Data driven

Houston artificial intelligence startup aims to impact medical field

InformAI can use its data technology to help doctors with preventative care and diagnoses. Courtesy of InformAI

A Houston-based startup has a new technology that allows hospitals and medical establishments better access to its own data – which translates into more effective diagnoses and preventative care.

InformAI — founded by Jim Havelka, CEO, in 2017 — is introducing the technology to the Texas Medical Center. Havelka saw a need within the medical industry for this type of service.

"There were several things missing," says Havelka. "One was access to very large data sets, because it wasn't really until the last five or 10 years that digitalization of data, especially in the healthcare vertical became more widespread and available in a format that's usable. The second convergence was the technology, the ability to process very large data sets."

InformAI currently offers four unique solutions using artificial intelligence and deep learning algorithms: Paranasal Sinus Classifier; Brain Cancer Classifier; Patient Outcome Predictors; and Surgical Risk Predictors.

According to the website, both medical image classifiers assist physicians in detecting the presence of medical conditions. The Paranasal Sinus Classifier detects and distinguishes medical conditions prevalent in the paranasal sinuses. The classifier assists physicians by evaluating sinus medical conditions at the point of care, speeding up radiologist workflow by flagging medical conditions for further review, and providing a triage of pending sinus patient study reviews. The Brain Cancer Classifier focuses on several tumor types and has the potential to provide radiologists and surgeons with additional insights to inform their diagnoses and treatment plans.

In addition to the classifier solutions, the predictors are also key to patient care, as InformAI patient outcome predictors evaluate the risk of adverse outcomes from a surgical procedure.

"Our data set has 275,000 surgical procedures that we can use to look for patterns, and then use that to understand how a patient may react to going through that surgical procedure and that's a very valuable input to surgeons," Havelka tells InnovationMap. Patient outcome risks include mortality, stroke, prolonged ventilation, infection, re-operation, and prolonged hospitalization.

"The innovation is the ability to use artificial intelligence to augment the capabilities of the physician and flag diagnostics for them to consider," says Havelka. "For example, one of our image classifiers that reads three-dimensional CT head scans has the equivalent of thirty lifetimes of an ENT contained in the AI labeled training dataset. It would take thirty lifetimes of an ENT to see that same number of scans and associated disease state patterns. InformAI currently has 10 full-time employees and works with radiologists-in-residence in building solutions and conducting research. The startup partners with the Texas Medical Center, Nvidia, Amazon, and Microsoft.

"They're quite interested in what we've built because it's really cutting edge technology that we're doing," Havelka tells InnovationMap.

Havelka and his team also work with some of the largest physician groups and imaging companies in the country to build products. "At the end of the day our core competency is the ability to take data, medical images or patient data, and put it into a usable format to assist physicians in making better treatment decisions for patients," says Havelka. "We can flag and detect patterns, disease states, and risk profiles that can improve the decision making of the physicians for the patient.

InformAI has plans to fundraise this year, with a goal of raising $5 million to $6 million in a round.