A new blood test for liver cancer

20 November, 2022
Share with a friend

A new blood test for liver cancer

Researchers from Johns Hopkins University have revealed a new tool based on artificial intelligence to test blood and detect liver cancer.


The new technique succeeded in diagnosing most of the individuals undergoing a test that determined their presence of liver cancer based on blood plasma samples, by 80% of the 724 people.


The blood test, called Delphi, detects changes in DNA fragments produced by cancer cells in the bloodstream.


The researchers noted that the experiment included a first-of-its-kind test in analyzing the human genome in different ethnic groups, in which there are multiple causes of liver cancer. In their study, published in the scientific journal “Cancer Discovery”, the researchers indicated that 400 million people around the world are at risk of developing liver cancer as a result of chronic liver diseases.

Prof. Dr. Victor Velculescu, co-author of the study, said: “Increasing the early diagnosis of liver cancer can improve survival, in light of the low efficiency of current tests in diagnosing cancer.” According to the Medical Express website.


The researchers diversified between the sources of samples in the experiment, as they collected 501 samples from the United States and the European Union, as well as 223 samples from Hong Kong, and the researchers also typed between chronic liver injuries in the subjects of the experiment. The researchers sought to train the machine learning model by using artificial intelligence algorithms on various data to improve the accuracy of testing and injury diagnosis.


The Delphi technique relies on examining the size and quantity of DNA fragments in the blood on a large scale of the genome, and distinguishes between healthy DNA present in an accurate and organized manner within cells, and cancerous DNA scattered in cells and blood.


The technology is low cost as it requires less coverage of genome sequencing, which improves people's ability to conduct preventive examinations.

The results of the experiment varied among people with different liver diseases, but the technique was able to detect cancer in its infancy with an accuracy of 80%. The researchers are seeking to expand the trial and integrate the technology into clinical use.


Share with a friend