New AI technology detects cancer mutations in minutes

Haifa, Israel – A new Israeli technique that analyzes digitized biopsy images in real time can now detect cancer mutations in minutes instead of weeks.

This artificial intelligence (AI) tool was the initiative of a young computer scientist who lost her mother to ovarian cancer five years ago.

Dean Bittan, 34, told The Epoch Times that he knew about the long and difficult weeks cancer patients waited for a diagnosis because he was on the front lines.

“My mother passed away from cancer, which is also why I started the company with my partner,” he said.

Bitan co-founded Imagene, an Israeli company that uses an AI algorithm that detects changes in the appearance of cells on digital biopsy slides to help doctors cut mutation detection from weeks to minutes and CEO.

The company has developed a special algorithm that can identify the unique appearance of cells with different mutations.

‘Literally lifesaving’

The combination of digital pathology and Imagene’s AI “allows us to identify mutations at literally life-saving speed,” said Ido Wolf, professor and director of the Department of Oncology at Tel Aviv Medical Center. told the Epoch Times.

Not currently a diagnostic tool. More research and evidence is needed. However, Wolf said there are individual circumstances in which patients die without immediate treatment.

Dov Hershkovitz, professor and director of the Institute of Pathology at Tel Aviv Medical Center, told The Epoch Times about a 41-year-old mother of three who recently came to the oncology department with headaches. A CAT scan was performed and found a mass in the lung and a metastasis in the brain.

They biopsied the mass and realized that this patient needed urgent care. “We don’t have time. If we wait until the full genetic test is done, this patient will have to undergo radiation to the brain,” Dr. Hershkovitz said.

“We really wanted to save her from this treatment,” he said. The disease was progressing very rapidly.

The hospital had already been working with Imagene for research purposes for two years, he said.

“They took two minutes,” he said.

Following Imagene’s morphological diagnosis, Hershkovitz said he knew the target gene.

AI tools have allowed medical staff to plan accurate treatments for patients.

Hershkovitz said he was able to learn the diagnosis and start treatment less than two days after the patient showed up at the oncology department.

“The improvement has been amazing,” he said. “She didn’t need radiation to her brain.

Imagene’s AI enables the analysis of digital slides without performing molecular analysis, DNA extraction, or very time-consuming and expensive assays, Hershkovitz said. specific mutations can be transmitted.

“If we knew what the expected mutations were, we could save money, save time, and save lives,” he said.

Epoch Times photo
Imagene genetic modification detection. (Courtesy of Imagene)

digital pathology

Until the last few years, pathologists have looked at tissue wrapped in paraffin blocks by cutting very thin samples of tissue and attaching them to slides. After exposing tissue sections to various dyes, they are examined under a microscope by a pathologist.

Digital pathology can scan tissues at high resolution, allowing samples to be analyzed at a much higher level than the human eye can see under a microscope.

AI scans digitized samples and finds small differences that distinguish cancer types according to different mutations.

Cancer cells with one type of mutation look different than cancer cells with another type of mutation. Mutations not only affect the genetic structure of the mutation, but also affect the entire structure of the cell.

A combination of digital pathology and AI can detect specific mutations with high probability.

“This is the revolution,” said Wolfe.

‘High precision’

Sheba Medical Center, Israel’s largest hospital, announced the study in September with Imagene. modern pathology Journal by Nature.

In this study, we compared the performance of traditional testing methods and AI methods.

“Of the 140 examinations performed using the AI ​​solution, only one case showed discrepancy with the final pathology report, demonstrating the high accuracy of the AI ​​solution,” the study said. says.

About a third of patients “unfortunately, when they arrive with metastatic lung cancer,” Iris Bershak, professor and director of the Institute of Pathology at Sheva Medical Center and author of the paper, told the Epoch Times. likely to die,” he said. [and have to wait] Until pathology laboratories around the world find the answer to a suitable treatment. ”

Barshack, who is also president of the Israeli Association of Pathologists, said Sheba had “trained” the algorithm on more than 600 cases before using the technology in the clinic in real time on nearly 100 cases.

At this stage, it’s not a regulated diagnostic tool, so it validates the lab findings she said. can start”.

Imagene’s AI algorithms, after going through a full validation process, will replace the lab work she said.

“We’re still cutting turnaround times to hours instead of weeks, and that’s saving lives,” says Barshack.

Avital Rbani, a spokesperson for Imagene, told the Epoch Times in an email: “We are in the process of filing with the FDA and CE. A mark that has been tested to comply with the European Union Marketing Directives.

Rabani added that Imagene works with many top-tier medical centers and pharmaceutical companies around the world, and its identity remains a secret.

Leah Onelly


Lia Onely reports from Israel for The Epoch Times.