DECODE – Improving the Clinical Diagnosis of Dementia Dementia, the name for a group of symptoms that include difficulties with memory, thinking, problem solving, language and perception, is a condition that can affect young and old.

Dementia is caused by several brain disorders of which the most common is Alzheimer’s Disease. In the UK, it is forecast that over one million people will be living with dementia by 2025.

The early stages of dementia-causing diseases can be subtle and easily confused with normal signs of aging. To aid non-specialist clinicians in diagnosing the condition, researchers at Exeter University have developed DECODE (DEmentia identification COmputerized DEcision support system), a questionnaire that takes into account a variety of known risk factors to produce personalised feedback for a patient. Having developed an early prototype, they approached OCC for its expertise in the development of medically approved software.

From a Prototype to a Clinical Trial

In order for DECODE to be used as a medical device, it must be shown to be effective in a clinical context. The goal of the work for OCC was to develop the application to the point where the Medicines and Healthcare Products Regulatory Agency (MHRA) , the regulatory body for medical devices in the UK, would approve it as safe for use in a clinical trial by issuing a “notice of no objection”. This meant adhering to strict standards of code quality, documentation and testing during software development, in accordance with a standards-compliant quality management system. OCC and the DECODE team partnered with Beverley Scott of BCS Clinical Consulting to achieve this.

Improving the UI

The prototype consisted of a wizard-like sequence of questions culminating in a calculated probability of dementia and advice to the clinician on the next steps to take. To make it suitable for a clinical trial, OCC added a number of new features including a login page, a storage system for completed assessments, a button enabling trial administrators to download a spreadsheet of assessments for analysis, and a variant sequence of questions suitable for use by GPs in a primary care context, as well as improving the design of the results screen to make the final result clearer.

Safety and Security

The application was originally developed as a single page web application using the Angular framework. For the additional features, OCC created a web API based on .NET Core, allowing login and assessment data to be centrally stored and monitored. The key algorithms underlying DECODE are also accessed via an authenticated API, enabling intellectual property to be protected. To ensure that the software met its design specifications, OCC implemented a rigorous testing framework using xunit unit tests for the algorithms and the Cypress framework for simulation of the interactions between the user interface and the back end.

Future Work

DECODE is currently undergoing final tests prior to its submission to the MHRA. In the future, if clinical trials are successful, the application will be CE-marked and made available for general use.