MyDiabetesIQ

This project brings together a unique partnership of niche expertise across NHS,
academia and the commercial sector to create a new and world leading diabetes
artificial intelligence product (MyDiabetesIQ) driving tailored decision-support for
clinicians and patients. This project aims to design, develop, test, implement and
evaluate MyDiabetesIQ in Scotland and NW London to demonstrate its
commercial potential.
Focus: Scotland is world-renowned for diabetes care, driven by a unique national
data-linked complete shared electronic health record for clinicians (SCI-Diabetes)
and patients (MyDiabetesMyWay (MDMW)). MDMW has already been
commercialised through a university spin-out company, MyWay Digital Health Ltd
(MWDH), and there is huge potential to build on this success. Computer aided
learning in this project will focus around key areas of diabetes care including i)
prediction of medicine response in individual patients: to improve prescribing,
reduce waste and side effects ii) prediction of chronic complications (including
lower limb/ cardiovascular disease), iii) prediction of acute complications such as
hypoglycaemia (low blood glucose) and ketoacidosis to enable early intervention,
and iv) prediction of diabetes type (e.g. type 1, type 2, genetic diabetes (MODY),
latent onset autoimmune diabetes) to prevent misclassification, and thereby
supporting accurate and safe treatment.
Innovation This proposal is highly innovative; it moves beyond current
approaches using algorithms. MyDiabetesIQ will aim to combine a wide variety of
data sources i.e. NHS data (laboratory, anthropometrics, medical history, medications), and patient recorded data (activity tracking, home glucose
monitoring, insulin pumps and self-reported mood, outcomes, experience and
activation/ motivation scoring) as available from large national datasets.
MyDiabetesIQ will facilitate prediction of multiple diabetes-related outcomes
creating a sophisticated learning health system for diabetes. The outputs will
drive practical evidence based advice via individual and population based alerts,
reports, and dashboards, to clinicians and patients, breaking AI out of the realms
of academia into everyday practice, with the potential to re-model for other longterm
conditions. MyDiabetesIQ will feed into commercial products
MyDiabetesMyWay (patient portal) and MyDiabetesClinical (clinician portal)
serving a market demand for cost saving digital health interventions in a climate
of high costs and rising disease prevalence.