Comparison of the manual, semiautomatic, and automatic selection and leveling of hot spots in whole slide images for Ki-67 quantification in meningiomas

Digital image analysis of Ki67 hotspot detection and index counting in gastroenteropancreatic neuroendocrine neoplasms

The Ki-67 proliferative index plays a pivotal role in the subclassification of neuroendocrine neoplasm (NEN) according to the WHO Classification of Digestive System Tumors (5th edition), which designates neuroendocrine tumor (NET) grades 1, 2, and 3 for Ki-67 proliferative index of <3 %, 3-20 %, and >20 %, respectively. Proliferative index calculation must be performed in the hotspot, traditionally selected by visual scanning at low-power magnification. Recently, gradient map visualization has emerged as a tool for various purposes, including hotspot selection. This study includes 97 cases of gastrointestinal neuroendocrine neoplasms, with hotspots selected by bare eye and gradient map visualization (GM). Each hotspot was analyzed using three methods: eye estimation (EE), digital image analysis (DIA), and manual counting. Of the NENs studied, 91 % were NETs (26 % for G1, 55 % for G2, and 10 % for G3). Only 9 cases were neuroendocrine carcinoma (NEC). Between two hotspot selection methods, GM resulted in a higher grade in 14.77 % of cases, primarily upgrading from NET G1 to G2. Among the counting methods, DIA demonstrated substantial agreement with manual counting, both for pathologist and resident. Grading by other methods tended to result in a higher grade than MC (26.99 % with EE and 8.52 % with DIA). Given its clinical and statistical significance, this study advocates for the application of GM in hotspot selection to identify higher-grade tumors. Furthermore, DIA provides accurate grading, offering time efficiency over MC.

Comparison of vascular endothelial growth factor/vascular endothelial growth factor receptor 2 expression and its relationship to tumor cell proliferation in canine epithelial and mesenchymal tumors

The vascular endothelial growth factor (VEGF)/VEGF receptor 2 (VEGFR2) signaling pathway plays an important role in tumor angiogenesis. VEGFR2 is expressed not only in vascular endothelial cells but also in tumor cells; however, the relationship of VEGF/VEGFR2 expression and tumor proliferation has yet to be elucidated. In addition, since several studies have reported that VEGFR2 inhibitors are more effective against epithelial tumors than mesenchymal tumors, there may be a difference in VEGF/VEGFR2 expression between epithelial and mesenchymal tumors. The purpose of this study was to elucidate differences in VEGF/VEGFR2 expression between epithelial and mesenchymal tumors and the relationship of VEGF/VEGFR2 expression and proliferation in canine tumor cells. We assessed 29 epithelial and 21 mesenchymal canine tumors for microvessel density (MVD), mRNA transcription levels of von Willebrand Factor (vWF) and endoglin, expression of VEGF, VEGFR2, and phosphorylated VEGFR2 (pVEGFR2), and proliferation index (PI) using real-time reverse transcription polymerase chain reaction and immunohistochemistry. VEGFR2 expression on vascular endothelial cells, MVD, and mRNA transcription levels of vWF and endoglin were not significantly different between the two groups. However, expression of VEGF, VEGFR2, and pVEGFR2 was higher in epithelial tumors (P<0.01). Moreover, PI correlated with pVEGFR2 expression in only epithelial tumors (P<0.01, Rs=0.543). These results suggest that the activity of VEGF/VEGFR2 signaling in tumor cells is raised in epithelial tumors, and that this signaling pathway may be related to tumor cell proliferation in epithelial tumors.

