Interfaces and Interactions for Collaborative Decision Making for Dementia Diagnosis

This project develops a platform for Consensus Diagnosis in dementia research: from concept to minimal viable product, with User Experience considerations for non-technical users.

Initial situation

Alzheimer's research has historically focused on white Western populations, limiting understanding and slowing progress in diagnosis and treatment for the majority of the global population.
The Davos Alzheimer's Collaborative (DAC) is working to close this gap through studies in low- and middle-income countries. One such study is underway in India, where patients undergoing eye surgery are assessed through cognitive testing and dementia risk screenings. The goal is to assign a clinical diagnosis – healthy, mild cognitive impairment, or Alzheimer's disease – based on multi-source data.
The process used by DAC researchers to reach a diagnosis is known as Consensus Diagnosis, which makes use of ratings of multiple independent experts, who review the provided data, and come to a conclusion. If no agreement is reached, the experts discuss the case until a definite diagnosis is made. Consensus Diagnosis is especially useful in the diagnosis of diseases without definite biomarkers or other clear tests, such as dementia.
A previous project (IP5) developed a validated design-prototype for a diagnostic interface. This includes a streamlined workflow for clinicians and a web-based proof-of-concept for displaying diverse clinical data (e.g., structured forms, PDF files, and images).

Problem statement / Project goal

Solution developed and its benefits

Implementation

The developed application fully covers the core workflows of the conceptualized online Consensus Diagnosis Platform. It provides role-based access for administrators, moderators and clinicians, and allows them to view and rate each case as their permissions allow.
The platform combines a modern frontend framework (React & TypeScript) with a Java Spring Boot backend and a PostgreSQL database to create a complete web application. Additionally, a dedicated external API for future integrations with real study databases was designed and implemented. The generic approach followed in this design allows for the tools to be used by multiple cohorts with only small developer interference needed, should the structure of the assessments per study location change.

UX Research

To further allow the platform to stand on its own, the User Experience Research direction was chosen to target the customizable configuration of the assessment data structure. After an initial concept-phase, two versions of such a configuration editor were implemented and A / B tested with several non-technical users that provided valuable feedback. Bugs were then fixed in the final integration into the platform, as well as a further concept proposed for the optimal solution to this complex topic.
Having this component within the platform will allow administrators to react to any changes to the data model in the study database – and therefore makes the adoption of the tool to other cohort studies easy.

Further Research

Due to current emergent technologies and little budget on DAC’s side, Artificial Intelligence tool integration is incredibly interesting in this dementia research context. For this reason, additional considerations, along with drawbacks and benefits, have been made into how such an integration might look like for the Consensus Diagnosis Platform, such as providing an AI that can look at the case data and act as an additional clinician in the entire diagnosis process.

Conclusion

This project has further proven the feasibility of adopting the use of an online Consensus Diagnosis Platform in DAC’s ecosystem.
Because the necessary infrastructure was not ready during this project, no integrations to production systems (such as a real study database or authentication management) were done. Despite this limitation, the focus on extensibility and modularity during the development will allow a seamless adoption across DAC’s diverse research cohorts.
The application will be further developed in collaboration with the IIT at FHNW, building upon a solid and modern basis.

Key Terms

• Dementia A group of conditions causing progressive decline in memory, thinking, and daily functioning.
• Consensus Diagnosis A clinical decision agreed upon by multiple healthcare professionals after reviewing all relevant data.
• User Experience The overall feeling, ease, and satisfaction a person has when interacting with a product or service.
• Full-Stack Application This project combines a modern frontend framework (React & TypeScript) with a Java Spring Boot backend and a PostgreSQL database to create a complete web application.
• Usability Testing A method to evaluate a product by testing it with representative users to identify usability issues and improve user experience.

Clients

Davos Alzheimer’s Collaborative (DAC)
a global partnership of like-minded organizations that is mobilizing the world against Alzheimer’s disease
Dr. Irene Meier, Dr. Vaibhav Narayan
https://www.davosalzheimerscollaborative.org

Team