Structure and context in prostatic gland segmentation and classification
MICCAI'12 Proceedings of the 15th international conference on Medical Image Computing and Computer-Assisted Intervention - Volume Part I
| Hi-index | 0.01 |
Computer-aided Gleason grading of prostate cancer tissue images has been in rapid development during the past decade. Automated classifiers using features derived from multi wavelet transform, fractal dimension and other measurements, and using text on forests have shown considerable successes. This paper presents our study on application of cardinal multiridgelet transform (CMRT) to prostate cancer images to extract texture features in the transform domain. CMRT can provide cardinality, approximate translation invariance and rotation invariance simultaneously. With 32 images of Gleason grade 3 and grade4 as a training set and using texture features extracted there from, a support vector machine with Gaussian kernel has been trained to classify grade 3 and grade 4. The leave-one-outcross-validation showed its accuracy of 93.75% and AUC of0.9651. 10 test images of grade 4 showed 100% accuracy.