AI-Powered Hospital System for Autism Assessment

This project is a HIPAA-compliant hospital software designed to support doctors in assessing and monitoring autistic patients. The system enables patients to complete AI-assisted video-based tests that analyze focus, facial expressions, and other behavioral markers, while delivering structured results directly to doctors for diagnosis and progress tracking.

Alongside the AI-driven assessment module, the system also includes electronic medical record (EMR) features, scheduling, secure patient-doctor messaging, and billing integration — making it a complete solution for autism-focused healthcare facilities.

• Role: Design, Development & Maintenance
• Duration: 25 days
• Team Size: 4

Hospitals and clinics often struggle with standardizing autism assessments and integrating them into digital health systems. Traditional methods rely heavily on manual observations, which can be subjective, time-consuming, and difficult to track over time.

The key challenges were:

• Ensuring the platform adhered to HIPAA requirements for patient privacy.
• Developing an AI model capable of analyzing subtle patient behaviors via video.
• Integrating assessments with a full hospital management workflow (records, appointments, billing).
• Building a platform that doctors, patients, and caregivers could all use intuitively.

We built a secure hospital system with an AI-powered autism assessment module at its core. The solution combined machine learning, healthcare-grade security, and user-centered design to create a tool that enhances diagnostic accuracy and streamlines patient management.

Key Features Delivered:

AI-Powered Autism Assessments – Patients take video-based focus tests where the AI model tracks:

• Eye contact duration
• Attention span / focus levels
• Facial expression analysis (smiles, frowns, confusion indicators)
• Head/gesture movements
• Responsiveness to visual/audio cues

• Doctor’s Dashboard – Secure dashboard for doctors to view test results, reports, and historical comparisons.
• Electronic Health Records (EHR/EMR) – Patient history, diagnoses, and prescriptions stored securely.
• Appointment Scheduling – Patients and caregivers can book appointments, both online and in-person.
• Billing & Insurance Integration – Automated invoicing, insurance claims, and payment tracking.
• Secure Messaging – HIPAA-compliant chat between doctors, patients, and caregivers.
• Admin Panel – Manage doctors, patients, roles, billing, and reports.
• Accessibility-Friendly UI – Built with a focus on inclusivity for neurodiverse patients and caregivers.

Within the first 6 months of rollout in a partner hospital, the system achieved:

• 1,200+ autism assessments conducted successfully.
• 30% reduction in doctor assessment time, thanks to AI pre-analysis.
• 95% satisfaction rate among doctors using the dashboard.
• 100% HIPAA compliance audit clearance.
• Improved patient engagement, with 70% of caregivers reporting better follow-up consistency due to digital tracking.

Building a HIPAA-compliant healthcare platform with AI capabilities presented several challenges:

• Problem: Handling sensitive patient data under HIPAA compliance.
• Solution: Implemented end-to-end encryption, role-based access, secure audit logs, and anonymized AI training datasets.

• Problem: Ensuring AI accuracy across different patients.
• Solution: Trained models with diverse datasets, applied transfer learning, and added confidence scoring for doctors to interpret results reliably.

• Problem: Making the system intuitive for caregivers with limited technical literacy.
• Solution: Designed a minimalist, step-by-step UI with accessibility features like large fonts, high contrast, and multi-language support.

• Problem: Integrating AI assessments with traditional hospital workflows.
• Solution: Designed modular APIs to connect the AI module with EMR, scheduling, and billing systems seamlessly.

The platform was built using a combination of robust, scalable technologies to ensure performance, security, and ease of use:

• Frontend: React.js, TailwindCSS, Framer Motion
• Backend: Django REST Framework
• Database: PostgreSQL (for medical-grade reliability & auditing)
• AI Module: TensorFlow + OpenCV (video-based behavior tracking)
• Authentication & Security: JWT, Two-Factor Authentication, HIPAA-compliant encryption standards
• Hosting: AWS (HIPAA-ready infrastructure)
• Other Tools: Docker, GitHub Actions (CI/CD), Figma for design prototyping

Building this AI-powered hospital system was a deeply educational experience that reinforced several important lessons in healthcare technology development:

• Successfully delivered a HIPAA-compliant hospital-grade system.
• Integrated AI-driven diagnostics into healthcare workflows.
• Balanced security, performance, and usability for both doctors and patients.
• Learned advanced practices in data privacy, medical compliance, and AI-human collaboration.

The hospital system proved that AI can augment, not replace, human expertise in healthcare. By combining objective AI analysis with professional medical oversight, it not only improved efficiency but also brought greater accuracy and consistency to autism assessments.