VichitraBodh

📝 Notes

This project was developed to assist early diagnosis of Autism Spectrum Disorder (ASD) using facial image analysis. It was part of a healthcare AI initiative focusing on pediatric disorders.

• Driven by the need for accessible, non-invasive diagnostic tools in early childhood.
• Aligned with current research linking facial morphology with neurodevelopmental conditions.
• Designed to support pediatricians and mental health professionals with a secondary diagnostic aid.
• Developed in a research setting and intended for educational and prototyping purposes only.
• Ethical considerations were addressed, including dataset anonymization and bias evaluation.

📄 Project Description

VichitraBodh is a deep learning-based system that detects Autism Spectrum Disorder from pediatric facial images using transfer learning with VGG16. The model achieved 74.99% accuracy on a curated dataset. It integrates image preprocessing, data augmentation, and fine-tuning strategies for robust performance.