🧬 AI-Powered Neonatal Diagnostics
Early Detection
for Every Newborn
NeoScan AI combines four deep learning models to detect critical neonatal conditions — from retinal disease to facial syndromes — in seconds.
👁
Eye Analysis
Detects ROP stages 0–3, cataract and leukocoria using EfficientNetB4 trained on 3,105 retinal images.
● Live — 98.38% accuracy
🩺
Skin Analysis
Classifies jaundice, cyanosis and anaemia from skin photographs using EfficientNetB3 with class-aware augmentation.
● Live — EfficientNetB3
🦴
Body Analysis
Detects bone fractures from skeletal X-Ray images using ResNet50 (Transfer Learning) trained on Multi-Region Bone Fracture dataset.
● Live — ResNet50 · 83% accuracy
🫥
Face Analysis
Detects Cleft Lip/Palate and Down Syndrome using VGG16 + ViT ensemble with 68-point facial landmark analysis.
● Live — VGG16 + ViT Ensemble