Israel is making a big bet on big data in medical research and treatment. There are dozens of startups—and large companies—that are parsing anonymized medical data, outcomes and results for patients, and patient data for clues on how diseases develop, how they can be treated more effectively, and how they can be prevented.
To that end, the Israeli government is investing some $300 million in a big data digitization project, to make anonymized data available to researchers, pharmaceutical companies, and medical institutions.
It is often said that data is the new oil. In the case of medical data, it is an unbelievable resource that can be used to save lives and enable personalized medicine. Electronic health records, when combined with big data analytics and machine learning, can provide insights into patient conditions, treatment possibilities, and likely outcomes—enabling doctors to even predict, and subsequently intervene before a condition or illness appears.
Systems like these could examine thousands of records of patients suffering from a specific condition—say, cancer—to determine commonalities that could indicate the best ways to treat them. Such a system would look at thousands of data points and analyze them, building treatment models that would be tailor-made for specific patients. The system would ensure that patients get the most effective treatment possible, enabling staff to more efficiently treat them.
I've seen how this works in the field, as a family physician. A patient with advanced colon cancer arrived at my clinic, and I wondered if I could have predicted this. I decided to check this patient's routine blood tests and saw the hemoglobin levels for this patient decreased throughout the years, however, they never reached below the medically desired levels.
At Maccabitech—the research and innovation Institute at Maccabi—we conducted research on other patients with the same cancer who did not have anemia and found that for 1,074 patients, the hemoglobin levels had decreased for three years. This was a strong indication this trend of decreasing hemoglobin levels could help predict patients at risk for colon cancer.
These findings led me to realize that we needed a tool that allows doctors to identify the risk of developing this disease in advance, especially in 30% of the population who did not perform a colonoscopy or fecal occult blood test for 10 years. Based on those findings, we developed, together with Medial EarlySign, a startup company with brilliant mathematicians, an algorithm that can predict the onset of colorectal cancer based on a normal blood count, which in Israel 80% of people get done in any case.
This tool is implemented today in Maccabi, basically running in the background. When a high "colon score" result is found, the doctor is alerted and refers the patient to a colonoscopy. In the last two years, thanks to this development, we have found over 300 people with colonoscopy findings who would not have been discovered otherwise. If this tool was based on checking a few data points in a relatively small sample of patients, imagine what we could do with tens of thousands of patient records, analyzed for thousands of data points.
Since then, we have expanded our research to various types of data. One example is voice analysis. We work with a company called Beyond Verbal who is using big data to match vocal biomarkers in recordings from chronic patients to identify cardiovascular conditions. The technology extracts various acoustic features from a speaker’s voice, giving insights on personal health with the potential to revolutionize conventional patient care.
Another example of the use of big data to enhance care is K, a mobile app that enabled patients to receive personalized search results based on patients like them. K utilizes massive amounts of data collected from hundreds of millions of anonymized doctor visits in Maccabi over the last 25 years and offers users a personalized and highly reliable reading of what their symptoms might be saying. To use it, the patient answers a series of simple and clear questions, and the app measures the answers against similar cases collected during general practitioner visits.
One of the most innovative uses of big data in medical tech today is the collaboration between Israel's Maccabi HMO's MK&M Big Data Science Institute and IBM.
In this project, we aim to assist doctors in automatically identifying breast cancer using mammography images from the HMO's database, essentially “teaching” computers how to identify breast cancer findings in those images. Based on vision technology, the system analyzes millions of images and looks for markings, lesions, and other features that could indicate the presence of cancer.
This system can assist doctors in deciding how to treat patients, and eventually to create a predictive analysis algorithm that could actually help prevent breast cancer. Today, breast cancer comprises nearly 30% of all diagnosed cancers in women across the globe. While much work has been performed to facilitate breast cancer detection, including advanced computer-aided diagnosis tools for mammography images until now, little or no work has been done in the creation of advanced decision-support tools that combine multi-modal image analytics and clinical data analysis.
The enormous amount of medical data that has accumulated can be used for scientific breakthroughs, due to ever-growing storage capabilities and data analytics technologies. The data can be cross-referenced with other data, thus predicting diseases before they occur—and enabling us to intervene.
These systems and others have the potential to eliminate many diseases, provide effective treatment for many others, reduce costs, assist doctors and alleviate human suffering. And the technology to do all this exists right now. What are we waiting for?
Varda Shalev, M.D., MPH, is the head of the Morris Kahn and Maccabi Research & Innovation Institute.