Lunch & Learn: Delivering insight into complex patient journey with graph analytics - June 04 2020
1 like298 views
The document presents an overview of how graph analytics can deliver insights into complex patient journeys in life sciences and healthcare. It covers the use of knowledge graphs for drug discovery, patient stratification, and clinical trial outcomes, showcasing a demonstration using a synthetic patient population with extensive data integration. Additionally, it highlights upcoming events and workshops related to the application of graph technology in various sectors.
Lunch & Learn: Delivering insight into complex patient journey with graph analytics - June 04 2020
- 1. 1 Delivering Insight into Complex Patient Journeys with Graph Analytics
- 2. 2 • Welcome & Introduction • Graphs in Life Sciences: Patient Journey insights • Patient Journey Case Study / demonstration • Q&A David Rosenblum – Sr Field Engineer Joe O’Donnell – Account Executive, Life Sciences Alicia Frame, Data Science Product Manager
- 3. WE LIVE IN A CONNECTED SOCIOECONOMIC WORLD
- 5. Neo4j Innovation Lab The Wrong Tool for the Job
- 7. Social Media Graphs analyze social relationships Website Search Graphs rank web pages in their search engine results Product Recommendations Graphs understand behavior and preferences of online buyers Knowledge GraphSocial Graph Customer 360 Graph Property Graphs = Relationships
- 8. 8 • + 40,000 publications, case statistics, genes and functions, molecular data and more • Explore papers, patents, existing treatments and medications around the family of the corona viruses and other connected diseases Neo4j in Healthcare & Life Sciences Product Brand Management RWE Knowledge Graph • Connecting 50 research databases, 100k’s of Excel workbooks, 30 bio-sample databases • Bytes 4 Diabetes Award for use of a knowledge graph, graph analytics, and AI • Customized views for flexible research angles Drug Discovery • Competitive evaluation project with 3 yrs of visits, tests & diagnosis with 10’s of Bn of records • Finding similarities in patient journeys • Graph algorithms for identifying communities & Key Opinion Leaders
- 9. Graph Transactions Graph Analytics Data Integration Development & Admin Analytics Tooling Drivers & APIs Discovery & Visualization Developers Admins Applications Business Users Data Analysts Data Scientists Enterprise Data Hub You need more though…
- 11. Node ● Represents an entity within the graph ● Can have labels Relationship ● Connects a start node with an end node ● Has one type Property ● Describes a node/relationship: e.g. name, age, weight etc ● Key-value pair: String key; typed value (string, number, list, ...) 11 Property Graph Model
- 12. 12 Leveraging Patient Journeys Algorithms & Machine Learning Global analyses to find trends and predict outcomes Queries Fast, local decisioning and pattern matching You know what you’re looking for and are answering questions or extracting data You’re learning the overall structure of your graph to make inferences
- 13. Example questions: - What do the clinical histories of patients with a particular disease look like? - Can we identify off-label drug use or unexpected clinical outcomes? Applications: - Stratify patients into relevant subgroups - Use graph features to predict outcomes 13 Patient Journey for R&D
- 14. Example questions: - Can we identify patterns associated with specific outcomes and change them? - Are resources (clinicians, equipment, diagnostic tests) constrained? Are there alternatives? Applications: - Improving health outcomes - Cost savings measures 14 Patient Journey for Interventions
- 15. Example questions: - What happened to the patients in my treatment group? Were they different from the control? - Can we identify adverse outcomes and their causes? Applications: - Data lineage and transparency - Causal modeling & advanced ML 15 Patient Journey for Regulatory and Clinical Trials
- 17. Data loaded into Neo4j Synthea - Synthea is an open-source patient population simulation. - The Data is fake. - There are 1MM patients in our graph, with their encounters, providers, observations, procedures, conditions and prescriptions. - You could have additional data -- genomic profiles, expression data, test results
- 18. The Model
- 19. It’s a Database - you can query it
- 20. Visualizations to Drive Decision Making…
- 21. …. with drill through capabilities
- 23. ….Or “top n” Journies
- 25. Bloom… with guided paths
- 26. Node similarities via GDS Louvain & Community Detection algorithms
- 28. For more info/register, visit neo4j.com/connections Upcoming Connections 30 June 2020 Graphs in Life Sciences & Healthcare 28 July 2020 Graphs in Risk Mitigation 25 Aug. 2020 Graphs in Telco 17 Nov. 2020 Graphs in Government A digital event series to keep the graph database community connected, energized, and engaged in education.
- 29. Other Digital Events Join Us! 17 June 2020 Patient Journey Graph Data Analytics Workshop 30 June 2020 Graphs in Life Sciences & Healthcare Event Survey Coming in Email – Your Feedback is Important. Contact us: Ø Alicia Frame – Alicia.frame@neo4j.com Ø David Rosenblum – David.Rosenblum@neo4j.com Ø Joe O’Donnell – joe.odonnell@neo4j.com