Value of KI-67/MIB-1 labeling index and simpson grading system to predict the recurrence of who grade I intracranial meningiomas compared to who grade II

Simpson grading of resection has been used as a predictor of intracranial meningioma (IM) recurrence. Histopathological findings, like the Ki-67/MIB-1 labeling index, may be useful in the assessment risk of recurrence. Our objective was to analyze the predictive value of meningioma recurrence using both parameters. We retrospectively studied 322 consecutive patients with histopathological diagnosis of IM WHO grade I and 43 patients with IM WHO grade II in a 13-year period. Multivariate survival analysis was performed. In the WHO grade I IM group, recurrence was observed in 28 patients (8.69%). The Cox regression model for WHO grade I IM, provided a significative hazard ratio (HR) for Ki-67/MIB-1 index ≥3 (HR = 36.35, p < 0.001) and Simpson's grading resection, grade II (HR = 2.03, p = 0.045), grade III (HR = 3.41, p = 0.034) and grade IV (HR = 19.75, p ≥ 0.001). In the WHO grade II IM group, recurrence was observed in 10 patients (23.25%). The Cox regression model for WHO grade II IM, provided a significative hazard ratio (HR) for Ki-67/MIB-1 index ≥3% (HR = 1.66, p < 0.001) and Simpson's grading resection grade III (HR = 3.96, p = 0.027). The Kaplan-Meier survival curve showed a similar distribution of survival between WHO grade I IM with Ki-67/MIB-1 ≥3% and WHO grade II IM. In WHO grade I meningiomas, the Ki-67/MIB-1 index and Simpson grading were both independent predictors of recurrence. A similar management protocol should be advisable for WHO grade I with Ki-67/MIB-1 ≥3% and WHO grade II meningiomas.

piNET-An Automated Proliferation Index Calculator Framework for Ki67 Breast Cancer Images

Simple Summary

Approximately 2.1 million women are affected by breast cancer every year. Invasive disease accounts for 80% of breast cancer cases and is the most common and aggressive type of breast cancer. Early diagnosis is the key to survival. Ki67 biomarkers have been shown to be a promising prognostic biomarker in this regard, but manual proliferation index (PI) calculation is time consuming and subject to inter/intra observer variability which reduces clinical utility. Computational pathology tools can aid pathologists to make the diagnostic process more efficient and accurate. With the advent of deep learning, there is great promise that this technology can solve problems that were difficult to tackle in the past, but more work needs to be done to combat the challenge of multi-center datasets. In this work, a novel Ki67 PI calculator based on deep learning is proposed, called piNET, which is shown to be accurate, reliable, and consistent across multi-center datasets.

Abstract

In this work, a novel proliferation index (PI) calculator for Ki67 images called piNET is proposed. It is successfully tested on four datasets, from three scanners comprised of patches, tissue microarrays (TMAs) and whole slide images (WSI), representing a diverse multi-centre dataset for evaluating Ki67 quantification. Compared to state-of-the-art methods, piNET consistently performs the best over all datasets with an average PI difference of 5.603%, PI accuracy rate of 86% and correlation coefficient R = 0.927. The success of the system can be attributed to several innovations. Firstly, this tool is built based on deep learning, which can adapt to wide variability of medical images—and it was posed as a detection problem to mimic pathologists’ workflow which improves accuracy and efficiency. Secondly, the system is trained purely on tumor cells, which reduces false positives from non-tumor cells without needing the usual pre-requisite tumor segmentation step for Ki67 quantification. Thirdly, the concept of learning background regions through weak supervision is introduced, by providing the system with ideal and non-ideal (artifact) patches that further reduces false positives. Lastly, a novel hotspot analysis is proposed to allow automated methods to score patches from WSI that contain “significant” activity.

Hormone receptor expression, clinical and histopathological analysis in feline injection site sarcomas

