Pediatric diseases can be tricky for even the most seasoned doctors, but a new artificial intelligence (AI) system developed in China can diagnose childhood diseases as well as an experienced physician, raising hope that the technology could improve the diagnosis process and increase efficiency.
According to a paper titled “Evaluation and Accurate Diagnoses of Pediatric Diseases Using Artificial Intelligence,” the AI system was able to achieve high accuracy in diagnosing many dangerous conditions such as acute asthma exacerbations (97 percent), bacterial meningitis (93 percent), varicella (93 percent), influenza (94 percent) roseola (93 percent).
The study was published in the journal Nature Medicine, the first time a top medical journal has published research findings on the use of natural language processing (NLP) technology in making clinical diagnoses.
How the AI was trained
The AI system uses YITU Healthcare Technology’s NLP system to extract “clinically relevant information and establish a diagnostic system, the team proposed a model for electronic health record data that integrates prior medical knowledge and data-driven modeling. It has been applied in a large pediatric population and shows a strong performance in accuracy across multiple organ systems.”
YITU Healthcare Chief Executive Officer, Ni Hao, said their findings prove that AI technology is able to assist doctors to deal with massive data and diagnoses, “as well as provide support in uncertain and complex medical cases. Pediatric diseases can be tricky for doctors. An AI assistant will profoundly improve the diagnosis process and increase efficiency.”
As per the press release, researchers collected 101.6 million data points from more than 1.36 million outpatient visits between January 2016 and July 2017 at the Guangzhou Women and Children’s Medical Center to train and validate the model.
“The primary diagnoses included 55 diagnosis codes encompassing common diseases in pediatrics and representing a wide range of pathologies. The diagnostic system achieved a robust performance for two categories of pediatric disease: common conditions and potentially life-threatening conditions,” the statement said.
Optimistically, the researchers believe the AI system has great potential for clinical use, particularly for triage procedures. They said the system works by generating a predicted diagnosis with inputs of basic medical history, vital signs and physical exams from patients.
“These predictions help doctors make better use of precious time. The system can also assist physicians to diagnose complex or rare conditions to avoid misjudgments or biases. Most importantly, the system will provide high quality healthcare service as well as ease the tension of severe shortage of experienced pediatrics in China, benefiting both patients and doctors,” a statement said.
The researchers conceded that the impact of their study could be felt more in areas where healthcare providers are in relative shortage, but insisted that the “benefits of such an AI system are likely to be universal.”
Works like a human doctor
One of the co-authors of the study, Kang Zhang, a researcher at the University of California, San Diego, said the new technology was the first in which AI absorbs unstructured data and “natural language” to imitate the process by which a physician figures out what’s wrong with a patient.
“It can mimic a human pediatrician to interpret and integrate all types of medical data — patient complaints, medical history, blood and imaging tests — to make a diagnosis,” Zhang was quoted as saying.
Giorgio Quer, director, Artificial Intelligence at Scripps Research Translational Institute, said the study demonstrated the growing usefulness of AI in medicine, adding: “An interesting application of this system – that would require new data and that should be clinically validated – would be in the emergency department. This system could potentially analyze vital signs, basic history and notes from a physical examination by a nurse to help in the prioritization of patients who should be seen first by a physician.”
Now to the age-old question on whether machines are about to replace humans. Not quite yet, argued Duc Pham, a professor of engineering and head of the school at the University of Birmingham.
“Although the authors’ results show that on average their system performed better than junior doctors, it will not replace clinicians,” he said.
This is a view expounded by Quer, who said that although the system cannot yet outperform an experienced physician, it could assist junior physicians.
The study had 70 scientists. Included among them were researchers from the VA Healthcare System in San Diego who worked in collaboration with scientists from cities in China: Hangzhou (Hangzhou YITU Healthcare Technology Co. Ltd), Guangzhou (Guangzhou Medical University, Guangzhou Kangrui Co. Ltd. and Guangdong (Regenerative Medicine and Health Guangdong Laboratory).
The growing importance of AI in healthcare cannot be overemphasized. Recent studies have shown the technology outperforming humans. Last year, Google AI announced that its deep learning program named LYmph Node Assistant (LYNA) could detect advanced breast cancer with 99 percent accuracy.
Also last year, a computer algorithm, developed by the National Institutes of Health (NIH) researchers, was able to successfully identify changes in the cervix, pointing at cancer cells better than a human expert looking at the same images through a microscope.