AI, indeed, seems to be the next great wave of transformation in healthcare. Countries worldwide are challenged by constant growth, and growing complexity, in their healthcare needs with an aging population and increase in disease incidence. Among the most promising clinical applications of AI is diagnostic imaging, and mounting attention is being directed at establishing and fine-tuning its performance to facilitate detection and quantification of a wide array of clinical conditions. AI carries huge promise for healthcare, with applications that already provide benefits to patients in fields such as telehealth, personalized medicine, screening, cancer diagnostics, and quality control, to name just a few. The pandemic further crystallized the need to accelerate the adoption of new technologies that facilitate telehealth, such as cancer diagnosis done remotely.
IDTechEx expects the market for AI-enabled image-based medical diagnostics to grow by nearly 10,000% until 2040 and the market for AI-enabled image-based medical diagnostics to exceed $3 billion by 2030 across areas such as cancer, cardiovascular diseases, respiratory, retinal, and neurodegenerative diseases.
Pioneers In Cancer Diagnostics
Founded in 2016, Ibex Medical Analytics pioneered cancer diagnostics in pathology and is the first and so far the only company to deploy live AI solutions in laboratories and hospitals across the world, with concrete evidence showing that its own solution helps physicians improve the quality and speed of cancer diagnosis. Simply put, Ibex uses artificial intelligence to help physicians and healthcare providers transform cancer diagnosis. Its team of computer scientists and machine learning experts develops AI algorithms that mimic the work of a pathologist and detect, at a very high degree of accuracy, cancer and other clinical features in tissue biopsies. Such algorithms are then used in labs around the world to analyze live cases and provide an automated, AI-based “second opinion” that alerts when detecting cancers that were missed by pathologists. This way, AI is used as a “safety net” for pathologists and patients and helps in reducing, and nearly eliminating, errors and misdiagnosis.
“Another way pathologists can use our AI technology is as a tool for decision support, a kind of “trusted advisor” that helps them sort through cases and complete diagnoses much faster. With the global increase in cancer incidence putting constant pressure on diagnostics labs, AI could make pathologists considerably more productive and remain focused on the more complex cases, a huge benefit to health systems that struggle to maintain their service levels,” Dr. Daphna Laifenfeld, Ibex’s Chief Scientific Officer shares with me in an interview.
Laifenfeld has dedicated her career to the field of personalized medicine, beginning with her tenure in academia (at the Technion–Israel Institute of Technology and Harvard Medical School) where she focused on pathways underlying neuropsychiatric disorders such as major depression and Alzheimer’s disease. She quickly came to recognize that personalized medicine is inherently about the intersection between medicine, its underlying data, and technology – an approach she implemented in various industry positions and as a co-founder of GenetikaPlus, focusing on personalized medicine in Depression. Daphna joined Ibex after having headed Diagnostics and Personalized Medicine at Teva Pharmaceuticals. Her knowledge, passion, and experience in using technology to progress medical practice for immediate impact on patient lives through AI-based diagnostics are very obvious, and she is one of those people who can explain scientifically complicated things in a very simple way (at least for us without a scientific background, that is).
The Power Of An AI Algorithm
Ibex’s platform is used by pathologists – the physicians tasked with diagnosing various diseases including cancer, who typically work in pathology laboratories found in large medical centers, community hospitals, and the private sector. As Laifenfeld highlights – “an AI system is only as good as the data used to train it”. That’s why Ibex has partnered with Maccabi Healthcare Services, a large healthcare provider in Israel that owns one of the largest digitized clinical datasets in the world. Maccabi’s archives include millions of pathology slides and fully digitized pathology reports – a real goldmine for developers of AI algorithms for pathology.
“We augment the Maccabi dataset with datasets from other pathology institutes and work with pathologists who manually annotate biopsy images. These annotations are used during the training phase, resulting in a model which is then tested on a new set of images and compared to what we call ‘ground truth’, typically determined by other pathologists. It’s an iterative and meticulous process that ends only after rigorous validation with independent pathologists that determine that the algorithm is accurate and meets its performance goals,” she explains.
