JONA2

Investigation of the clinical introduction of automated decision support software for neuroimaging in dementia

Summary

Over 800,000 people in the UK have dementia and this is expected to rise to 1.1 million by 2025. Patient care costs approximately £26 billion a year. Early detection and timely diagnosis for patients improves clinical outcomes, including quality of life, by enabling early signposting of patients to appropriate resources. The NHS spends approx. £50-100M per year on brain imaging to assist in dementia diagnosis but Interpreting brain scans in dementia is difficult. Research studies have developed software to assist in interpretation but current, commercially available, software is not able to combine images from different modalities, does not cover the imaging modalities most commonly used in the NHS (CT scans) and is expensive.

A NIHR Post-doctoral Research Fellowship (RH) has produced prototype analysis software for brain images produced by several different scan modalities (CT, MRI & SPECT) and methods for measuring and compensating for differences caused by the specific imaging scanner. It is able to compare an individual's brain images with features data extracted from a brain image library for many patients with a known diagnosis (confirmed by extended clinical follow-up). This system - JoNA2 - is designed to provide additional information to the clinician to support decision-making. This software has been the major outcome of and preliminary testing as part of this study has been promising.

This grant will enable us undertake the necessary work to prepare an application for further funding (NIHR Invention for Innovation, i4i).

We propose the NIHR funded i4i project work will incorporate several workstreams:

a) User evaluation of the software and machine characterisation procedures in 4 clinical neuro-imaging centres.

b) Establishing clinical effectiveness of JoNA2 software in correctly identifying diagnostic image features in earlier scans obtained from patients who later have a confirmed diagnosis. (NOTE images from these patients will not have been used during the Fellowship project work).

c) Develop a pathway to commercialisation for the software and/or test phantoms.