Israeli smart software harnesses flood of data to manage diabetes care

DreaMed Diabetes’s cloud-based Advisor Pro helps doctors craft or modify insulin plans based on 288 daily glucose level readings

Infographic showing how the Advisor Pro cloud-based diabetes solution developed by DreamMed Diabetes works (Courtesy)
Infographic showing how the Advisor Pro cloud-based diabetes solution developed by DreamMed Diabetes works (Courtesy)

Petah Tikva-based DreaMed Diabetes has developed cloud-based software that uses machine learning to help doctors and healthcare professionals better monitor their patients with type 1 diabetes. The firm has received a CE mark from the European Union, which means it can start marketing its product in Europe.

Type 1 diabetes is a chronic condition in which the pancreas produces little or no insulin, a hormone that allows glucose to enter cells and provide them with energy. When there is a deficiency of insulin, sugar builds up in the bloodstream, causing life-threatening complications.

The condition affects more than 90 million people worldwide and the global market for diabetes-related products is expected to reach $20.7 billion in 2022, an increase of more than 60 percent over 2017. Despite global efforts, type 1 diabetes has no cure to date and its treatment primarily focuses on managing blood sugar levels with doses of insulin.

Whereas patients used to be required to prick their fingers to monitor their blood sugar levels via glucose meters, new technology has seen the sprouting of continuous glucose monitors (CGM), whose sensors function round the clock and can generate around 288 glucose level readings a day with no finger pricking needed.

However, along with all the benefits CGMs yield, they also create a new headache for physicians: a flood of previously unavailable data to review in order to determine the best course of treatment for their patients.

To this end, DreaMed Diabetes’ product, Advisor Pro, uses algorithms, machine learning and fuzzy logic — a form of logic in which a concept can have a degree of truth between 0.0 and 1.0 —  to collect all of the data from the various insulin pumps, glucose monitors, and the patient’s food consumption. The software then processes all of this data and suggests insulin plans or changes to insulin plans to the physicians, simplifying and speeding up their work.

“As technology has changed so have the methods to optimize treatment. A decade ago glucose meters were the only way to measure blood sugar. Now with the use of CGM taking more than 288 measurements a day, there is a ton of data that we can aggregate together,” said DreaMed Diabetes CEO Eran Atlas.

Not all healthcare providers have the expertise or time to analyze the vast amounts of data available to determine optimal insulin treatment for their patients, said Atlas.

“Our mission is to simplify treatment,” Atlas said, by providing “intelligent, personalized diabetes decision support solutions.”

Infographic demonstrating how the Advisor cloud-based diabetes solution works (Courtesy)

Advisor Pro is the only diabetes decision support solution that has received regulatory approval for optimizing insulin therapy based on continuous glucose monitoring data, he said.

Pilot studies performed at the Schneider Children’s Medical Center of Israel found that the recommendations made by Advisor Pro were nearly identical to the changes in insulin therapy made by physicians.

DreaMed Diabetes is now conducting a multi-center clinical study in the US, Europe, and Israel in partnership with Glooko, a leader in mobile and web applications for diabetes, to evaluate Advisor Pro. Results of the trials are expected by the end of 2018 with an eye to getting FDA approval soon after, Atlas said.

The company said it is planning to gradually launch the product in Europe this summer.

Founded in 2014, DreaMed Diabetes has been developing solutions for people with type 1 and type 2 diabetes. The company’s first product, GlucoSitter, was  licensed to Medtronic, a US medical devices manufacturer.

Most Popular
read more: