Dementia diagnostics using artificial intelligence; computer tool for memory clinics
Project aim: DAILY aims to translate digital technology to clinical practice, to support clinicians in their daily challenge of dementia diagnostics. We develop a computer tool, in close collaboration with clinicians and patients, that transforms all complex diagnostic test results to an easy-to-understand and clinically useful report.
Background: Dementia is a health priority in the Netherlands as well as globally. To provide adequate care in dementia, early and accurate diagnosis is essential. Despite great progress in diagnostic biomarkers, advances in computer technologies and in artificial intelligence, dementia diagnostics to date remains suboptimal in many places: tools are often outdated, variability in diagnostic workup and quality is high, with diagnostic accuracies as low as 40% having been reported . Simple and intuitive tools are needed for enabling better and more uniform diagnostic quality at all levels of healthcare.
Approach: We develop, in co-creation with patients and professionals, a novel reporting module utilizing data from 1) a novel minimal cognitive test battery, 2) prototype web-based cognitive testing at home, and 3) prototype CT brain quantification. The module transforms all complex data and the outputs of artificial intelligence models to an easy-to-understand and clinically useful report. We validate this novel computer tool (cNeuro-light) prospectively in daily practice, focusing on clinician and patient outcomes
Deliverables: With this data the project paves the way for the optimal use of artificial intelligence tools in clinical practice and empowering patients, to take the first step in personalized medicine by translating diagnostic test results directly to the patient. DAILY delivers reports on development, feasibility and acceptability of cNeuro-light and recommendations for optimal use of computer tools in differential dementia diagnostics (Fig 1). This leads to a more efficient, accurate and consistent diagnostic process in local memory clinics and can harmonize care across
Figure 1: Factors identified in the DAILY project that influence the adoption of digital tools in memory clinics, divided into four subcategories: healthcare professionals, technology, human environment, and organizational environment.
The figure displays the different factors identified for each category. All factors are centered around the healthcare professionals, who need sufficient knowledge of the tool, time to learn and become familiar with it, and assurance of their clinical autonomy. The key considerations regarding the tool itself are ease of use, usefulness, relevance, and validity. Furthermore, the use of the tool depends on the specific environment, divided into the human and organizational environment, indicated by the half circles, in which it is used by the professionals. One important environment is the human environment, where important factors include the interaction between the patient and the professional, patient preferences, and the ability to tailor the tool accordingly. The organizational environment is final important environment to consider. The effectiveness of the tool depends on its ability to be flexible and adaptable to fit seamlessly into the clinical setting and its specific workflow and in which the workload is generally high. Finally, providing training, education, and real-time (administrative) support are important requirements to encourage adoption.
