Successful Projects for the Spark Health Digital Innovation Programme 2020

Dr Ehsan Vaghefi – Toku Eyes
“THEIA – mobile diabetic retinopathy screening AI”

Diabetes and its vision-associated complications are one of the most common causes of blindness in New Zealand. The project objective is to increase the reach of THEIA™, a Diabetic Retinopathy (DR) and Diabetic Macular Oedema (DMO) artificial intelligence by pairing it with portable fundus cameras through the AWS cloud platform. Using this platform, patients at risk of developing DR or DMO will be quickly identified and referred to specialist centres and receive treatment months or years ahead of the current process.

Dr Martin Paul Than – Canterbury District Health Board
“Wayfind – data drive clinical decision support pathways”

Wayfind, a next generation decision support tool, which can be linked to and used in conjunction with the HealthPathways clinical guidance system. Wayfind advances health pathway guidance by electronically capturing key predictors to calculate and display risk prediction and guidance for clinicians in real-time. The project will use incoming data to inform risk prediction model development such as a Neural Network model for predicting the risk of a heart attack. Alesha Smith – Airmed Ltd “Predicting services required at emergency medical incidents” One of the biggest challenges for emergency medical services (EMS) is to dispatch the most appropriate resource to provide the right care at the right time. A machine learning approach for air ambulance prediction on an EMS dataset may help to develop a decision support system for dispatchers. The objectives of this study are to develop real-time reporting to support EMS which should positively improve patient outcomes.

Randall Britten – Auckland District Health Board
“Automated sleep study review using artificial intelligence”

Obstructive sleep apnoea (OSA) is characterised by repeated episodes of complete or partial obstructions of the upper airway during sleep, and if untreated reduces quality of life. Patients typically have to wait 8 months before being diagnosed and treated for OSA because the current diagnostic methods require time-consuming manual review of sleep data. The objective is to develop an AI tool that will assist experts in OSA, so that people who suffer from OSA receive the best care sooner.

Dr Ehsan Vaghefi – Toku Eyes
“AI-enabled laser biometer”

Early detection of and active treatment of myopia in young children can significantly reduce the risk of visual loss. Toku Eyes® has developed a new technology to address the urgent need to develop an inexpensive tool to detect myopia in young children. The objective is to train a neural network to detect and then remove the effect of children’s random eye movements from the analysis and then build this capability into our current OLB prototype, which will accurately and inexpensively diagnose and guide the treatment of pathological myopia in children.