What does Doctor AI say?

After a challenging career in the IDF’s Unit 8200, Ronen Lavi retired and left the army at the rank of Lieutenant Colonel. He describes developments related to obtaining information in the world of intelligence from the beginning of the cyber era until his retirement as “a dramatic, maybe even traumatic process”.

Lavi describes how intelligence personnel encountered copious amounts of diverse information, much more than humans can cope with: “We understood that because there is so much information, we fail to separate the wheat from the chaff and provide decision makers in the field with the relevant information”.

About ten years ago and after forming a think tank to address this issue, a new unit was created with the aim of establishing supportive decision-making systems for all types of data. Lavi was appointed Head of the R&D Division: “We understood that if we want information to keep pace with decisions, it is necessary to remove the human factor from the process and introduce a machine”.

The extraordinary systems developed by the unit, won them the Israel Defense Prize. “We went from being hunters of knowledge to data gatherers with the aim of improving the decision-making process”, says Lavi.

When he reached retirement age, Lavi decided to recruit his longstanding friend and partner from the unit – Shay Perera, an AI expert. The pair made a clear decision: to progress from the world of cyber to action based on the accumulated knowledge which has a genuine impact on people. Lavi and Perera quickly identified the healthcare sector. Their startup (Navina) was established by Lavi in 2018 who serves as its CEO while Shay Perera is the company’s CTO.

As Lavi explains, the unique feature of the medical world is the incessantly increasing quantity of information, while the decision-making process has remained unchanged for decades.

After conducting a field study, the two realized the need for healthcare to be more proactive, accessible, and efficient in order to be cost effective. On the other hand, they also understood that the best area in which to make the change was family medicine – precisely because this field is considered to be at the bottom of the economic food chain. This was because economizing operations and processes in family medicine can dramatically reduce the system’s costs by saving time and hospitalizations.

Lavi explains that the understanding of preventative medicine’s importance also applies to the US market. A situation whereby a family doctor has only ten minutes to talk with the patient, acquaint himself with his medical history, and diagnose and treat him is precisely the point to introduce technology.

“Unlike a doctor who doesn’t remember what happened six months ago, technology does remember”, says Lavi. “Navina says: I will give you a tool to get to know the patient, his medical history, and the latest tests. The system takes all the information in a patient’s medical file and informs the doctor of all the important data in the form of a user-friendly clinical profile of the patient.

“In this way, every visit will become more accurate and effective, leading to better quality treatment. The result is overall satisfaction – for the patient, physician, health system and insurance companies. Navina helps physicians provide better healthcare.

“Providing physicians with tools that they can’t do without is our vision and our biggest challenge”, Lavi continues. He explains that the product’s development is being led by a large team of doctors that has been working closely from the outset with other physicians from clinics in the US to tailor the product to their specific “challenges”. The result is a solution which the physicians view as having value on a daily basis. Indeed, Navina’s usability is especially high compared to the competitors, and as of now, is used by thousands of physicians in the US.   

Lavi describes the physicians who participated in building Navina as genuine partners who joined them in the trenches from the very beginning. Thanks to the thorough understanding regarding the needs and difficulties of the end clients – the physicians – one of the company’s sources of strength is its high level of usability compared to that of Navina’s competitors.

One of the problems that characterizes the American healthcare market is the significant lack of order pertaining to information located in the local health systems. Using AI technology, Navina’s system can successfully identify documents, read a scanned text, distribute the relevant codes of medicines etc.  – and integrates them into a clear picture: the patient’s metrics, the drugs he uses, and results of his lab tests.

“Our product is very actionable”, Lavi explains. “We know how to give the physician reasoned clinical information and to place it within the relevant medical context. It is not enough to show him that the glucose level is high – you need to give him a context. That’s the reason we outshine our competitors who provide information in a summarized rather than updated form”.

In other words, Navina takes all the information in the medical file and highlights only the important data, even showing proposals for diagnoses originating from several data points. “We don’t say what the problem itself is”, Lavi stresses. In practice, Navina quickly reads the results, going back even ten years, and presents the physician with insights and recommendations. The physician can accept or reject the diagnosis, a process which, in turn, enables Navina to learn and improve. As of today, up to 80% of Navina’s diagnoses are accepted, Lavi says proudly, “which proves that Navina is an auxiliary tool and not a substitute for the physician, and his knowledge and empathy”.

Navina has raised 44 million dollars so far, capital which has been invested in developing the AI technology and clinical algorithms, and in accelerating the product’s marketing to clinics throughout the US. Although the current target population is physicians in the US, the company is located in Israel where it operates its development center.

“We have been working with the Innovation Authority since the seed stage and they have always believed in us. The collaboration with them was amazing. During the Covid period, the Innovation Authority operated in an impressive manner. They were the first to provide funding and also succeeded in accelerating processes. This enabled us to work better and with peace of mind.

What about the future? According to Lavi, Navina is currently focusing on offering a solution for 60 million clients who will be insured under the value-based care insurance model by 2030. The ultimate vision is for each of the 340 million patients in the US to receive service based on our system each time they visit the family doctor. “We want to break through beyond the existing worlds so that doctors will say that they can’t start their day without Navina”, he concludes.

The Genomic Revolution Starts Here

Moshe Einhorn, CTO of Genoox and Amir Trabelsi, the company’s CEO, also became acquainted during their military service, this time in the ‘Mamram’ (Center of Computing and Information Systems) Unit’s programming course. After leaving the military, the two joined the cyber world in separate companies – Einhorn in fields of information security and software development and Trabelsi in product management and sales.

