The company Diagnostic Robotics, which developed an artificial intelligence system for medical prediction, announced a fundraising of 45 million dollars. Robotic diagnosticsS was founded by Yonatan Amir, together with Dr. Kira Radinsky who became famous at a very young age fromeBay acquired the artificial intelligence prediction company SalesPredict. Another founder is Prof. Moshe Shoham, the inventor of the technology of Mazor Robotics, which was sold for $1.6 billion to Medtronic, and of several other companies. Radinsky is today the CEO of the company, after replacing Amir, who was appointed president.
The investment was led by the StageOne investment fund with the participation of Migdal Insurance, Clal Insurance and Bank Hapoalim and the leading American medical center Mayo Clinic. Private investors Bradley Bloom, founding partner of Berkshire Partners, and Gigi Levy also participated in the current round. In the past, Excelmed, Maverick Ventures Israel, Alpha Capital, Dr. Kobi and Dr. Judith Richter and others invested in the company. To date, the company has raised about 84 million dollars including the current round.
Diagnostic Robotics was established in 2017 to use artificial intelligence to improve processes in hospitals, for example to predict which patients in the queue may deteriorate quickly, or to manage the patients’ visit to the various doctors and tests, in order to optimize the use of resources and shorten the queues.
However, in recent years it changed the model, and it focuses more on medical prediction in the community, rather than in the hospitals. The company has contracted with several insurance companies in the US, and already monitors about 28 million users. It registers revenues that Radinsky calls “significant”.
“The state needs to know the patients and the people at risk”
The company made headlines at the beginning of the corona epidemic, when it sought to use the predictive capabilities through the artificial intelligence it had developed, to predict where new corona foci would develop and what the disease’s spread patterns would be. Radinsky previously dealt with epidemic prediction using artificial intelligence in her academic work. But with Corona, Radinski explains to “LeGoves”: “The need for forecasting was not high. We didn’t need to locate centers of spread because Corona was everywhere, and it was possible to understand the trends by looking at the numbers, even without a forecasting system.” Even the prediction of the spread of monkeypox today is not an interesting story or an interesting market for the system like the company’s, she says. But streamlining the processes of locating patients who can benefit from the drug paxlobide to prevent the deterioration of corona – maybe yes.
This is part of what Radinsky calls “community triage.” The term triage is usually intended to describe the process of prioritizing patients who arrive at the emergency department. But the new approach is to catch them before they even get to the ER. “The state needs to get to know the patients and the people at risk, and address them proactively, to prevent them from deteriorating,” says Radinsky.
At this stage, there is the best option to prevent their deterioration, Radinsky explains. “We base ourselves on information from the patient’s personal file as well as questionnaires that we send to patients from time to time, in order to identify people who are on the verge of deterioration, using our artificial intelligence system that has already seen tens of millions of files. We have shown that we succeed in preventing unnecessary hospitalizations. In some cases, our agreement with companies The insurance allows us to receive a reward that is derived directly from the savings we created, and the insurance companies already recognize the economic value that the company’s tools have generated for them and are already rewarding us for it.”
“We are not looking for the sickest people, but those in the middle”
The company has products relevant to about 25 different diseases, five of which have already gone through the entire validation process, and the intention is to eventually combine them into one predictive tool that takes into account the patient’s health holistically. Meanwhile, the company’s leading product is designed to treat heart failure. The economic cost of hospitalization following heart failure is very high, and preventing the attacks outside of hospitals is possible with simple drugs, if the deterioration is caught in time. A number of companies, including Israeli ones, are developing wearable or even implanted devices to predict the attack. “The problem is the low response of the patients to the wearable devices,” says Radinsky. “Our approach is to send out questionnaires or have nurses call the patients and ask us about symptoms, and we also have one monitoring device – a digital scale. We are not looking for the sickest people, who would even agree to an implanted predictive device, but those in the middle, who live their lives and don’t notice to deterioration but may be damaged by it in the most significant way, and for them prevention is especially important.”
Other products of the company are, for example, a system for detecting endometriosis. The company’s product identifies through the patients’ medical files who may be presenting a clinical picture of endometriosis, and begins sending them questionnaires designed to identify the cyclical nature of the disorder. “Currently, it takes 3-8 years to reach a diagnosis of endometriosis, and our goal is to shorten this significantly,” says Radinsky. Additional products are designed to diagnose COPD, complex mental illnesses, or deterioration of metabolic diseases.
“We see that a patient has received a diagnosis of pre-diabetes and his data are not improving, perhaps we need to hold his hand more. Or a patient with a chronic disease that we see has stopped going to the doctor and taking medication, perhaps he has also developed depression alongside the chronic disease. We are looking for the patients for whom the intervention can to help them. And there are also those who we recognize that this type of intervention will not help them, and then we don’t try either. For example, we can predict that a certain patient has no point in contacting again with messages about nutrition and diet. For him, it is better to invest already in drug treatment.”
“We started the recruitment after the market started to collapse”
Apparently, a family doctor who knows the patient and is in constant contact with him, could have noticed some of these things, and also adjusted the treatment according to his familiarity with the patient’s character. But mainly in the US, where the company’s main activity is, a family doctor who really gets to know the patient in depth is a disappearing world. “Today, all these programs are run by nurses, with some of the activity being digital only,” Radinsky says. “As part of the ‘great resignation’ In the US, there is a lack of nurses. In some clinics even a third of the nurses resigned at once, or the price to hire them has gone up. So our system also tries to optimize the nurse’s time.”
In the USA, community activity is the company’s leading product, and in South Africa and Israel, the other two countries where the company operates, the hospital management products are still more in demand.
Raising 45 million during this period was not a simple matter, Radinski says. “We started the recruitment after the market started to collapse. It was not trivial this time,” she admits. “Many investors have not only become pickier or haggle more, but simply freeze their investments.” Nevertheless, the company finished raising the round quite quickly. “It happened because we are a growing company, which already has receipts in the market,” Radinsky says.