The following research summary has been graciously provided by Jugal Manoj Dave and David V. Coppini of Bournemouth University in the UK. November 2020
Comprehensive risk assessment for Diabetic Neuropathy
J.M. Dave1, V.N. Dubey1, D.V. Coppini2, J. Beavis1
1Faculty of Science and Technology, Bournemouth University, Bournemouth, UK, 2Department of Diabetes, Poole Hospital NHS Foundation Trust, Poole, UK
There are around 415 million people worldwide who are living with diabetes and more than 46% of the people are still undiagnosed. There are many devastating complications of diabetes in which diabetic neuropathy is one of them. Diabetic neuropathy affects nerves and causes early symptoms like numbness, tingling and loss of touch sensation which starts from the big toe and gradually affects proximally if it remains untreated. This project is based on early detection of diabetic neuropathy for which an intelligent system was developed. The system consists of two parts: one is an innovative device named VibraScan which was developed by keeping the Neurothesiometer device as the base device; and, second is a risk assessment tool developed to determine individual risk depending on clinical characteristics of the subject.
As per the literatures, vibration-based devices are considered as the gold standard for neuropathy detection. VibraScan was considered advanced in many aspects compared to existing technology. The advantages are first that it takes less time to perform the test, normally is around 40 to 60 seconds. Second, the automatic interpretation of results in terms of various severity levels based on detected vibration perception threshold (VPT) indicated like the traffic-light classification i.e. Green light represents low risk, yellow as medium risk and red as high risk neuropathy. Finally, and most importantly, subject can comfortably run the device without a need of an operator since device software ensures the placement of the feet at the required position and monitors the pressure throughout the test using a pressure sensor.
The risk assessment tool is based on historical data and so to upgrade it, there is a need to collect more patient data. The larger the dataset, the more accuracy and precision can be achieved. For collecting more data, VibraScan can be used to determine current VPT of the patient. In this way, VibraScan can be tested on patients for a further trail. And, by using the risk assessment tool on same patients, the clinical data can be collected from the backend in the database to form the data. In this way larger datasets can be collected to improve accuracy using neural network model. Hence, the tool can be updated in the timely fashion and people can use this tool as the mobile application without the need of any device to determine the risk level of developing diabetic neuropathy. This is still a work is in progress.Publications
1. Dave, J.M., Dubey, V.N., Lowes, V., Beavis, J. and Coppini, D.V., 2018. A ‘smarter’ way of diagnosing the at risk foot: Development of a novel tool based on vibratory measurements in subjects with diabetes. In: Diabetes UK Professional Conference, 14-18 March 2018, London.
2. Dave, J.M., Dubey, V.N., Coppini, D.V. and Beavis, J., 2018. VibraScan: A smart device to replace Neurothesiometer for measuring diabetic vibration perception threshold BioMedEng18, (ISBN 978-1-9996465-0-9), Page no.200.
3. Dave, J.M., Dubey, V.N., Coppini, D.V. and Beavis, J., 2019. Predicting diabetic neuropathy risk level using artificial neural network based on clinical characteristics of subjects with diabetes. In: Diabetes UK Professional Conference, 6-8 March 2019, ACC Liverpool.
4. Dave, J.M., Dubey, V.N., Coppini, D.V. and Beavis, J., 2019. Comprehensive risk assessment of diabetic neuropathy using patient data BioMedEng19.
5. Dave, J.M., Dubey, V.N., Coppini, D.V. and Beavis, J., 2020 VibraScan: A platform-based tool for the assessment of diabetic neuropathy DFSG virtual conference.
6. Dubey, V.N., Dave, J.M., Beavis, J. and Coppini, D.V., 2020. Predicting diabetic neuropathy risk level using artificial neural network and clinical parameters of subjects with diabetes. Journal of Diabetes Science and Technology.