success story

UTHealth accelerates medical research by leveraging AI and ML in collaboration with Virtusa, Cardinal Health, and AWS

Virtusa collaborated with The University of Texas Health Science Center (UTHealth), Cardinal Health, and Amazon Web Services (AWS) to advance medical research using the latest artificial intelligence (AI) and machine learning (ML) technologies. This research focused on finding the best treatment and management strategies for subarachnoid hemorrhage (SAH) based on healthcare data and computer simulations.

Ensure that you open the links to enable the Next button, assuming you have read all the documents.
Okay & NextClick to Next
The Challenge

While the future of life sciences is driven by outcome-based innovation with a high degree of personalization, the life sciences industry has been a late adopter of innovative technologies despite the large R&D budgets. As a result, the industry is grappling with the challenges brought in by the sudden surge in digital disruption. 

Moreover, most life sciences companies generate terabytes of data that sit in silos. For a long time, companies did not consider it suitable for retrospective analysis. However, things are changing. The life science industry is now ready to unlock the power of simulated data.

The adoption of AI technologies has become more critical than ever before. Experts are now pushing to create intelligent solutions to make lives better and businesses more efficient. However, the fundamental challenge in fuelling this technological and economic growth is quantifying data in algorithmic predictions and decisions. 

The Solution

Virtusa and Cardinal Health simulated a comprehensive dataset comprising electronic health records (EHR) representing more than 30,000 patients. We leveraged vLife®, Virtusa’s cloud-based platform, with an extensive HIPAA-compliant data lake with multiple data sources, pre-built APIs, AI, and ML models.  It is used to uncover hidden trends that led to new treatment strategies and cures for various illnesses.

UTHealth leveraged this simulated data to train and evaluate ML models to predict treatment outcomes for specific illnesses. About twenty faculty and students from Biostatistics and Data Science department at UTHealth School teamed with various medical and data experts to generate data. Meanwhile, the data provided through Cardinal Health’s proxy patient population simulation required minimum data cleaning and preparation. Once ready, the data was run in an ML application model to generate accurate predictions.

AWS services used:

  • AWS Redshift
  • AWS S3
  • AWS EC2
  • AWS Elastic Beanstalk
  • AWS CloudFront
  • AWS CloudFormation
The Solution
The Benefit

We helped UTHealth accelerate research and augment existing real-world evidence using simulated, ready-to-use clinical data. UTHealth utilized the simulated data for data comparisons, training machine learning models, and verifying model training methodologies. The fact that the data was simulated and, therefore, non-PHI enabled UTHealth to publish the data to validate their research openly. 

Related content