Scrum development process design for AI-driven medical software development in brain health

Abstract

This thesis examines how a mobile health (mHealth) application can be designed and assessed to better support children with Attention-Deficit/Hyperactivity Disorder (ADHD), focusing on ways to improve their quality of life (QoL) through structured caregiver input and closer clinical follow-up. Although digital tools are becoming more common in behavioral healthcare, many ADHD-related apps still fall short. They often lack the usability required for consistent use, miss opportunities for personalization, and fail to fully comply with medical data regulations—challenges that limit their effectiveness in real-world clinical environments.

The objective of this study is to develop a customized Scrum-based development framework for building an ADHD-focused mobile survey tool. The methodology follows the Design Science Research (DSR) paradigm, combining iterative design, stakeholder engagement, and empirical validation. The artefact—a functional prototype—was developed across multiple sprints and evaluated through four rounds of feedback involving clinicians, researchers, and caregivers.

This research makes several key contributions. First, it introduces an agile development process that’s been adapted specifically for the constraints and expectations of medical software. Second, it delivers a working prototype—a mobile application that enables caregivers to collect and share quality-of-life (QoL) data for children with ADHD in a structured, clinically useful way. And third, it offers a practical evaluation model that connects user feedback to ongoing design improvements. Throughout, the work emphasizes both ethical considerations and clinical relevance, laying a foundation that can support future expansion.

While artificial intelligence (AI) wasn’t part of the initial build, the study points toward it as a promising direction for future development. Features like real-time QoL prediction, adaptive feedback, and therapeutic guidance could significantly extend the system’s capabilities. More broadly, the findings show that agile methods—when combined with continuous stakeholder input—can be adapted to meet the demands of regulated digital health environments.