A neural network approach for classification of placental tissues using discrete wavelet transform

This paper proposes an efficient method for the classification of placental development with normal tissues. The proposed method consists of selection of tissues, feature extraction using discrete wavelet transform and classification of the tissue by the multi layer perceptron. The method is tested for placental images acquired by ultrasound techniques; resulting in 95% success rate. The proposed method showed a good classification rate. The method will be useful for detection of the anomalies those concerning premature birth and intra-uterine growth retardation. (author)