“At some point, I wanted to make a professional change and went to study for a master’s degree in bioinformatics at Tel Aviv University”, says Einhorn. “That opened my mind to an amazing world I hadn’t encountered before. At this point, we decided to establish Genoox together. We both had experience in software and big data, each with his own background and expertise which, together with our academic education, gave us the push towards the field of personalized healthcare and, specifically, genomics”.

The genomics revolution occurred about twenty years ago when the human genome project was completed”, says Einhorn. “The project cost billions and took over a decade but ultimately unlocked the map of the human genome. However, while we now know how to read the genetic sequence data, we have not yet fully deciphered it. Although we are all 99% genetically identical, there are still millions of variants that make us different.

The ability to check the DNA of each patient enables a new kind of medicine: genomic medicine. If, in the past, patients with a certain disease would all receive the same treatment and drugs regardless of their personal genetic structure, today’s personalized genetic healthcare is something that happens daily in a variety of clinical areas.

Genomic medicine will allow physicians to deduce from the DNA which drugs and treatments will help each patient, how to reduce their side effects and how to make them safer and more efficient.

That’s where Genoox comes into the picture. The platform the company developed enables to improve patients’ treatment by transforming genetic data into practical and valuable information. Anyone can in fact receive personalized treatment according to their own clinical background according to the genetic profile. This can now help in a very diverse range of clinical applications for rare diseases, oncology, and screening tests and will later be applied to complex diseases such as diabetes, Alzheimer’s etc.

Let’s take for example a child who, at a very early age, develops neurological symptoms, a serious developmental delay or other symptoms that could indicate a rare disease with a genetic background. In cases like this, the child will be sent to a genetic sequence lab to identify the mutation causing the disease.

Genetic sequencing technology enables us to generate DNA from a blood or saliva sample and to extract the raw genetic sequence. The laboratory performing the sample’s sequencing needs to analyze it i.e., find the variant causing the clinical problem which is like looking for a needle in a haystack.

The raw genetic data includes millions of variants and deciphering this data by a geneticist or by a clinician is a Sisyphean process that takes valuable time and is susceptible to errors and oversights.

Genoox’s system – the main component of which is an AI engine – can receive the genetic data together with the clinical background, rank all the findings, and present them in an orderly and understandable way within just a few minutes. By doing so, it enables medical professionals to perform a quicker and more accurate diagnosis and to use genetic analyses as a simple, cheap daily tool.

Beyond the human benefit, genetic tests can save extremely significant sums of money for the health system by enabling to diagnose the problem, save expensive and unnecessary treatments and, of course, to save lives. 
Cancer is another field in which genomics is a revolutionary tool. Today, the lab can take a biopsy from a tumor, sequence its DNA, and examine its genetic profile. Analysis of the genetic profile using Genoox’s system allows an oncologist to recommend pharmaceutical treatment that will help the patient and avoid those that will actually exacerbate his situation. These treatments have recently been introduced to the state-subsidized basket of health services and the Ministry of Health funds these tests for certain types of cancer.

We all Have a Different Profile

“When we started developing the system, we placed significant emphasis on Artificial Intelligence”, Einhorn says, “After a while, we noticed that the genetic labs work independently from each other. This led us to our main differentiator today – a community-based platform”.

The Genoox team built a network that connects between labs that work with the platform and connects the data revealed in each lab. This means that professionals receive more information which is not just limited to their personal patients.

If we continue the example of the child with genetic problems, the genetic physician treating such a case may encounter a mutation but not be convinced if that is what is causing the disease. “What can help him is if he can reach out to another physician who encounters a similar or identical mutation in a patient with similar symptoms or, alternatively, if he encounters someone healthy who has the mutation which means that it’s probably not the cause of the disease.

We are the first and largest professional genetic community in this field”, Einhorn says proudly. “This gives us tremendous power and bolsters genetic physicians’ work”. He emphasizes that, naturally, all data sharing is undertaken in a way that protects patients’ privacy in accordance with regulatory requirements. A further unique feature of Genoox is that it is the only platform today that also enables these analyses on any sequencing platform or genetic test.

Genoox’s target population is genetic clinicians – genetics labs that have a genetic physician or bioinformatic professional, pathological labs that perform cancer analyses, specialist doctors such as oncologists and neurologists, academic researchers and pharma companies. The company is currently working with a Freemium model where some of the features are offered for free and others only available in exchange for payment suitable for large organizations with more advanced needs.

The Genoox system currently operates in more than two thousand organizations worldwide. “We are seeing an exponential growth both in the number of users – over 30,000 monthly users – and in the number of analyses performed. This is happening, among other reasons, because the costs of performing a genetic analysis are constantly decreasing as more tests are introduced to the state basket of health services.

So far, the company has raised almost 20 million dollars from several leading venture capital funds in Israel and overseas. “We work with the Innovation Authority and have received several grants”, Einhorn says. “The R&D grant is what enabled us to build the community and many of its capabilities”.

One of the future challenges identified by Genoox is the integration of genetic data as an inherent part of developing pharmaceuticals. “Our current focus as a company is to expand collaboration with pharma companies that can take aggregated and anonymized data derived from the community and to make use of this data to develop the next generation of drugs. If, for example, a company develops a drug that suits a certain patient’s genetic profile, and they want to recruit participants for a clinical trial, Genoox can help them to identify patients with an appropriate genetic profile.

“The direction is clear”, Einhorn concludes: “Genetic sequencing will become as routine as a blood test is today. We believe that everyone will have their own unchanging genetic profile according to which it will be possible to offer them medical and healthcare recommendations. Once the tests become standard and accessible, we will be able to turn genetic data into practical available information, thereby improving treatment for everyone”.