Delivery Date AI uses advanced artificial intelligence, specifically deep learning algorithms like convolutional neural networks and transformers, to analyze ultrasound images. The AI has been trained on over 8 million ultrasound images, allowing it to identify complex pixel patterns associated with gestational age and preterm birth risks. When an ultrasound scan is performed, the images are instantly sent to Delivery Date AI’s cloud-based platform, which processes them in real-time and generates a predicted delivery date. The system integrates seamlessly with existing ultrasound equipment, requiring no additional hardware, and delivers
results within seconds.
Unlike traditional fetal biometry, which relies on specific measurements of fetal anatomy, Delivery Date AI utilizes a more comprehensive approach:

• Pixel-level Analysis:
  ▪ Examines individual pixels and patterns across the entire ultrasound
  image
  ▪ Identifies subtle features and textures not visible to the human eye
• Maternal Anatomy Focus:
  ▪ Analyzes changes in maternal tissues and structures
  ▪ May include cervical length, uterine wall thickness, placental
  characteristics
• Holistic Image Interpretation:
  ▪ Considers the entire ultrasound image as a complex data set
• Looks for patterns and relationships across different anatomical structures
  • Machine Learning Algorithms:
    ▪ Uses advanced AI techniques to identify correlations between image
    features and pregnancy outcomes
    ▪ Continuously improves predictions based on new data
  • Beyond Standard Measurements:
    ▪ Does not rely solely on traditional measurements like fetal crownrump length or femur length
    ▪ Can potentially detect risk factors not apparent in standard biometric
    measurements
• This approach allows Delivery Date AI to potentially capture a broader range of indicators for preterm birth risk and more accurate delivery date prediction.