CT and AI

Ping An’s technology can read CT scan faster and more effectively using AI | Source: Pexels

Chinese insurer uses AI to transform healthcare

IN CHINA, the uneven distribution of medical resources is a chronic problem, especially in the remote parts of the country. Challenges associated with obtaining proper medical services and the high cost associated with it have led to the impoverishment of less fortunate households.

Ping An Technology, the financial and health technology subsidiary of Ping An, the Chinese life insurance giant, is using cloud technologies, artificial intelligence (AI), and big data to empower the country’s healthcare community as part of its social responsibility efforts.

The possibilities offered by the combination of AI and healthcare seem endless. By integrating AI into the healthcare scenario, Ping An’s solutions have applications across the entire treatment process.

Starting from the moment the patient walks through the door of the doctor’s office or the entrance to the emergency room, all through the medical examination, right up to the final diagnosis and treatment.

In addition to improving the quality of the experience when dealing with the country’s healthcare system, the company is also using technology to deliver more convenient, thoughtful and intelligent medical services.

In the traditional healthcare scenario, patients spend a lot of time waiting for their turn in the doctor’s chambers, repeatedly filling out forms, going through medical examinations, and receiving a diagnosis, before finally receiving any kind of treatment.

In addition to the time-consuming, costly, and frustrating process, the high costs associated with diagnosis and treatment is also another reason why people are generally unhappy with the current state of medical services.

Faced with these challenges across the medical service sector, here’s how Ping An’s new solutions will improve the traditional healthcare scenario, boost efficiency in medical processes, and exponentially improve the patients’ experience:

Interpreting CT scans:

Ping An integrated the theory of Robust Control with AI algorithms involving deep as well as transfer learning to help AI interpret CT scan results.

In January of this year, Ping An’s technology broke the world records for detection of lung nodules and reduction in the number of false positives with the respective accuracy rates of 95.1 percent and 96.8 percent.

AI-based interpretation of CT scans, which is gradually being installed in more and more hospitals, is expected to help pathologists reduce the time needed to interpret a scan by more than half, effectively decreasing the number of missed diagnoses and misdiagnosis due to human factors.

Face recognition for medical treatment:

The insurer’s new technology has made it possible to use facial recognition as a component of the medical treatment process and as a mode of payment.

Today, face recognition for authentication is used in many medical procedures.

For patients, the technology enables them to make appointments, prove their identity prior to the hospital visit, prevent the issue of one person posing as an imposter for another in physical examinations.

For doctors and nurses, it further optimizes the environment surrounding the diagnosis and treatment process, standardizes diagnosis and treatment practices, and relieves the pressure on medical staff.

For medical organizations, the technology enables checking on work attendance through facial recognition to address loopholes in attendance systems and efficiency practices, significantly enhancing the service level.

Disease prediction through AI and Big Data:

The company has created the world’s first macro + micro method of disease prediction through AI and big data.

Ping An’s technology uses intelligent disease prediction and screening models to forecast the occurrence of infectious diseases one week ahead of an outbreak, and guides the public in the prevention of acute diseases.

According to data from the Chinese Center for Disease Control and Prevention, the prediction models for influenza and hand-foot-and-mouth disease both have an accuracy of more than 86 percent and, during seasonal peaks, accuracy can rise above the 90 percent mark.

The intelligent screening model for COPD has an accuracy of 92 percent. Both of these signal that there’s great potential in using big data and AI for disease prevention on a large scale.