Abstract
Many studies have reported the limitations of computer-aided diagnosis systems using a single T2-W MRI which include weak texture descriptors and an extensive amount of noise. Therefore, researchers have used multiparametric MRI to improve the performances of their methods. We propose a computer-aided diagnosis (CADx) method for prostate cancer within the peripheral zone using a single modality of T2-W MRI and qualitatively compared our results with some of the methods in the literature. The proposed method was tested based on 418 T2-W MR images taken from 45 patients and evaluated using 9-fold cross validation with five patients in each fold. The results demonstrated a comparable performance with CADx systems using multiparametric MRI. We achieved area under the receiver operating curve (Az) 88% ± 9% and 87% ± 10% for
Random Forest and Naive Bayes classifiers, respectively, while the combined classifier
achieved 91% ±7%.
Random Forest and Naive Bayes classifiers, respectively, while the combined classifier
achieved 91% ±7%.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of the 19th Medical Image Understanding and Analysis Conference |
| Pages | 207-212 |
| Number of pages | 6 |
| Publication status | Published - 15 Jul 2015 |
