Event Details

Underdiagnosis of ADHD in women and improve gender-specific diagnostics in mental healthcare.

January 29 - April 7, 2025
Virtual Event
PSU, Riyadh, Saudi Arabia

WiDS Datathon 2025

The Women in Data Science (WiDS) Datathon 2025, organized by Prince Sultan University in collaboration with Stanford University, challenges the global data science community to develop AI models for detecting ADHD in women using fMRI data. This initiative aims to address the underdiagnosis of ADHD in women and improve gender-specific diagnostics in mental healthcare.


Why This Matters

  • ADHD in women is often underdiagnosed due to differing symptom presentations
  • AI can uncover neurological patterns unique to women
  • Aligns with Saudi Vision 2030 and UN Sustainable Development Goals
  • Opportunity to make real impact in women's healthcare

Workshops Details

Wednesday, Jan. 29, 2025 • 12:00 - 12:30 PM

Introductory Session: Overview of PSU WIDS DATATHON

Presented by Prof. Tanzila Saba, Dr. Souad Larabi, and Dr. Anees Ara.

Wednesday, Feb. 5, 2025 • 5 - 7 PM

Workshop 1: Data Analysis: Preprocessing and Visualization

Learn techniques for data preprocessing and visualization.

Wednesday, Feb. 12, 2025 • 5 - 7 PM

Workshop 2: Machine Learning/Deep Learning in MRI data

Explore applications of ML/DL in analyzing MRI datasets.

Wednesday, Feb. 19, 2025 • 5 - 7 PM

Workshop 3: Model Training & Testing using Python

Hands-on session on building and evaluating models with Python.

Wednesday, Feb. 26, 2025 • 5 - 7 PM

Workshop 4: Model Validation using Python

Focus on validation techniques to ensure model robustness.

Monday, Apr. 7, 2025 • 12 - 1 PM

Workshop 5: Success Stories by previous DATATHON Winners

Inspirational talks and insights from past winners.


Competition Timeline

Registration Opens

December 2024

Datathon Period

January 29 - April 7, 2025

Winner Announcement

Celebration of achievements and announcement of DATATHON 2025 winners.

May 1, 2025 at WiDS PSU Annual Conference

Technical Specifications

Platform

Kaggle (virtual event)

Visit Kaggle

Tools

  • Python (Primary language)
  • TensorFlow/PyTorch
  • Scikit-learn
  • Neuroimaging libraries (NiBabel, Nilearn)

Dataset

Curated fMRI scans from 1,200 women (600 with ADHD, 600 controls)
Pre-processed De-identified

Challenge

Develop AI models to identify ADHD biomarkers in women's brain scans with ≥85% accuracy
Classification Task

Event Gallery