Health Tech, Artificial Intelligence

VisualDx gets Gates grant to build offline diagnostic tool

VisualDx got a grant for $655,535 from the Bill and Melinda Gates Foundation to build a tool that can allow health workers to diagnose skin conditions without a signal. The app uses images and a series of questions to point users toward a diagnosis.

Screenshots of VisualDx’s software show the workflow for physicians. The company has built a diagnostic tool that allows providers to quickly diagnose skin conditions and other diseases.

A New York-based company won a grant from the Bill & Melinda Gates Foundation to build a diagnostic tool that can work without an internet connection. VisualDx won a $655,535 grant, which it will use to build an app that health care workers can use to recognize and treat skin conditions.

The grant is an offshoot of VisualDx’s existing work. The company has been building clinical decision support tools for providers since the early 2000s. Co-founder and CEO Art Papier, a dermatologist, said he created the company with the idea of helping busy doctors remember things.

presented by

Around that time, Google images didn’t yet exist, and Kodak was still making the shift to digital imaging. VisualDx did early work during the Anthrax attacks of 2001, where its system was used to diagnose the infection.

“We did a lot of work for state health departments that pushed our product into emergency rooms in the early 2000s,” Papier said.

VisualDx’s systems are used in more than 2,300 hospitals and clinics, and more than 90 medical schools. Two years ago, it spun up that work into a consumer-facing app called Aysa, that lets users look up their skin conditions at home.

Now, it is using its machine learning expertise and clinical database to build an app for clinicians that can diagnose skin conditions without an internet condition. It immediately analyses a photo of the skin lesion on-site, and then asks a few other questions to drive a better diagnosis, such the patient’s location and what other symptoms they’ve been experiencing.  The system doesn’t need an internet connection except to sync with the cloud when available, Papier said.

presented by

The idea is to put this tool in the hands of physicians and health care workers in areas with few available clinical resources.

“A lot of people think of dermatology as acne, warts and skin cancer,” Papier said. “But that skill to look at a rash and say, ‘I think that patient has measles, or dengue fever,’ is key to improving diagnosis for infectious disease. We’re improving the ability for non-experts to recognize the clues.”

In the years since its founding, Papier said VisualDx has built a database of  clinical information. The company only gets images from well-qualified sources, he said, and physicians review it to make sure its results are accurate. The company also makes a point of collecting information on dermatological conditions across different skin tones to ensure its results are accurate for all patients.

The company also has a database of information on symptoms and disease locations that helps it give a more precise diagnosis.

“Medicine requires precision. There’s a lot of hype from AI and big data. But it’s garbage in, garbage out,” Papier said. “The stakes are very high, it involves people’s lives, and you need high reliability systems. … We’ve made a conscious effort to lay a foundation that’s rock solid.”

The one-year grant from the Gates Foundation will help VisualDx develop the tool. But Papier plans to continue the project indefinitely, with a unit at VisualDx focused on global health. He also hopes to expand the tool beyond skin conditions in the long run, as VisualDx has done with its other systems.

“The reason we’re working in skin so much because there’s a tremendous need for skin diagnosis,” he said. “There’s a global shortage of experts.”

Photo Credit: VisualDx