Deployment of a Clinical Data Lake for ADHD: Architecture and Preliminary Insights

Author : Sandra GarciaPonsoda, Javier Sanchis, Juan Trujillo, Alejandro Mate, Miguel A Teruel

Abstract :Attention Deficit/Hyperactivity Disorder (ADHD) generates diverse clinical, behavioral, and physiological data that are typically fragmented across systems. To address this, we present the deployment of a clinical data lake within the BALLADEER project, designed to integrate multimodal information, including EEG, physiological signals, and gamified task performance, into a single repository. The architecture preserves raw data formats under a schema-on-read approach, supports synchronization across modalities via event markers and timestamps, and enforces privacy through pseudonymization and controlled access. Standardized acquisition protocols and quality-control checks ensure data reliability, while a consistent file hierarchy facilitates findability and reuse. The dataset currently includes 164 participants aged 6-18 years and incorporates serious games developed within the project, such as Attention Robots and Attention Slackline. At the time of writing, the system is being extended with a new task, Attention Mistakes, demonstrating its adaptability. This deployment establishes a scalable foundation for future Lakehouse extensions, federated deployments, and advanced analytics to support biomarker discovery and decision support in ADHD research.

Keywords :Data Lake, Big Data, Healthcare Analytics, Machine Learning, Attention/Deficit Hyperactivity Disorder.

Conference Name :International Conference on Distributed Systems and Applications (ICDSA-25)

Conference Place Budapest, Hungary

Conference Date 15th Oct 2025