Chinese scientists make breakthrough in early detection of ‘king of cancers’

0 21


The model combines a non-contrast computed tomography (CAT) scan with an AI algorithm. The team reported in a paper published Monday in the peer-reviewed journal Nature Medicine that the specificity of the pre-screening model reached 99.9 percent, indicating that there was only one false-positive case in every 1,000 tests.

Meanwhile, the sensitivity, or ability to detect pancreatic tumors, can reach 92.9 percent, beating the average radiologist’s performance by 34.1 percent.

03:28

Researchers in Singapore are using AI to read the mind through brain scans

Researchers in Singapore are using AI to read the mind through brain scans

One of the paper’s reviewers, Li Rujiang, associate professor of radiation oncology at Stanford School of Medicine, said the work is “an important step in the right direction for pancreatic cancer screening.”

According to a cancer hospital doctor at the Chinese Academy of Medical Sciences, AI-based imaging applications have not been approved by Chinese authorities. So while the initial results are impressive, there is still a long way to go before the technology can be used in clinical practice.

The team developed an early screening model for pancreatic ductal adenocarcinoma (PDAC), the most common subtype of pancreatic cancer, accounting for more than 95 percent of all cases. PDAC causes approximately 466,000 deaths annually worldwide.

in america, Pancreatic cancer It is currently the fourth leading cause of cancer in both men and women, and trends show that it will become the second leading cause of cancer death in the country by 2030.

In the year According to a study published in April by the US National Institutes of Health (NIH), which looked at trends in age-standardized cancer incidence, survival and death rates between 2000 and 2019, the number of people dying from pancreatic cancer has been steadily increasing – 0.2 percent annually from 2006 to 2019.

In the midst of the drug crisis, Chinese cancer medicine has made a historic campaign in the United States

Early or incidental detection can greatly improve a patient’s chances of survival. Studies show that high-risk patients with PDAC diagnosed at an early stage have a median overall survival of 9.8 years, while those diagnosed late live an average of 1.5 years.

However, there is a lack of effective and easily accessible screening technology for the general public.

with Expansion Pancreatic cancer is relatively low—less than 13 cases per 100,000—making the use of expensive contrast-enhanced CAT scans in the general population uneconomical.

“On the other hand, current pancreatic cancer early detection tools generally have poor accuracy, leading to many misdiagnoses and unnecessary panic,” Cao Cai, head of the Shanghai Institute of Pancreatic Diseases, said in an interview with mainstream media website Zhishifenzi. .

In a discussion last year between Cao and Lu Le, the head of the DAMO Academy’s medical team, they proposed AI to help with early cancer diagnosis.

The pair acted quickly and started a research project with over 10 leading medical institutions that aims to develop technology by combining CAT scans, which are widely used in medical facilities and hospitals, with AI. An ideal model for large scale pancreatic cancer screening.

Apple founder Steve Jobs died at the age of 56 of pancreatic cancer, which is difficult to diagnose. Photo: AP

Their brainchild was an algorithm to “identify pancreatic cancer with artificial intelligence” – known as PANDA. They were trained based on more than 3,200 image sets from a high-volume pancreatic cancer institute in China, about 70% of which were from patients with pancreatic lesions.

Thanks to the large data set, deep data management and innovative training strategy design, PANDA was highly intelligently trained. AI imaging expert.

DAMO Academy researchers have discovered that subtle density differences in non-contrast CAT scans, which may be difficult to detect with the naked eye, can be picked up by AI.

Applying PANDA to real-world clinical scenarios involving 20,530 patients to confirm its accuracy, the researchers found that the AI ​​tool could achieve an impressive sensitivity of up to 92.9 percent and specificity of 99.9 percent.

According to the data provided by Alibaba Cloud, the PANDA model has been used more than 500,000 times in hospitals and medical examinations, and has detected many early-stage pancreatic cancers that had previously been missed.

“The accuracy of the PANDA algorithm is superior to many recognized screening methods,” German clinician Jörg Klef and colleagues wrote in a commentary published in the same issue of Nature.

Beijing to limit the use of artificial AI in online health care activities

However, they cautioned that further evaluation is needed before AI-based testing becomes a widespread practice.

They suggested that any pancreatic cancer screening method should identify early stages such as “T1 lesions” that are smaller than 2 cm (0.79 inches) in diameter, but the AI ​​model from China did not provide specific and predictive values ​​for this subgroup. .

“The value of any cancer screening method is in reducing overall mortality. The study was retrospectively designed and could not assess the effect of the test on patient mortality,” he said, adding that AI-based screening should be evaluated in the same way as conventional screening.

“This AI model is still in its infancy and warrants further validation efforts,” said a doctor at the Chinese Academy of Medical Sciences. He added that given the low prevalence of pancreatic cancer, the demand for this AI tool would be limited.



Source link

Leave A Reply

Your email address will not be published.

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More