Deep Learning for Pattern Recognition of Fetal Growth Using GSA and Frost Filter

Antenatal analysis uses ultrasound to visualize the fetal growth. The accurate pattern recognition is difficult due to noises in the images. In order to improve the pattern recognition accuracy, deep learning for pattern recognition of fetal growth using GSA and frost filter technique is proposed.The DLPRGSFF technique takes the antenatal videos as input for pattern recognition. The antenatal videos are segmented into frames. The DLPRGSFF technique includes three major processes namely preprocessing, feature extraction and pattern recognition. The matching results provide the accurate pattern recognition. Experimental measurement is conducted for analyzing the performance of DLPRGSFF technique against the two state of-the-art methods with different metrics such as pattern recognition accuracy and computational time with respect to a diverse number of images. The results obtained with DLPRGSFF technique is encouraging than other approaches.