Multiparametric MRI maps for detection and grading of dominant prostate tumors

Abstract

Purpose: To develop an image-based technique capable of detection and grading of prostate cancer, which combines features extracted from multiparametric MRI into a single parameter map of cancer probability.

Materials and methods: A combination of features extracted from diffusion tensor MRI and dynamic contrast enhanced MRI was used to characterize biopsy samples from 29 patients. Support vector machines were used to separate the cancerous samples from normal biopsy samples and to compute a measure of cancer probability, presented in the form of a cancer colormap. The classification results were compared with the biopsy results and the classifier was tuned to provide the largest area under the receiver operating characteristic (ROC) curve. Based solely on the tuning of the classifier on the biopsy data, cancer colormaps were also created for whole-mount histopathology slices from four radical prostatectomy patients.

Results: An area under ROC curve of 0.96 was obtained on the biopsy dataset and was validated by a "leave-one-patient-out" procedure. The proposed measure of cancer probability shows a positive correlation with Gleason score. The cancer colormaps created for the histopathology patients do display the dominant tumors. The colormap accuracy increases with measured tumor area and Gleason score.

Conclusion: Dynamic contrast enhanced imaging and diffusion tensor imaging, when used within the framework of supervised classification, can play a role in characterizing prostate cancer.

Copyright © 2012 Wiley Periodicals, Inc.

Figures

Figure 1

ROC curves for the biopsy data, for different groups of features acquired by changing the decision threshold, Pc, from 0 to 1.

Figure 2

Left: T2-weighted MRI of the mid-gland of a patient with biopsy confirmed cancer in mid-left region. Right: The SVM-based cancer probability map, with hot colors showing higher Pc. The Gleason score: 4+5, the average Pc in the tumor area: 0.9.

Figure 3

The <D> and Ktrans maps for the patient presented in Figure 2.

Figure 4

Cancer colormaps and MRI images with corresponding histopathology slide for case 1. The main pathologic finding is a 3+4 tumor in the transition zone, visible in two consecutive cross sections.

Figure 5

The <D> and Ktrans maps corresponding to the case presented in Figure 4, row 1.

Figure 6

Cancer colormaps and MRI images with corresponding histopathology slide for case 2. The main pathologic finding is a tumor that is of Gleason score 3+4 in one cross section and 3+3 in the neighbor cross section. The hot-spot in the transition zone of the prostate is a false positive detection. This is most likely due to the fact that our SVM is trained only on data from peripheral zone.

Figure 7

Cancer colormaps and MRI images with corresponding histopathology slide for case 3. The main pathologic finding is a tumor that is of Gleason score 4+3 (with considerable tertiary Gleason score of 5) in one cross section and 4+5 in the neighboring cross section.