Introduction

Oracle Life Sciences AI Data Platform is a cloud-based platform that allows healthcare organizations, research institutes, and life sciences companies to find insights using large-scale, de-identified patient data sets.

Oracle Life Sciences AI Data Platform gathers clinical data from the Oracle Learning Health Network, which includes patient records contributed by many healthcare institutions. The data is ingested, normalized, standardized, and de-identified using the Expert Determination methodology as defined in HIPAA Privacy Rule ยง 164.514(b)(1), and made available in both the Oracle Core Data Model and the Observational Medical Outcomes Partnership (OMOP) Common Data Model, with mapping to standard clinical ontologies such as ICD-10 and SNOMED.

Access Oracle Life Sciences AI Data Platform through your organization's dedicated Oracle Cloud Infrastructure (OCI) tenant. Through the tenant, you can access Oracle Autonomous Data Warehouse (ADW) and the de-identified data. Then you can use the Oracle Data Science Service to access notebook interfaces to create SQL queries, create cohorts, and analytics using R, Python, and Spark. Once the notebooks have been created, you can use Oracle Analytics Cloud (OAC) to visualize and investigate the data and validate or disprove your research hypothesis.

You need to sign in to the following applications separately to use Oracle Life Sciences AI Data Platform. Currently, Oracle Life Sciences AI Data Platform does not support federated authentication:

Oracle Analytics Cloud is the cloud-based platform that Oracle uses to enable customers to conduct analytics and business intelligence, analyze data, build dashboards, and apply analytics without needing to maintain physical infrastructure such as servers. While Oracle Analytics Cloud can connect to large datasets, performance and report response time could lag severely unless you curate the dataset in advance. For example, Real-World Data contains more than 30 billion records and queries could take hours or days without first filtering.

Oracle recommends using Oracle Analytics Cloud to:
  • Prepare a cohort (or list) of patients in Database Actions, then gather the Real-World Data for those patients. That way you start with a more manageable dataset.
  • Curate aggregates. Prepare summarized datasets grouping by attributes of interest instead of data at the most granular level.

Note:

Oracle does not recommend connecting Oracle Analytics Cloud to the full Real-World Data dataset.