Automatic grading of pathological images of prostate using multiwavelet transform

Histological grading of pathological images is used to determine the level of malignancy of cancerous tissues. This task is done by pathologists. Pathologists are inconsistent in these judgments from day to day and from person to person. So the grades are very subjective and furthermore in some cases this is a difficult and time-consuming task. This paper presents a new method for automatic grading of pathological images of prostate based on the Gleason grading system. According to the Gleason grading system, each cancerous specimen is assigned one of five grades. In our method the decision is based on features extracted from the multiwavelet transform of images. Energy and entropy features are extracted from submatrices obtained in decomposition. Then a k-NN classifier is used to classify each image. We also used features extracted by wavelet packet and second order moments to compare various methods. Experimental results show the superiority of the multiwavelet transform compared to other techniques. For multiwavelets, critically sampled preprocessing outperforms repeated row preprocessing and has less sensitivity to noise. We also found that the first level of decomposition is very sensitive to noise and thus should not be used for feature extraction.