Welcome to our store!

New collections added on a weekly basis!


for all orders over $99.99 within CONUS

Imagene Profiles Cancer Biomarkers in Real Time

Medgadget Editors |

Cancer therapies have proliferated over the past few decades, improving outcomes for many patients. But this variety requires accurate diagnostics and appropriate decision making to choose the optimal course of therapy. The current gold standard of identifying which cancer mutation is present is new generation sequencing (NGS), which provides a comprehensive genetic report, but can take up to six weeks to complete, requires a significant tumor sample size, and can only be done in an advanced laboratory at a high cost.

Imagene, a company headquartered in Tel Aviv, Israel, has developed AI-based image analysis software that analyzes digitized biopsy slide images to provide immediate tumor mutation profiling. Thanks to support from Biomed Israel, the leading international Life Science and HealthTech conference in Israel (see more below about the upcoming event), we had a chance to learn about Imagene’s technology from Dean Bitan, CEO and Co-Founder of the company.

Dean recounted how only two decades ago we only had access to one-size-fits-all treatments that were effective for some patients, but did nothing for others. Thanks to DNA sequencing and the ability to understand the DNA, and the bugs within it, which lead to mutations, it became possible to develop drugs that better target cancers. This is precision medicine, but it requires a long and complex process, a high cost, and biopsy samples that can be difficult to obtain. This is very difficult psychologically for the the patient (and doctor) to wait for a long time to obtain results and not start treatment, as the cancer can deteriorate while waiting.

Imagene was founded in 2020 to make use of the capabilities of AI to be able to identify a variety of biomarkers using only a biopsy image. The company’s technology may significantly improve the diagnostic path for many patients on the way to treatment, making it quicker, cheaper, and available to many more people.

Currently, as few as 28% of cancer patients receive a comprehensive analysis of all 9 actionable cancer biomarkers (based on NCCN guidelines) and up to 64% of lung cancer patients don’t get the optimal treatment that’s available. This can be attributed in large part to insufficient testing for biomarkers of the nature of the mutation. In many cases, including with NGS, there’s a much lower sensitivity than expected. Moreover, interpretability can be poor because it requires well trained specialists that know what they’re looking at.

Imagene’s technology takes a biopsy image and within a couple of minutes provides a report ready for a diagnosis of the biomarker, identifying the actual cancer mutation present in the biopsy.

The time between biopsy and when the treatment is finally delivered, during which the nature of the cancer is being investigated, is measured in weeks right now, usually more than a month. Since this is too long for many patients, Imagene’s technology promises to significantly shorten the waiting time till treatment initiation for many patients. Potentially, patients may know what the diagnosis is and what the treatment will be on the same day.

Unlike many other diagnostic modalities, Imagene’s tech does not need any additional dedicated tissue. In fact, one digitized biopsy image, the diagnostic slide that was originally prepared for the initial pathology diagnosis, is all that is needed for an immediate gene alteration identification. It will not necessarily replace existing technologies, but Dean Bitan believes that his company’s capabilities will improve the value of existing diagnostics by focusing them on where their value shines the most and improving how specific indications are arrived at.

Artificial intelligence, if thought of as a generalization of models, works quite well to detect cancer mutations in all kinds of patients and in all kinds of places because patterns of cancer repeat themselves. Imagene’s merges self-supervised learning and other AI methods, where unlabeled data can be used to improve the results, as well as proprietary processing of prepared data to analyze images. Imagene is already working with 28 different biomarkers in eight different organs and is showing how their technology is standardized and providing the same level of results is accurate enough in order to assist in navigating clinical decision making. Currently the technology is still under clinical research, and soon it will be commercialized under the required regulatory requirements. Eventually the technology should be agnostic to any tissue type. Such work requires a lot of data points and they’re working with medical centers, labs, pharma companies, and using public data to continue developing the technology.

We hope to see Imagene showing off its technology at Biomed Israel, the leading international Life Science and HealthTech conference in Israel. This year it is scheduled for May 16-18, 2023 in Tel Aviv, and topics range from medical robotics, to bio-convergence, to the impact of AI on biopharma. Over 6,000 industry leaders, scientists, engineers, physicians, and investors will be attending for the 21st consecutive year of this conference. It is the largest event in Israel that brings together Israeli healthcare professionals and industry experts with international colleagues to work for three consecutive days on business opportunities, develop partnerships, and to seek new collaborations. Hundreds of Israeli life science firms will be showing off their products and technologies to attendees from all over the world. More info can be found at the Biomed Israel website. The conference is co-chaired by Ruti Alon, Founder and CEO of Medstrada, Ora Dar, PhD, Senior Expert, Medical Sciences and Health Innovation, and Nissim Darvish, MD, PhD, Managing Partner, Eliraz Ventures.

Here’s a video report from an Israeli news channel about Imagene’s tech:

Link: Imagene homepage…

Leave a comment

Please note: comments must be approved before they are published.