Among the ten companies that completed the sixth cycle of the IBM Alpha Zone accelerator, several were engaged in various aspects of artificial intelligence. One of the companies is Medivizor which tries to deal with the mountains of scientific articles that come out every day for every possible symptom
This week, the Rabin Museum in Tel Aviv hosted the closing event of the sixth cycle and the opening of the seventh cycle of the IBM Alpha Zone accelerator, a 20-week program during which IBM helps start-up companies build solutions for the corporate market, interested in a future business partnership with the blue giant.
Of the ten companies that completed the accelerator's sixth cycle, several were engaged in various aspects of artificial intelligence. One of the companies is Medivizor which tries to deal with the mountains of scientific articles that come out every day for every possible symptom.
The CEO of the company, Tal Gaboli, former chief scientist of Amdocs, explains in a conversation with the science website that he and his friends, including Prof. Steve Kaplan, the chief medical director, realized that the intelligence of computers should be connected in order to improve the results of medical treatments.
"We realized that no doctor is able to read and keep up to date, so we built a system based on artificial intelligence. The patients tell us about their medical condition, medical history, medications they are taking, and of course the symptoms, and the system explains the patient's condition in a vernacular language, and recommends to the doctor a personalized, relevant treatment. The system uses artificial intelligence to review thousands of research articles, clinical trials, medical guidelines and various warnings, and derives patient-specific explanations from them. The system updates the patient and his doctor about current clinical trials, treatment options and research findings suitable for the patient's medical conditions. We also use the system to match the right patients to medical trials. "
"Personalization is the most important element. Every year over a million scientific articles in the field of medicine are published in the world - only for people who already have a diagnosis. If God forbid you are diagnosed with any disease, you wonder what the relevant science is, are there new treatments that my doctor may not know about? And what is scientifically true, when you can't trust Dr. Google.
How do you check if it is indeed a real study?
Givoli: "We rate the quality of each study, according to the journals in which it is published and according to other parameters in the analysis of the studies. We reveal this rating to the patients. Of course, we usually only give the most reliable information. In the event that there are not many studies in a certain field, we will present the few studies that were, but we will explain how they are limited.
What is the future of AI in medicine?
"Artificial intelligence has a very bright future in the field of medicine, there is no doubt about it at all. A doctor cannot contain all the required information in his mind and process it. It will not replace the doctor, but can be an aid in any field, starting with the transcription of the meeting with the patient, through the diagnosis and assistance in choosing the operation. It is not a question of if but when. Medicine is a huge information problem. There is no doubt that just as a computer defeated the world champions in chess and Bego, the diagnostic ability of computers will develop. and surpass that of Adam. The augmentation of the human mind by the computer mind is going to move us forward.
How did you train the system?
The system automatically reads thousands of studies, and doctors have taught the system how to decipher them, sometimes it is necessary to show the system a hundred examples of each disease, until you reach a situation where you can rely on the system to give recommendations that deal with human life.
More of the topic in Hayadan:
One response
A system that learns based on only 100 examples - does anyone think that this haphazardness will lead to anything?
Givoli's successes and public relations are apparently the kind of stalks or warning weeds he left behind after many years and huge salaries he raked in at Amdocs.
People don't change. Only the friars who throw money at illusory investments change. a waste
you have been warned!!