The University of Texas Health Science Center at Houston has partnered with a trio of tech companies for extensive new research into how electronic health record data can be harnessed by new AI algorithms to improve diagnosis and treatment.

WHY IT MATTERS
The Chair of UTHealth's Department of Biostatistics & Data Science, Hulin Wu, PhD, will lead a team of researchers focused first on innovating strategies for managing an acute condition, subarachnoid hemorrhage, and a chronic one, diabetes, using advanced computer simulations.

Virtusa and Cardinal Health will help UTHealth build a comprehensive dataset of deidentified EHR data from som 30,000 patients, made available using Virtusa's vLife cloud-hosting platform, deployed via Amazon Web Services' AWS Marketplace.

Using vLife technology, which comprises data lake capabilities, pre-built APIs, machine learning models and more, UTHealth researchers – data scientists, clinicians, epidemiologists, informaticians and computer scientists – aim spot hidden trends in the data that might lead to new treatments or cures for those and, eventually, other medical conditions.

The simulated data will be used to train and assess a series of new machine learning models to help predict various treatment outcomes. Cardinal Health's Proxi technology helps enable easier data hygiene and preparations for use in AI algorithms.

THE LARGER TREND
Artificial intelligence is gaining big momentum across healthcare as a key enabler to innovative diagnosis and treatments and improved outcomes. From precision medicine to predictive analytics, AI and machine learning tools are being put to work in new and interesting ways. 

To build on this momentum, Partners HealthCare this past week launched two new funds to spur AI development and digital tools. Its Artificial Intelligence and Digital Translation Fund has initial funding of $30 million over five years, and the Translational Innovation Fund will provide $50 million in funding over six.

ON THE RECORD
"We are developing novel statistical methods and machine learning approaches to interrogate EHR databases to identify the best treatment strategies and risk factors for a variety of diseases using real-world evidence," said UTHealth's Wu in a statement. "The EHR data simulated by Virtusa's platform will help us test and validate our new predictive models and machine learning algorithms before applying to the real EHR data."

"As data becomes the fuel driving technological and economic growth, a fundamental challenge is how to quantify the value of data in algorithmic predictions and decisions," said Anthony Lange, senior vice president, life sciences at Virtusa.

"Through the use of our technology, organizations like UTHealth can work with simulated clinical records to accelerate research, augment existing real-world-evidence, and save time by using clean, ready-to-use linked data," added Jeff Graham, principal, big data analytics at Cardinal Health.