Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. Learn More
How does a medical professional stay aware of the right procedures and treatments for patient ailments in the modern world? While many often rely on experience, there is another way that could have life-saving consequences. The trick is, it relies heavily on the power of artificial intelligence (AI).
New York-based medical startup H1 released a new update to its HCP Universe platform today to inject a dose of healthcare AI into medical intelligence. The HCP Universe platform is currently used by medical affairs teams at life sciences companies, which make sure doctors are aware of and use the latest science and medicine.
With HCP Universe, medical affairs teams can target the right doctors and educate them about the latest and most important medical treatments and which patients should receive that treatment.
“Our mission for this product is to make sure that the latest medicine is used on the right patients, so that patients get the right treatment,” Ariel Katz, H1 cofounder and CEO, told VentureBeat.
Join us in San Francisco on July 11-12, where top executives will share how they have integrated and optimized AI investments for success and avoided common pitfalls.
Using healthcare AI to improve adoption of new treatments
For H1, the use of healthcare AI is all about providing the intelligence to help medical affairs people find the right doctors – proactively.
“What we have done in the past is provide a platform for users to go and search and find doctors for a specific field or treatment, but that’s not the right thing to do,” Katz explained.
Instead, identifying and reaching the right doctors in a given field of medicine is key. Katz said that the updated HCP Universe platform has AI-powered features to highlight and help drive evidence-based medicine usage around the world.
The hardest part of making the data actionable for medical affairs, Katz explained, was relating the data together and putting it in the right taxonomies. For example, if a user searches for “obesity,” there are any number of medical people that could be involved, including endocrinologists, dieticians or even psychiatrists.
“If a patient searches for obesity, they don’t just want to find a doctor that specializes in diets, they want to find one that is relatable to that person’s needs,” Katz said. “So it’s relating the data together and then the machine learning libraries learn from user behavior to drive relevance.”
Why a graph database wasn’t enough
The idea of connecting relationships together is a common concept in graph databases. In fact, the HCP Universe platform is built with a graph database. But on its own, Katz said that his company has discovered that this is not accurate enough when it comes to important healthcare treatment decisions.
“If a recommendation engine is 80% accurate for a restaurant where you just want a bagel, you’re probably happy,” Katz said. “If it is 80% accurate and you’re trying to find a doctor and you should have been diagnosed with cancer, that’s not okay.”
With the machine learning libraries, H1 is able to learn from the data and can correlate complex relationships that the graph database doesn’t identify on its own. H1 uses AWS machine learning tools, including Sagemaker, to help power its healthcare AI efforts, Katz said.
When H1 launched, Katz noted that the biggest problem it had to solve was aggregating and collecting sources of information on medical professionals.
“We started by solving a data problem first and making sure all the information is accurate, trustworthy and reliable,” Katz said. “The next generation, which is what we’re launching, is how do you actually make it smart and turn that data into insights?”
Looking forward, H1 will be training its AI to provide clinical quality scores for medical professionals. For example, if a user wants to identify the best medical professional for treating bladder cancer, the system will help identify the best doctor based on multiple factors, including patient surveys and hospital readmission rates, among other pertinent factors.
“This information, on who is a better doctor and what’s a better hospital, will change the experience for many people who are engaged with the healthcare ecosystem,” Katz said.
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.