“At the end of the day, your algorithm relies heavily on the quality of the dataset but more importantly, it’s the know-how and skills of the development team, as well as the methodology they chose, that determine the quality of your AI. One important characteristic of our R&D process is the fact that we engage pathologists throughout the entire product development cycle. They help us adopt the clinician’s point of view on what’s important and how they are able to detect, with their eyes, of course, specific features, for example, cancer in a tissue sample. It’s our understanding of their thought process, which helps tremendously in developing an algorithm that is supposed to mimic their work. We also decided very early on to develop Strong AI algorithms – these are algorithms trained to perform more than one task, or in our field - detect more than just cancer. There are many non-cancerous features that pathologists are trained and actually required to detect and report. Training our algorithms to detect many types of features has improved their accuracy in detecting cancer as well and helped pathologists embrace our solutions more easily,” she continues.
Some of Ibex’s major successes include developing a first solution for detecting prostate cancer, which they deployed initially at Maccabi Healthcare Services. Its AI proved its utility within weeks when it alerted on cancer that was missed by pathologists – the first-ever reported case of a misdiagnosed cancer that was detected in real-time by an AI solution. Since then the company has deployed its AI solution in labs across the world and they routinely detect missed cancers – making sure that the case is reviewed and corrected. Laifenfeld and her team also continued R&D efforts, adding a solution for breast cancer, making Ibex’s platform the first-ever multi-tissue AI solution (for breast and prostate) deployed in routine practice in pathology.
“We are now working on solutions for additional tissue types, expected to hit the market later this year, as well as for new applications for additional workflows in pathology that have already demonstrated significant productivity gains. Finally, we are engaged with several partners on projects for the development of AI-markers for prognostic and predictive applications used in cancer management and drug development,” adds Laifenfeld.
The U.K. is at the forefront of changing cancer care, according to Laifenfeld, and it’s led by ground-breaking, U.K. government-supported digital pathology and AI projects such as PathLAKE+, NPIC, and iCAIRD. This is an opportunity for Ibex to implement clinical-grade AI across fully digitized pathology networks deployed in multiple regions.
“We started off in the U.K. by teaming with LDPath, a London-based provider of digital pathology services to non-less than 24 NHS trusts throughout the majority of the U.K., including large teaching hospitals and district general hospitals. At the same time, we’ve created a partnership with major teaching hospitals, led by Imperial College London, and won a share of a £50 million fund as part of the U.K.’s AI in Health and Care Award, an initiative led by NHSx and the National Institute for Health Research (NIHR). This project, which is rolling out this year, will enable the deployment of Ibex’s AI platform in six NHS trusts and involves researchers from Imperial College London, University College London, University Hospitals Coventry & Warwickshire, and other institutes. It will enable the demonstration of the benefits of broad-scale implementation of AI technology to cancer patients.”
A $38 Million Series B Financing Found
Last week, Ibex announced a $38 million Series B financing round led by Octopus Ventures and 83North, with additional participation from aMoon, Planven Entrepreneur Ventures, and Dell Technologies Capital, the corporate venture arm of Dell Technologies. The investment brings total funding of Ibex to $52 million since its founding in 2016 by Joseph Mossel and Dr. Chaim Linhart. This new funding will help the company meet the growing demand for AI and digital pathology rollouts, support an expanding customer base and grow talent across its teams. “We intend to expand the Galen™ solution portfolio at Ibex, bringing new AI tools for more tissue types, including novel AI-based enhancements of the pathology workflow and oncology-focused AI-markers,” adds Laifenfeld.
“Raising money during a once-in-a-century pandemic is indeed an interesting experience. If you had told me a year ago that we would meet new investors, pitch the technology, prove our business case, develop necessary trust and finally ink $38 million in funding – all without boarding a plane even once, I would have thought you are out of your mind. But necessity is the mother of invention, and we all had to get used to doing things differently. We were confident in our vision and ability to execute on it, and the fact that our team continued working in full force throughout 2020 without major impediments helped us in gaining the trust of our investors. As a side note, it is worth mentioning that the pandemic accelerated the pace of digital transformation, particularly in healthcare. From this perspective, it was easier to make the case for a tech venture that helps physicians with remote work and provides efficiencies,” concludes Laifenfeld.