Automatic fetal organs detection and approximation in ultrasound image using boosting classifier and hough transform

In this paper we proposed a system for automatic fetal detection and approximation in ultrasound image. We used Adaboost. MH based on Multi Stump Classifier to detect fetal organs in ultrasound. After fetal organ detected, it is approximated using Randomized Hough Transform. Experiments result show that mean accuracy of the fetal organs detection reaches 93.92% with mean kappa coefficient value reaches 0.854 and mean hamming error reaches 0.032. Proposed method has better performance compared to other five methods proposed in previous researches. Fetal Organ shape approximation performance reaches 81% for fetal head, 57% for fetal abdomen, 72% of fetal femur, and 66% of fetal humérus.