Hi-tech system can find life-threatening conditions years before you get it

High-speed algorithms and artificial intelligence are keys to early detection of diseases.

DECEMBER 22, 2016 23:24

Updated: DECEMBER 26, 2016 11:53

Medical early sign (photo credit: Courtesy) — Medical early sign

(photo credit: Courtesy)

Results from a successful year-long implementation of Medial EarlySign’s groundbreaking tool to identify colorectal cancer in the early stages were published by Maccabi Healthcare Services on Wednesday.

The company, based in Kfar Malal in central Israel, used artificial intelligence and mathematical algorithms to scour and analyze hundreds of millions of medical records.

That analysis allowed them to detect early signs and predisposition to medical conditions such as colon cancer, Medial’s co-founder and chairman Ofer Ariely told The Jerusalem Post.

The company’s first clinical run was done in partnership with Maccabi Healthcare Services, which used Medial’s predictors to successfully identify 111 findings that required intervention, including 11 colorectal cancer cases last year.

“The traditional medical system only reacts to visible symptoms and treats you after identifying existing symptoms.

Our solution is prediction- based. A doctor tells you what you have, we tell you what you are highly likely to get in six months, two years, five years, or... in such an early stage that doctors can’t identify symptoms yet,” Ariely said.

Medial was founded in 2009 by Ariely, Ori Geva and Nir Kalkstein, who used the proven concept of predictive algorithms as the basis for their technologies. Kalkstein was co-founder of Israel’s pioneering, high-speed algorithm- based financial trading firm, Final. He took the basis of Final’s success – the algorithms – together with his financial resources, and invested them in healthcare.

Medial’s first operational platform, ColonFlag, was developed in partnership with Maccabi. The system was embedded into Maccabi’s database, with access to years of digitized medical records from more than 2 million clients.

The system received the records in an encrypted and anonymous format to protect privacy.

The practice run started in 2015. Since then it has yielded 38% of diagnostic findings that required medical intervention.

Within the year, ColonFlag found 100 individuals in the precancerous stage and 11 with active colon cancer.

None of the 111 individuals had received a colonoscopy or been tested for blood in stool. Without Medial’s screening, they would have had no way of knowing they had cancer or were at risk until they suffered signs and symptoms of the disease.

The system works by analyzing millions of records, then cross-referencing them with computerized medical knowledge, known trends and similar cases to produce a patient-specific prediction.

Moreover, it utilizes “machine-learning” artificial intelligence that learns from accumulated data and past results and turns that knowledge into information about what actions are needed.

Since the first trial with Maccabi, Medial has begun using its platform in the United States, the United Kingdom and Poland.

“When we ran the platform on English medical records we also reached the same results.

The English are different from Israelis in genetics and in nutrition and environment, and we still reached the same results. We ran in the US and got the same results. In Poland, also the same results. The algorithm works everywhere the same way,” Ariely said.

Medial has developed predictors for 100 different conditions such as colon cancer, stomach cancer, diabetes, sepsis and organ dysfunction, he added.

“We developed 100 different predictors for 100 different diseases and medical conditions that cause the most suffering, cost the most, and can result in death. Right now only the ColonFlag predictor is in use [by medical providers], as the deployment of these platforms still needs to go through several hurdles, mainly, that doctors are used to working in a certain way, and this system is entirely different in nature to traditional medical practices,” Ariely said.