Lumiata powers Predictive-First Healthcare by combining the brilliance of physicians with the analytical power of big data


I had a chance to catch up with Ash Damle, the Founder and CEO of Lumiata at the recent Health 2.0 Conference in Santa Clara.

Ash grew up in a family with multiple generations of physicians. By the age of 4, he was hacking away at computers and by 22, obtained a double degree in math and computer science from MIT. It’s not surprising that his passion is figuring out how can we make greater sense of health through data, the power of computation, and the move towards value-based care. It’s exciting to see where he’s taken this quest.

Last year, Ash established Lumiata (formerly known as MEDgle), an idea he’s been working on for more than six years. Lumiata is a health data science company in Silicon Valley backed by Khosla Ventures and Blue Shield Blue Cross Venture Partners. Lumiata’s goal is to take predictive analytics farther than anyone has gone before.

[Here’s a link to the video.]

 

Data science + medicine

The company has compiled vast amounts of medical data on symptoms, diagnoses, procedures, and medication from a variety of sources, including (but not limited to) textbooks, medical journals, and public data from medical Institutions, such as the Centers for Disease Control, the National Institutes of Health and the World Health Organization. He also has patient data from de-identified charts. If you can think of a relevant data source, the Lumiata folks probably have it – or plan to get it. According to Ash, the company has crunched more than 160 million data points so far.

Lumiata patient TimeLineUsing proprietary algorithms honed by more than 25,000 physician hours, Lumiata’s platform and associated APIs allow the analysis and organization of this massive amount of information into interconnected Medical Graphs, similar in concept to Google’s Knowledge Graph and Facebook’s social graph.

 

Mimicking how the best physicians think

In a way, Lumiata has built a tool that mimics the way the best trained physicians think—connecting all of the information gleaned from their interactions with patients (e.g., history, physical finding, lab tests, scans etc.) with the kind of knowledge they have crammed into their brains from years of medical school, residency and ongoing reading and practice. The plan is not to replace doctors with Lumiata’s tools, but rather to enhance their predictive capabilities—transforming average doctors into “super-docs”. These tools can also enable hospitals and other healthcare organizations, risk-bearing entities (aka insurers), and non-physician clinicians to also enhance their ability to do what they do for patients and populations.

Lumiata delivers it’s insights in the form of health stories that capture the fact that our health doesn’t exist in isolated data points. Understanding a patient’s complete story (clinical, social, financial) is critical to building true predictive models. Lumiata transforms data from claims, EHRs, sensors and more into patient stories, contextualized and up-to-date, to provide a 360 degree descriptive and predictive view of an individual or population. It can serve as a common model for the physician, care managers and payers to manage the future health, risk or cost of those individuals. These stories are the basis for understanding answers to key questions facing all healthcare practitioners everywhere: What to do? For whom? When? and Why?

A common question Ash gets is, “Will Lumiata replace physicians?” The answer is NO. Today, Lumiata’s team consists of clinicians, data scientists, care delivery experts and engineers who are working towards empowering ALL members of the care team—risk bearing organizations, physicians, care providers, health organizations, and patients themselves—to be predictive first, so they can drive the move towards value-based care.


13 COMMENTS

  1. This seems like a very well-conceived predictive model and I wish them the best of luck. The ground is littered with predictive models that basically “predicted” that high-cost patients would remain high-cost. This sounds like it’s thrown out the old model altogether. While I think they have their work cut out for them, I’d hate to be competing with them right now.

LEAVE A REPLY

Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.