- Traditional ultrasound analysis is often manual and subject to human error, requiring significant expertise which may not always be available.
- There is a need for an efficient, accurate, and accessible method to analyze ultrasound images to improve maternal and fetal health outcomes.
- At its core lies the creation of a comprehensive dataset and a suite of deep learning models.
- Comprising over 15,000 high-quality ultrasound images of a fetus at 23 weeks of gestation, this dataset is annotated with diagnostic planes, fetal biometric values, anatomical features, and bounding boxes for various fetal structures.
- Fetal orientation Model That can detect the structure of the foetus. --- Fetal thorax ,Maternal cervix , fetal brain, fetal femur.
- Maternal Health ---- it can detect mother health – safe , Risk , Severe on the basis of Calories, trimester , Sleep time and Heart