Duke Digital Health Science Digest – Issue 71
In this issue of the Digital Health Science Digest: Machine learning and breast cancer, a new schema for assessing web-based weight loss interventions, and more!
The Digital Health Science Digest is a bimonthly newsletter compiled by Duke Digital Health. We bring you the most interesting research publications, policy news, and other fun digital health science stuff.
More here: dukedigitalhealth.org
Subscribe here to get these digests direct to your inbox!
Like it? Click here to forward this email to a friend
The University of Nebraska Medical Center is launching the PREDICTS study (Preventing Diabetes with Digital Health and Coaching for Translation and Scalability), which aims to assess the effectiveness of a weight loss and HbA1c reduction program among 500 patients. The program is built on Omada Health’s platform and is modeled after the National Diabetes Prevention Program.
A new project at MIT, Massachusetts General Hospital, and Harvard Medical School is testing artificial intelligence (AI) and machine learning to improve detection and diagnosis of breast cancer. When tested on 35 high-risk lesions, the model correctly diagnosed 97% of the breast cancers, thus reducing the number of unnecessary surgeries by more than 30%.
Researchers (including our very own director, Gary Bennett) examined 49 randomized controlled trials on eHealth interventions for weight loss in order to characterize the use and function of eHealth tools. Six descriptive domains were defined, relating to whether and how information from the eHealth tool was utilized by interventionists.
Results from a recent study among 14,000 adults who access the Cardiogram app on the Apple Watch demonstrate some early proof-of-concept for using wearable data on resting heart rate to predict risk for diabetes. Using Deep Heart, Cardiogram’s “deep neural network,” algorithms were able to predict with 85% accuracy who had diabetes and who did not.
A recent 3-month pilot study tested the mPulse Mobile interactive text messaging solution to improve medication adherence among partially adherent and non-adherent Medicare patients of Kaiser Permanente Southern California. The results sowed a significant difference in medication refill rates; those in the text messaging group had a higher refill rate by 14 percentage points.
On Wednesday, Senate leaders announced a bipartisan budget agreement that included a $2 billion increase in funding to the National Institutes of Health. It also included funding of $6 billion for the opioid crisis over the next 2 years.
ICYMI, towards the end of 2017, the Centers for Medicare and Medicaid Services (CMS) announced it would not cover virtual coaching platforms in the Medicare Diabetes Prevention Program, saying the mHealth components aren’t reliable enough to warrant reimbursement. Of note, this quote: “CMS has long maintained that virtual programs that focus on self-reported weight loss won’t be covered because ‘self-reported weight loss is not reliable for the purposes of performance payment.'”
The LiveWell Rehabilitation Engineering Research Center is sponsoring an app development competition for middle, high school, and college students as well as non-professional developers. The app challenge is intended to promote awareness of the need for assistive and accessible technologies by fostering innovation. Submissions are due May 1. More information and guidelines here!
Apply now to become a MedTech Innovator for 2018! Applications due February 16th.
Digital Health Rules & Regulations: An Overview of FDA Guidelines
This seminar is open to Duke faculty, staff, and students, and will give an overview of the FDA’s approach to regulating mobile medical apps and software. Participants are encouraged to bring a mobile device to participate in case scenarios.
Tuesday, March 20th 9-10am
Duke North 2003
Sign up here!
Digital Health Symposium, “Digital Health Everywhere”
Hosted by UNC’s Health Sciences Library
Friday, February 23, 8:30am – 3:00pm
UNC Health Sciences Library
More information, agenda, and registration link!
If you’d like to submit an event or article for us to publish in our digest, please send at least 3 weeks before the event to: email@example.com