Open Science and Reproducibility: Empowering Research Through Innovation and Data Sharing
To advance transparency and reliability in cognitive science, I developed a custom software platform that not only facilitated my own study design, cognitive assessments, and working memory training, but is also openly shared to support reproducibility across the field. Reproducibility is fundamental to scientific progress—it strengthens the credibility of findings, enables others to verify and build upon results, and increases trust in research outcomes. By making both the software and research data available, I help ensure that other scientists can replicate methods, validate results, and accelerate discovery, ultimately reducing errors and misinformation. Innovative, open-source tools like this platform lower barriers to entry for rigorous experimental design, foster collaboration, and maximize the impact of research by enabling efficient data sharing and secondary analyses. This approach not only enhances the integrity and efficiency of scientific inquiry but also contributes to a more connected and trustworthy research community.
DS
All studies conducted on this platform have received ethical approval from the relevant Ethics Committee of the Department of Psychological Sciences at Birkbeck, University of London. All research involving human participants or sensitive data undergoes ethical review and receives approval before any research activities begin. This process ensures that all studies adhere to established ethical principles, including participant protection, informed consent, and data confidentiality, in line with Birkbeck's research governance framework.
NeuxScience® is a cloud-based software platform designed for cognitive assessment and working memory training, developed to advance research and clinical practice in cognitive rehabilitation. Below is a detailed breakdown of its purpose, architecture, and capabilities:
Core Purpose
NeuxScience® aims to:
1. Assess Cognitive Abilities: Measure working memory, attention, and other higher-order cognitive functions through gamified tasks.
2. Deliver Adaptive Training: Use machine learning to personalize working memory exercises based on real-time user performance.
3. Support Research and Clinical Care: Enable researchers and clinicians to analyze how cognitive training impacts transfer of learning, emotional regulation, and functional outcomes in populations with neurological or psychological conditions (e.g., ADHD, dementia, TBI).
Technical Architecture
The platform is built on a modern, scalable tech stack:
Frontend
- React/TypeScript for a responsive, user-friendly interface.
- Pixie 2D engine to create dynamic, gamified experiments and tasks with declarative programming.
Backend:
- Azure Cloud infrastructure for global scalability and reliability.
- Kubernetes for container orchestration, ensuring efficient resource management.
Database:
- Neo4j graph database to model complex relationships between users, tasks, and behavioral data (e.g., response times, movement patterns, task progression).
Key Feature
1. Granular Data Capture:
- Tracks every user interaction (clicks, response times, errors, movement trajectories).
- Logs emotional and environmental variables (self-reported mood, task context).
2. Knowledge Graph Integration:
- Stores data in Neo4j as interconnected nodes (users, tasks, sessions) and relationships (e.g., "performed," "improved," "struggled_with").
- Enables complex queries to uncover patterns (e.g., how anxiety levels correlate with working memory performance).
3. Machine Learning-Driven Adaptation:
- Algorithms adjust task difficulty and feedback in real time to optimize engagement and cognitive gains.
- Predicts individual training outcomes based on historical and population-level data.
4. Interoperability:
- Supports integration with neuroimaging tools and clinical assessment protocols.
- Exports data in formats compatible with R, Python, and BI tools for reproducible analysis.
Clinical and Research Applications
Neurological Populations:
- Traumatic brain injury (TBI) patients: Track recovery of attention and decision-making.
- Dementia: Monitor cognitive decline and test interventions to slow progression.
Mental Health:
- ADHD: Develop personalized working memory training to improve focus.
- Schizophrenia: Assess how cognitive training impacts emotional regulation.
Research Use Cases:
- Study transfer effects (e.g., whether working memory gains improve real-world problem-solving).
- Model biopsychosocial factors (e.g., how sleep quality or social isolation modulates cognitive performance).
Differentiation from Traditional Tools
NeuxScience® diverges from conventional cognitive testing software by:
Leveraging Gamification: Increases user engagement through game-like mechanics.
Prioritizing Data Depth: Captures micro-level behavioral data (e.g., mouse movements) for richer insights.
Emphasizing Flexibility: Researchers can rapidly design and deploy custom experiments via its declarative interface.
Enabling Longitudinal Analysis: Graph database structure simplifies tracking changes over time across diverse variables.
Mission
NeuxScience® seeks to bridge the gap between computational neuroscience and practical clinical care, providing tools that are as rigorous for researchers as they are accessible for clinicians and patients. By combining cloud-native architecture with cognitive theory, it aims to advance personalized, data-driven rehabilitation strategies for brain health.
London based? Why not participate in our study?
Location: MERLiN Laboratory, Birkbeck University of London
Interested in research collaboration?
Let's explore Opportunities for Joint Research Projects