass spectrometry and microarray experiments Textual Data
- 11. Ontologies Developed at Kno.e.sis • Drug Abuse Ontology – 83 classes, 37 properties • Depression Insight Ontology – ongoing work • Healthcare Ontology/ezDI Knowledge Graph – proprietary • Human Performance and Cognition “Ontology” – 2 million entities, 3 million facts (HPCO) • Ontology for Parasite Lifecycle – 360 classes, 12 properties (BioPortal) • Parasite Experiment Ontology – 142 classes, 40 properties (BioPortal) • Provenir Ontology - 88 classes, 23 properties (Provenir) – a key input to W3C provenance work Earlier at UGA: ProPreO (500+ classes), GlycO,…
- 12. Semantic Filtering Data Integration Knowledge Enrichment Entity Annotation Triple Extraction Sentiment Analysis/ Intent Mining/ User Modeling r1 Search/Browsing/S ummarization/ Trend/Analysis/Pre diction Knowledge Base Usages @ Kno.e.sis Data alone is not enough. KB+NLP+ML
- 13. Explanation Module Explained? Yes No Hypothesis Filtering Hypothesis Generation Hypothesis with High Confidence D D D DD D Patient Notes UMLS Knowledgebase Enrichment Sujan Perera, Cory Henson, Krishnaprasad Thirunarayan, Amit Sheth, and Sahas Nair. "Semantics driven approach for knowledge acquisition from EMRs." Biomedical and Health Informatics, IEEE Journal of 18, no. 2 (2014): 515-524.
- 14. 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 particular domain may contain required knowledge ● Partial knowledge about the domain can be used to efficiently acquire domain knowledge from data that can fill existing gaps in a knowledge base Data
- 15. Data Integration with Ontologies UniPort Internal Lab Data T.cruzi DB NCBI Data Sources PubMed T.cruzi immunology ontology Parasite Experiment ontology T.cruzi life cycle ontology Aligned Ontologies
- 16. ● Qualitative studies such as telephonic survey which suffer from limited population coverage and large temporal gaps. ● To address limitations of the qualitative studies, researchers have used various data sources such as social media (e.g. Twitter), web search logs, and neighborhood factors.....but in silos Depression Social Media Web Search log Neighborhood factors EHR data Depression Behavior
- 18. PREDOSE
- 19. Population Level Personal Wheeze – Yes Do you have tightness of chest? –Yes ObservationsPhysical-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> Wheezing ChectTightness PollenLevel Pollution Activity Wheezing ChectTightness PollenLevel Pollution Activity RiskCategory <PollenLevel, ChectTightness, Pollution, Activity, Wheezing, RiskCategory> <2, 1, 1,3, 1, RiskCategory> <2, 1, 1,3, 1, RiskCategory> <2, 1, 1,3, 1, RiskCategory> <2, 1, 1,3, 1, RiskCategory> . . . Expert Knowledge Background Knowledge 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 - continue Not Well Controlled – contact nurse Poor Controlled – contact doctor kHealth
- 20. Dealing with Heterogeneity He showed shortness of breath in last visit Dyspnea was observed in his last visit It is observed that patient has labored breathing The patient 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, indicating inadequate ventilation or low blood oxygen. rdfs:labelrdfs:label is_defined_as Expressing the Shortness of Breath explicit mention syntactic variation synonym positive implicit mention negative implicit mention individual literal
- 21. 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 pretty damn good didn’t do shit feel great Sentiment Extraction 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. Buprenorphine subClassOf bupe Entity Identification has_slang_term SuboxoneSubutex subClassOf bupey has_slang_term Drug Abuse Ontology (DAO) 83 Classes 37 Properties 33:1 Buprenorphine 24:1 Loperamide
- 22. Ontology Lexicon Lexico-ontology Rule-based Grammar ENTITIES TRIPLES EMOTION INTENSITY PRONOUN SENTIMENT DRUG-FORM ROUTE OF ADM SIDEEFFECT DOSAGE FREQUENCY INTERVAL Suboxone, Kratom, Herion, Suboxone-CAUSE-Cephalalgia disgusted, amazed, irritated more than, a, few of I, me, mine, my Im glad, turn out bad, weird ointment, tablet, pill, film smoke, inject, snort, sniff Itching, blisters, flushing, shaking hands, difficulty breathing DOSAGE: <AMT><UNIT> (e.g. 5mg, 2-3 tabs) FREQ: <AMT><FREQ_IND><PERIOD> (e.g. 5 times a week) INTERVAL: <PERIOD_IND><PERIOD> (e.g. several years) Smarter Data Generated with Ontologies
- 23. Thank You Visit Us @ knoesis.org Follow us @ facebook.com/Kno.e.sis One example of commercial applications: ezdi.com with additional background at http://knoesis.org/amit/hcls
- 24. Ohio Center of Excellence in Knowledge-enabled Computing - An Ohio Center of Excellence in BioHealth Innovation Wright State University
Editor's Notes
- #4: Cornell and Stanford?
- #6: https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcSucmCuyWvX4dFHv5XvS3KzjvD11hC8HwK9N4004LnBZOGLOgf6
- #9: https://encrypted-tbn3.gstatic.com/images?q=tbn:ANd9GcROFoUIaLPDWcvcCmoi1-sl8Bl3CPUtZooX5HHPuDiQKGI7oFZfuQ
- #21: We model the entities in a way that can be used to identify various styles of mentions of the same entity