Hillel's Tech Corner: Ibex: AI-powered cancer detection

As many as 9.6 million people succumbed to cancer in 2018. Cancer has also inflicted an economic cost on the world, as much as $1.6 trillion in 2010.

By HILLEL FULD
Ibex Medical Analytics (photo credit: Courtesy)
Ibex Medical Analytics
(photo credit: Courtesy)
By now you don’t need to be told about the disastrous effects of cancer and its global impact. You probably know someone affected by it. 
What you might not know is that cancer is the second leading cause of death globally. As many as 9.6 million people succumbed to cancer in 2018. Cancer has also inflicted an economic cost on the world, as much as $1.6 trillion in 2010.
Every time I hear of another person battling, or even worse, losing the battle to cancer, I always ask myself, “How is it that in 2019, we have autonomous drones, self-driving cars, robots that replicate human behavior with incredible precision, and so many other technological breakthroughs, yet cancer is still so common and technology cannot seem to disrupt it?”
Well, Ibex Medical Analytics, a pioneering developer of cancer diagnosis tools, is seeking to battle the epidemic of cancer by leveraging artificial intelligence, or AI.
The obvious application for AI with cancer is to prevent the inevitability of misdiagnosis by humans. We have all heard stories. The conventional approach used by pathologists involves the examination of bodily tissue on glass slides using a microscope. In recent years, pathologists have turned to the digital realm, using a scanner to take photographs at high resolution that create a 1GB image of the glass slide. The digital slides can be archived and retrieved without concern of physical damage. It is comparable to the transition from analog to digital photography.
Ibex aimed to enhance and improve the experience and effectiveness of the technology. The company has also gone one step further with its use of using AI as part of the process. While acknowledging the benefits of digital pathology diagnosis, Ibex also realized that the still present human factor raises concerns of misdiagnosis.
Based out of Tel Aviv, Ibex was founded by Joseph Mossel and Dr. Chaim Linhart after both joined the Kamet start-up studio as entrepreneurs-in-residence in May of 2016. Linhart has a PhD in computer science from Tel Aviv University and previously worked for 20 years in R&D positions for start-ups dealing with data analysis and algorithms. Mossel has a master’s degree in computer science and has prior experience in product management.
Through their experience in data analysis, they recognized that Israel was a place of significant promise for healthcare software due to the existing datasets in the country’s HMOs. With that in mind, they entered into a strategic agreement with the Maccabi Health Services HMO after launching Ibex, granting them access to the HMO’s huge pathology archive – the largest in Israel. Ibex also has a close relationship with the Maccabi Institute of Pathology. The head of the institute, Dr. Judith Sandbank, serves as chief medical officer for Ibex.
USING THE digital scanner in the Maccabi Institute, the first application developed by Ibex dealt with prostate biopsies in which relevant data sets were created. A digital pathology scanner surveyed thousands of slides, and based upon all aggregate data, the first algorithm was developed for diagnosing prostate biopsies. Cases from several years prior were further analyzed and after successfully detecting cases of missed diagnoses, Maccabi requested that the system be deployed.
The AI system developed by Ibex is called The Ibex Second Read™ (SR) system through which digital slides and summaries of respective pathology reports are fed. Subsequently, the system conducts its own automated diagnosis and then compares it with the results in the pathology report. If the system detects a discrepancy, it raises an alert on the relevant slide, which is then sent back for a second review by the pathologist.
The efficiency of the SR system was demonstrated in its first week of deployment back in March 2018. Within a few days it managed to detect a discrepancy in a 55-year-old who was misdiagnosed with benign prostatic hyperplasia. The system helped to correct the diagnosis to “malignant.” It was the first time a piece of software successfully diagnosed an instance of cancer.
Many have expressed concern that introducing AI into various industries may replace or minimize human involvement. However, the pathologist remains in charge with the SR system as detected discrepancies are sent back for review. The use of such a system provides much needed help in maintaining an efficiency in light of an existing shortage of pathologists in Israel and in other countries worldwide, as well as a continued increase in the number of cancer patients. Most importantly, AI serves as a very plausible method in dealing with the inevitability of human mistakes.
Ibex continues to work closely with several other medical institutes worldwide – most prominent among them being the University of Pittsburgh Medical Center. Ibex also works with a large private network of pathology labs in France.
Ibex’s work and advancement of the AI industry in the field of cancer diagnosis has already attracted $11 million of investments in Series A funding (bringing total investment in the company to $14 million). aMoon Partners, a leading life sciences venture capital fund that works toward advancing cures for life-threatening diseases as well as reducing major cost-drivers in healthcare, provided the bulk of the funds.
Ibex’s original partnership of two quickly became a team of 20 employees and growing – something that might give you a glimpse to this company’s seriousness when it comes to the future of cancer detection.
Artificial intelligence are two words you’ve heard countless times, but if Ibex’s successes to date are any indication, those two words might end up saving your life in the very near future.