Queen's University MMAI "AI for Good" Team Project

AidBrain – AI Early Detection

AI for Non-Invasive Early Alzheimer's Detection

Date: June 09, 2022

Overview:

AidBrain is an “AI for Good” initiative collaboratively developed by our team (“Team Dufferin”) during the Master of Management in AI (MMAI) program at Queen’s University. We conceptualized and designed an AI-powered mobile application aimed at the early, non-invasive detection of Alzheimer’s disease by analyzing digital biomarkers captured via smartphone sensors.

The Challenge:

Alzheimer’s disease is often diagnosed at advanced stages, limiting the effectiveness of interventions. Current early detection methods can be invasive, costly, and inaccessible, highlighting a critical need for more proactive and user-friendly assessment tools.

Our proposed solution centers on:

  • Non-Invasive Digital Biomarker Collection: Utilizing smartphone sensors to capture data related to:
    • Audio: Analyzing speech patterns from recorded sentences.
    • Dexterity: Assessing fine motor skills via on-screen writing or drawing tasks.
    • Eye Movement: Tracking gaze patterns during screen-based activities.
  •  AI-Powered Analysis: We envisioned leveraging machine learning, specifically deep learning technology to extract essential features from these multi-modal tasks. These features would then be fed into a classification model to generate a probability of Alzheimer’s risk.
  •  Key App Services:
    • Assessment & Monitoring: Providing AI-driven risk analysis.
    • Recommendations: Offering personalized lifestyle advice based on risk assessment.

Empathy & Insight: The Starting Point

Our Design Thinking process began with in-depth user interviews, leading to the creation of user personas and empathy maps to deeply understand user needs, pain points, and motivations.

Designing AidBrain: From Concept to MVP

Showcasing our user-centered design process, from initial wireframes to defining the Minimum Viable Product based on Design Thinking principles.

Business Model & Roadmap

Detailing AidBrain’s business model, target market, value proposition, and strategic product roadmap for development and growth.

The AI Engine: Data & Modelling

Outlining the core data journey for AidBrain, from digital biomarker collection and preprocessing to the AI modelling approach for early detection.

Impact & Responsible AI

Highlighting AidBrain’s potential societal impact, our commitment to ethical AI practices, data privacy, and mitigating risks like misdiagnosis and discrimination.

Key Learnings & Personal Contributions:

This “AI for Good” project was a pivotal opportunity to apply and further develop my professional expertise in a cutting-edge AI context. My key takeaways and contributions include:

Leveraging Product Management & User-Centricity in AI:

The project allowed me to directly apply my extensive experience in product management and user-centric design, which I’ve honed throughout my career. This involved leading user research initiatives and ensuring the insights gathered were central to our solution design, reinforcing the importance of this approach even in complex AI projects.

Bridging AI with Business Value & Marketing:

My background in marketing and entrepreneurial ventures was instrumental in defining AidBrain’s value proposition, identifying target customer segments, and contributing to the business model. This project confirmed my ability to connect technological innovation with market needs and create a viable business case.

Integrating Ethical AI Considerations:

A significant learning was the deep dive into the ethical dimensions of AI, particularly in healthcare. I focused on ensuring our proposed solution incorporated principles of fairness, transparency, and data privacy, aligning technological capabilities with responsible innovation.

Strategic Contribution & Value Creation:

My prior experience enabled me to contribute strategically to the team, ensuring our efforts were always tied to creating tangible value for users and addressing a real-world problem with a potentially profitable and impactful business solution. This project solidified my understanding of how to drive AI initiatives from concept to a well-defined, market-aware proposal.

Applied Concepts & Key Skills:

  • AI & Data Science: Machine Learning (Classification), Deep Learning (Conceptual), Digital Biomarker Analysis (Speech, Dexterity, Eye-Tracking), Predictive Analytics.
  • User-Centered Design (UCD) & UX: Design Thinking, User Research (Interviews, Personas, Empathy Maps), Prototyping (Wireframes to High-Fidelity).
  • Business Strategy: Business Model Development, Value Proposition Design, Market Needs Analysis.
  • Ethical AI & Responsible Innovation: Data Privacy, User Consent, Bias Awareness, “AI for Good” Principles.