Computer-Aided Detection of Prostate Cancer in MRI

Abstract

Prostate cancer is one of the major causes of cancer death for men in the western world. Magnetic resonance imaging (MRI) is being increasingly used as a modality to detect prostate cancer. Therefore, computer-aided detection of prostate cancer in MRI images has become an active area of research. In this paper we investigate a fully automated computer-aided detection system which consists of two stages. In the first stage, we detect initial candidates using multi-atlas-based prostate segmentation, voxel feature extraction, classification and local maxima detection. The second stage segments the candidate regions and using classification we obtain cancer likelihoods for each candidate. Features represent pharmacokinetic behavior, symmetry and appearance, among others. The system is evaluated on a large consecutive cohort of 347 patients with MR-guided biopsy as the reference standard. This set contained 165 patients with cancer and 182 patients without prostate cancer. Performance evaluation is based on lesion-based free-response receiver operating characteristic curve and patient-based receiver operating characteristic analysis. The system is also compared to the prospective clinical performance of radiologists. Results show a sensitivity of 0.42, 0.75, and 0.89 at 0.1, 1, and 10 false positives per normal case. In clinical workflow the system could potentially be used to improve the sensitivity of the radiologist. At the high specificity reading setting, which is typical in screening situations, the system does not perform significantly different from the radiologist and could be used as an independent second reader instead of a second radiologist. Furthermore, the system has potential in a first-reader setting.

Prostate cancer diagnosis using deep learning with 3D multiparametric MRI

“…The accuracy of mpMRI with PI-RADS is also largely dependent on the experience of the radiologist 5 . In the last few years, several machine learning algorithms have been proposed 6,7 , based on engineered features. However, feature engineering usually requires a significant amount of domain expertise and hence the accuracies of these models are highly variable depending on the quality of the engineered features.…”

A fully automatic computer aided diagnosis system for peripheral zone prostate cancer detection using multi-parametric magnetic resonance imaging

“…One of the most promising applications of mp-MRI is to detect prostate cancer (PCa) [1][2][3]. Current indications of mp-MRI include patients with rising levels of prostatic-specific-antigen (PSA) after one or more negative transrectal ultrasound guided (TRUS) biopsies [4,5]. One good reason why mp-MRI has not yet progressed to becoming a front-line imaging modality to detect PCa is because it is a labour-intensive examination and has a steep learning curve.…”

“…One good reason why mp-MRI has not yet progressed to becoming a front-line imaging modality to detect PCa is because it is a labour-intensive examination and has a steep learning curve. Indeed, interpretation requires experienced radiologists capable of analysing data extracted from the different MR sequences [5,6].…”

“…Similarly, Niaf et al [10] used a SVM classifier combined with a t-test featureselection method, and achieved an area under the ROC curve (AUROC) equal to 0.82. Finally, Litjens et al [5] included different stages in their CAD system, investigating a novel voxel classification step in combination with a candidate classification stage. None of the described studies has adopted a fully automatic registration step to align both the Dynamic Contrast Enhanced (DCE-MR) and the Diffusion Weighted (DW) to the T2-weighted (T2w) image.…”

Medical Image Segmentation for Liver Diseases: A Survey

“…Therefore, the diagnosis of diseases using computers has become an active area of research [7]. There is a strong need to have efficient CAD system that accurately examines the medical images and reduce the required time for accurate diagnosing of diseases.…”