In March, Microsoft co-founder Bill Gates warned that AI would replace most human jobs, including doctors and teachers, within a decade. Now we are beginning to see the signs of that prediction approaching reality.
On Tuesday, AI startup Ambience Healthcare unveiled a new medical coding model that beats doctors in the ICD-10 classification, a critical task for the medical business. According to the company, AI had 27% more coding accuracy.
Ambience builds AI tools that listen to doctors and patients’ conversations in real time and generate clinical notes. Its latest model can also identify the ICD-10 codes used to classify everything from influenza diagnosis to complex chronic diseases, as doctors record patient visits. This model was trained using OpenAI reinforcement learning tools customized for the medical field.
Ambience Healthcare’s AI platform outperforms clinicians’ performance by 27% in medical coding
There are over 70,000 ICD-10 codes. Getting them right is boring, but important. It affects hospital bill insurance companies, how illness is tracked, and how patient outcomes are measured. Traditionally, doctors or trained coders often get pressured and enter these manually at the end of a long shift. Mistakes can lead to false billing, compliance issues, and time for follow-up work.
Ambience says that the model can reduce its friction. In benchmark tests, the system was evaluated against the “Gold Panel” of the expert code. I then headed straight to 18 board-certified doctors. Results: AI performed 27% better than the physician average.
“Trained using OpenAI’s Enhanced Fine Tuning (RFT) technology, Ambience’s AI platform achieved a 27% relative improvement over physician benchmarks. It reduces coding errors by about a quarter and provides a powerful tool to tackle $266 billion in managed waste annually.
“We’re not replacing doctors and coders,” said Brendan Fortuner, head of engineering at Ambience. “What we’re doing is that we’re freeing them from management, and we’re fixing mistakes that will help improve healthcare and be safer and more cost-effective.”
The company’s chief medical officer, Dr. Will Morris, emphasized that ICD-10 data is more than just a deficit. This is the standard used to compare hospital and clinician outcomes. “If you think about it from a data perspective, it’s how you compare and contrast clinician A with B and B, or B to B,” he said. “It’s the cornerstone of quality,” he said in an interview.
Founded in 2020, Ambience is already in use by over 40 health systems, including Cleveland Clinic and UCSF Health. The company raises more than $100 million from Kleiner Perkins, Andreessen Horowitz and the Openai Startup Fund. Information says it is now seeking a fresh capital with a $1 billion North valuation. Vibe declined to comment.
The ICD-10 breakthrough adds to the growing set of features Amvience already offers. Startups previously deployed tools for CPT (current procedural terminology) code, but are now focusing on other bottlenecks in health care administrators, including previous approvals, usage reviews, and clinical trial matching.
The new model will be rolled out to Ambience customers this summer.
“Getting it right at the point of care is a fundamental change,” Morris said.
Founded in 2020 by Mike NG and Nikhil Buduma, the San Francisco-based atmosphere is supported by Andreesen Horowitz, the Open Startup Fund, Kleiner Perkins and other top-tier investors.
The founder of the atmosphere
🚀Want to share the story?
Submit your stories to TechStartUps.com in front of thousands of founders, investors, PE companies, tech executives, decision makers and tech leaders.
Please attract attention
Source link
