Machine learning approaches to analyze histological images of tissues from radical prostatectomies

Abstract

Computerized evaluation of histological preparations of prostate tissues involves identification of tissue components such as stroma (ST), benign/normal epithelium (BN) and prostate cancer (PCa). Image classification approaches have been developed to identify and classify glandular regions in digital images of prostate tissues; however their success has been limited by difficulties in cellular segmentation and tissue heterogeneity. We hypothesized that utilizing image pixels to generate intensity histograms of hematoxylin (H) and eosin (E) stains deconvoluted from H&E images numerically captures the architectural difference between glands and stroma. In addition, we postulated that joint histograms of local binary patterns and local variance (LBPxVAR) can be used as sensitive textural features to differentiate benign/normal tissue from cancer. Here we utilized a machine learning approach comprising of a support vector machine (SVM) followed by a random forest (RF) classifier to digitally stratify prostate tissue into ST, BN and PCa areas. Two pathologists manually annotated 210 images of low- and high-grade tumors from slides that were selected from 20 radical prostatectomies and digitized at high-resolution. The 210 images were split into the training (n = 19) and test (n = 191) sets. Local intensity histograms of H and E were used to train a SVM classifier to separate ST from epithelium (BN + PCa). The performance of SVM prediction was evaluated by measuring the accuracy of delineating epithelial areas. The Jaccard J = 59.5 ± 14.6 and Rand Ri = 62.0 ± 7.5 indices reported a significantly better prediction when compared to a reference method (Chen et al., Clinical Proteomics 2013, 10:18) based on the averaged values from the test set. To distinguish BN from PCa we trained a RF classifier with LBPxVAR and local intensity histograms and obtained separate performance values for BN and PCa: JBN = 35.2 ± 24.9, OBN = 49.6 ± 32, JPCa = 49.5 ± 18.5, OPCa = 72.7 ± 14.8 and Ri = 60.6 ± 7.6 in the test set. Our pixel-based classification does not rely on the detection of lumens, which is prone to errors and has limitations in high-grade cancers and has the potential to aid in clinical studies in which the quantification of tumor content is necessary to prognosticate the course of the disease. The image data set with ground truth annotation is available for public use to stimulate further research in this area.

Emerging Themes in Image Informatics and Molecular Analysis for Digital Pathology

“…More complex textural features can also be extracted; these include steerable and multiscale gradient features via mathematical operators such as Gabor filters (61), local binary patterns (62), and Laws filters (63). The shape and texture of nuclei within the stroma are significantly correlated with disease recurrence and patient outcome in breast (64), prostate (65), and oropharyngeal cancers (55). Figure 3 shows the digital stain representation of a routine H&E image, with overlays of nuclear architecture networks and capture of stromal and epithelial textural variations.…”

Classification of breast cancer stroma as a tool for prognosis

“…The accuracy of all three classifiers was high, wherein LBP gave the best performance (AUC = 0.995). Gertych et al 12 used the difference in pixel intensity characteristics of H&E images with texture histograms (joint histograms of local binary patterns and local variance) to classify regions in prostate tissue. They built a support vector machine (SVM) classifier to distinguish between stromal and epithelial regions.…”

Study of Image Analysis Algorithms for Segmentation, Feature Extraction and Classification of Cells

“…The proposed classifier is k-NN based and was implemented using MESSIF framework (Batko et al 2008). In comparison with SVM, the advantages of k-NN based approach in the implicit support of multi-class division and (Dogantekin, Avci, and Erkus 2013), (Karkanis et al 2003), (Gertych et al 2015), (Ashcroft et al 2011), (Stoklasa, Majtner, and Svoboda 2014), (Nanni and Lumini 2007), (Filipczuk, Krawczyk, and Woźniak 2013) Non-probabilistic: support vector machine, K-nearest neighbor and linear discriminant analysis.…”

“…(Liu and Liu 2013), , (Wang et al 2006), (Dogantekin, Avci, and Erkus 2013), (Ong and Chandran 2005), (Gertych et al 2015), (Abeysekera et al 2014), (Mao et al 2014), (dos Santos et al 2015), (Kayaaltı et al 2014), (Stoklasa, Majtner, and Svoboda 2014) Wavelet features.…”

“…Color coherence vectors (CCV), border/ Interior classification (BIC), color correlograms (Jaiswal et al 2003), (Huang et al 2006), (Karkanis et al 2003), (Filipczuk, Krawczyk, and Woźniak 2013) (Gertych et al 2015), skewness, and kurtosis. Other authors proposed a combination, modification or new variants of several well-known features.…”