QUARK Health, an artificial intelligence （AI）assistant developed by Alibaba Group Holding based on the firm’s proprietary foundation model Tongyi Qianwen, has become the first AI model in China to pass the national deputy chief physician qualification examination, marking a big step for AI’s clinical reasoning.

The Quark Health Large Model exceeded the pass mark in all 12 common medical disciplines at the exam Tuesday, according to Chinese media reports. It also reached the qualifying score for the chief physician title in general medicine, oncology, dermatology and sexually transmitted diseases, and psychiatry.

The other eight disciplines were common internal medicine, common surgery, obstetrics and gynecology, pediatric internal medicine, stomatology, otorhinolaryngology, ophthalmology, and anesthesiology.

The associate chief physician post is China’s second-highest medical professional title, behind only chief physician. Those holding it must independently handle complex and challenging cases in their specialty, guide junior doctors in diagnosis and treatment, and play a key role in emergency and critical care.

Quark Health’s ability to retrieve and filter valid information is remarkable, according to Jiang Jile, associate chief physician of Beijing Jishuitan Hospital’s spine surgery department “Traditional search engines often just list information related to a patient’s query, while Quark can provide specific professional assessments and recommendations,” Jiang was quoted by Yicai as saying.

When a patient describes a symptom, Quark Health can comprehensively evaluate potential underlying diseases and suggest appropriate treatment timelines and options which is a major advancement, commented Wang Jing, associate chief physician of the Oncology Center at Wuhan University People’s Hospital.

However, passing the exam does not mean Quark Health can replace doctors, Wang noted.

The future application of such AI models in the consumer market will primarily focus on assisting doctors in diagnostic decisions and reducing communication costs between physicians and patients, she pointed out.

As health AI models advance, patients could use them before seeking care to identify potential causes of their symptoms, helping them pick the most suitable specialist, Wang said, adding that they could also aid in early disease detection and reduce misdiagnoses and missed diagnoses.

(SD-Agencies)