Sunday, 2024 July 21

AI to boost doctors’ efficiency with imaging diagnostics: Inside China’s Startups

Hour-long waits are not uncommon in hospitals in China. But while that’s also ordinary in many countries, the COVID-19 pandemic has brought medical efficiency to the fore of public conversation globally.

Nowhere was this more evident than in the outbreak’s epicenter, Wuhan, where tens of thousands of fever-stricken patients rushed to hospitals as the country quickly realized that it was facing something much worse than ordinary flu season.

Luckily, Chinese tech companies came in support of medical professionals in the frontline against the pandemic, and Shukun Technology, a Beijing-based startup, started developing an AI detection system for pneumonia induced by COVID-19, rolling out the first operative product, in just 15 days.

The system, operated by trained medical professionals, can identify suspicious lung lesions in three seconds using CT scans, saving roughly half an hour per patient, according to the firm’s founder and CEO Ma Chun’e.

The company donated its new software to about a dozen hospitals treating COVID-19 patients in China, including the Central Hospital of Wuhan, the makeshift Leishenshan hospital, also in Wuhan, the Beijing’s makeshift Xiaotangshan hospital, and other medical centers, helping doctors screen and diagnose tens of thousands of patients.

Shukun’s contribution led to Beijing’s municipal government inviting Ma to a press conference in April, on how AI technologies played a role in the battle against the pandemic.

Shukun CEO
Ma Chun’e, Shukun’s founder and CEO Photo provided to KrASIA

In June, Shukun announced the closing of a Series B1 fundraising round, securing RMB 200 million (USD 28.3 million), funds that will be destined to apply Shukun’s AI technologies in diagnosing cardiovascular and neurological diseases, as well as cancer detection, in hospitals across the country.

KrASIA recently spoke with Ma about how she founded her company, her current projects, as well as Shunkun’s future plans.

Starting from heart disease

Before founding Shukun, in 2017, Ma worked at IBM for 10 years, in charge of incubating the global tech giant’s enterprise-facing AI platform, named Watson, while also leading R&D for its cloud platform.

After earning over a decade of experience with AI, Ma felt AI could do something to help increase efficiency in China’s medical sector, where doctors are usually in short demand, she told KrASIA. 

She thought about using AI to screen coronary artery disease. The company’s goal was to create an advanced and faster method to read the results of computed tomography angiography (CTA) tests for diagnosis.

“We chose this area because data from the CTA tests are generated by machines, which is naturally easier for machines to read and understand,” said Ma.

Shukun’s AI-powered system is used in a hospital in Zhejiang to help screen for heart disease. Photo provided to KrASIA

The CTA test, which combines a CT scan with an injection of a colored dye to produce pictures of blood vessels and tissues, serves to show possible abnormalities.

While normally a doctor in the radiology department carefully reads the CTA test results and generates a report for a cardiovascular doctor, who will then make a final decision on a patient’s physical conditions, Ma and her team believed AI could do the majority of the screening job.

In early 2018, the company rolled out CoronaryDoc, an AI system optimized to read results of CTA tests, helping imaging doctors in their assessments, based on what the system could find.

“AI can do almost 90% of what human imaging doctor based in China’s top-level hospitals have to do when reading these tests. However, the final work, the most important part about judgment, still requires a doctor’s judgment,” Ma said.

She remarked that AI cannot replace doctors, but can help them to work more efficiently, allowing them to diagnose more patients in less time, while also reducing patients’ waiting time.

Shukun’s CoronaryDoc system has been used in more than 100 hospitals in China on a trial basis as of the end of September 2019, according to the company’s website.

Venturing to combat more diseases, including COVID-19

Shukun Beijing
Shukun’s AI-powered CT scan screening system for COVID-19 is used in Beijing Xiaotangshan hospital. Photo provided to KrASIA

In addition to CoronaryDoc, Shukun has been developing AI products for screening diseases related to vessels in the head, neck, and the main artery. However, lung diseases were never within the startup’s business scope until the novel coronavirus hit China at the end of 2019.

“That was what we thought we could do at that moment,” she added, explaining why the company rushed to come up with an AI-powered CT scan screening system for COVID-19.

Shukun was not alone. China’s tech giants⁠—Alibaba, Baidu, Tencent⁠—all rolled out their respective COVID-19 AI detecting systems, as supplements to other medical tests.

Read more: Alibaba says its AI-powered algorithm can identify coronavirus infections with 96% accuracy

Sun Yu, a cardiovascular doctor based in a Guangzhou provincial hospital, told KrASIA that he happened to use two AI detecting systems developed by Beijing Infervision Technology, and Shanghai-based United Imaging respectively, when collaborating with the radiology department amid the pandemic.

“It is really fast for these two systems to recognize CT scans, and they can even figure out some minor abnormalities easily unnoticed by doctors,” he said.

However, the technology is yet to be perfect. Sun recalls that for example, these AI systems once identified blood vessel fractures as spots on the lung, also mistakenly reading some respiratory artifacts as pneumonia.

“In general, such AI-powered systems do increase efficiency greatly, but there is a long way for them to come up with accurate reports to fully replace a human doctor,” Sun said.

Still, there is a bright future for AI’s development in China’s medical sector, as the market is at a very early stage, according to Sun.

Like Sun, Shukun’s Ma also believes there is a promising future for AI in the medical sector. Shukun has a plan to expand its technologies to cover more diseases gradually, but that will depend on how much money the firm will have at its disposal.

“There is much that can be done, but we need to be down to earth,” she said.

Ma also revealed that Shukun is about to close its Series B2 round, adding that the company’s products have also been generating revenue since last year. However, she did not disclose additional details.

“As human lives are being increasingly digitized, AI and digital doctors will become more common in the long run,” Ma affirmed.

This article is part of KrASIA’s “Inside China’s Startups” series, where the writers of KrASIA speak with founders of tech companies in the country  

Jingli Song
Jingli Song
I believe Chinese innovation at various level needs to be known by the world.

Related Read