Skip to content

M*Modal: Using Artificial Intelligence and Natural Language Understanding to Perfect Speech Recognition

Many years ago, a group of "geeks" at Carnegie Mellon University did research aimed at trying to get speech recognition technology to really work. One of those researchers was Michael Finke. Today, Finke serves as the President and CEO of M*Modal, a company that specializes in the exact science he and his fellow researchers tried to perfect. [caption id="attachment_2593" align="alignleft" width="399"] M*Modal technology takes speech recognition to new levels in the medical transcription world.[/caption]

M*Modal, with offices in Squirrel Hill, provides cloud-based clinical documentation solutions that allow for efficient and accurate speech recognition, medical transcription, clinical documentation improvement and coding. This makes it simple for doctors to capture information about any particular patient.

“We founded the company around this topic,” Finke said. “How to make it easier for the doctor to capture information about a patient, how to save them time and how to make the notes more accurate than they used to be.”

Finke said that instead of using keyboards to type out medical records and information, doctors simply use a microphone to speak about what they’ve seen from a patient. M*Modal’s software and artificial intelligence (AI) organizes the information for the doctor and enters it into the medical records.

“It’s about twice as fast as typing,” Finke said. “You gain a significant productivity benefit because you don’t have to type and click. You can just say it.”

M*Modal doesn’t just obtain the information. It also understands it. The documentation from the doctors is run through M*Modal’s data centers and is analyzed based on what other information is known about a specific patient.

Then, Finke said, the data needs to be encoded. Natural language understanding and AI are applied in order to read the records and generate billing codes that are submitted to insurance companies.

“The doctors don’t want to directly dictate into the medical records,” Finke said. “They use our transcription services. We use our speech recognition technology to create a document, and then we have people who work for our company who review it for the doctor so the doctor can confirm and sign off on it.”

“It’s about twice as fast as typing,” Finke said. “You gain a significant productivity benefit because you don’t have to type and click. You can just say it.”

The company isn’t stopping there, however. Finke said the next step is to allow an ongoing dialogue between the patient, the doctor and the computer, all while the doctor is seeing the patient in real time.

“Instead of seeing the patient, letting the patient go and now start documenting, what we’re doing now is to help them to have the documentation already ready,” he said. “This is AI really at work. This is next level.

“The doctor today spends about 60 percent of their time working with the patient record, not working with the patient. We can eliminate a significant portion of that simply because the documentation thus becomes byproduct instead of a special thing they need to do. M*Modal is creating time to care for the patients.”

M*Modal’s breakthrough technology is certainly impressive, but according to Finke, the ultimate goal of the company is to save lives.

“Going to the extreme, that we could save lives with this, I think that should be a good message,” he said. “I think it’s what motivates people to come here. We’re not using technology for better placement of advertisements. We focus on trying to do a good thing for health care.”

The company is working on expanding and is looking for “talented, highly motivated and technically excellent people” to join the team, according to its website. Positions could potentially be open for virtual media scribe coders, software engineers or clinical documentation specialists. Anyone interested can apply on M*Modal’s website.