Researchers at UMass Medical School and UMass Lowell are collaborating on a new cyberinfrastructure technology with the aim to allow patients, researchers and physicians to transport and store large quantities of data, including sensitive information, through a secure system.
The system, Flexware, will be pilot tested in a clinical trial to more accurately estimate the caloric intake of obese patients. The project is being funded by a $1 million grant from the National Science Foundation. "We are doing this research because people often have difficulty recording what they eat, the portion size and where they eat," Yunsheng Ma, M.D., Ph.D., associate professor of medicine, UMass Medical School, said in a press release statement. "Accurate estimation of dietary intake, especially caloric intake, is important for assessing the effectiveness of weight loss interventions."
As part of the network cyberinfrastructure for Biomedical Informatics Innovation project, UMass Lowell researchers will create a network cyberinfrastructure for biomedical information flexible co-scheduling middleware engine (referred to as Flexware) that will enable patients, researchers and physicians to transport and store large quantities of data, including sensitive information pertaining to an individual's health, through a secure system, the researchers said.
"This project will have significant impacts on improving the quality of healthcare applications; providing clinical and scientific researchers with flexible and efficient network resource allocation for studying patient behavior; and training the next generation workforce in medical and engineering fields," said Vinod Vokkarane, PhD., associate professor of electrical and computer engineering at UMass Lowell.
As a means to test the network, researchers will also pilot a study that incorporates the Flexware network that considers how people record what and where they eat, as well as portion sizes. The current method of caloric estimation, called 24-hour-dietary recall, is a research technique that asks individuals to recall foods and beverages they consumed in the 24-hours prior to an interview.
However, that does not provide an accurate record of what a patient ate or drank because it relies too heavily upon memory, the researchers said. It also does not prompt people to be aware of what they're consuming. In order to increase the accuracy of self-reported dietary estimates and reduce patient-bias, technology-based enhancements, such as a snapshot or a video of the food intake are needed, they said. Although technology-based estimation of caloric intake is still a challenging task and an open research problem, the new machine-learning-based computing techniques the UMass Lowell team is actively developing will be secure and have huge potential to address this issue and to transform the field, the researchers believe.
The engineers and researchers said that they are sensitive to the photos and videos that will be shared over the network and the possibility that the individual sharing the information could be identified in the images. To address this challenge, they are developing Flexware in such a manner to ensure the privacy of the participant in the study by blurring the human body in the image while he or she records food intake.