Feline injection site sarcomas (FISS) are aggressive, with high recurrence and rarely metastasising. The objective of this study was to evaluate, by immunohistochemistry, the expression of oestrogen (ER) and progesterone (PR) receptors in FISS and correlate them with clinical and histopathological aspects. This was a retrospective study with 51 cases of FISS. Immunohistochemistry was performed to detect vimentin, ER, PR and Ki67 expression. Clinical, histopathological and immunohistochemical characteristics were predictor variables and the expression of ER and PR were the dependent ones. Twenty-eight (55%) of the 51 FISS cases were female and 23 (45%) male with 10.7 ± 4.2 years and median tumour size of 3 cm (2.0-5.4). The trunk was the most affected site, with 38 cases (84%). Histological grade III was observed in 57% of the cases, considering differentiation score, necrosis and mitotic index. ER expression, positive in 64% of cases, was associated with the mitotic index (P = .05) and degree of pleomorphism (P = .04). PR was not associated with the variables and 63% of cases were negative for this receptor. Thus, ER expression can affect tumour growth. The knowledge on the FISS hormonal expression is important to clarify the pathophysiological mechanisms. Further studies are needed to predict the value of ER expression in the prognosis of FISS.

J, Eriksson F, Ugleholdt H, Broholm H, Mathiesen T. Mitotic and Proliferative Indices in WHO Grade III Meningioma

Simple Summary

Malignant meningiomas are rare primary intracranial tumors associated with considerable morbidity and mortality. The diagnosis is based on the number of mitotic figures (mitotic index, MI). Consequently, the quantification of mitotic figures is prone to inter- and intraobserver variability. The mitotic marker, phosphohistone-H3 (PHH3), has been shown to be a more robust mitotic marker. Despite the prognostic value of MI across all meningioma grades, little is known of the prognostic value of the MI within malignant meningioma. Therefore, this study investigates the MI in a series of malignant meningiomas to analyze the association to progression-free survival and mitotic and proliferative indices. Furthermore, we investigated the precision (repeatability) of mitotic counts and the agreement between MI and PHH3 MI.

Abstract

Meningiomas with inherently high mitotic indices and poor prognosis, such as WHO grade III meningiomas, have not been investigated separately to establish interchangeability between conventional mitotic index counted on H&E stained slides (MI) and mitotic index counted on phosphohistone-H3 stained slides (PHH3 MI). This study investigates the agreement of MI and PHH3 MI and to analyze the association of progression-free survival (PFS) and MI, PHH3 MI, and the proliferative index (PI, Ki-67) in WHO grade III meningioma. Tumor specimens from 24 consecutive patients were analyzed for expression of Ki-67, PHH3 MI, and MI. Quantification was performed independently by two observers who made replicate counts in hot spots and overall tumor staining. Repeatability in replicate counts from MI and PHH3 MI was low in both observers. Consequently, we could not report the agreement. MI, PHH3 MI and hot spot counts of Ki-67 were associated with PFS (MI hot spot HR = 1.61, 95% CI 1.12–2.31, p = 0.010; PHH3 MI hot spot HR = 1.59, 95% CI 1.15–2.21, p = 0.006; Ki-67 hot spot HR = 1.06, 95% CI 1.02–1.11. p = 0.004). We found markedly low repeatability of manually counted MI and PHH3 MI in WHO grade III meningioma, and we could not conclude that the two methods agreed. Subsequently, quantification with better repeatability should be sought. All three biomarkers were associated with PFS.

Detection of Ki67 Hot-Spots of Invasive Breast Cancer Based on Convolutional Neural Networks Applied to Mutual Information of H&E and Ki67 Whole Slide Images

Ki67 hot-spot detection and its evaluation in invasive breast cancer regions play a significant role in routine medical practice. The quantification of cellular proliferation assessed by Ki67 immunohistochemistry is an established prognostic and predictive biomarker that determines the choice of therapeutic protocols. In this paper, we present three deep learning-based approaches to automatically detect and quantify Ki67 hot-spot areas by means of the Ki67 labeling index. To this end, a dataset composed of 100 whole slide images (WSIs) belonging to 50 breast cancer cases (Ki67 and H&E WSI pairs) was used. Three methods based on CNN classification were proposed and compared to create the tumor proliferation map. The best results were obtained by applying the CNN to the mutual information acquired from the color deconvolution of both the Ki67 marker and the H&E WSIs. The overall accuracy of this approach was 95%. The agreement between the automatic Ki67 scoring and the manual analysis is promising with a Spearman’s ρ correlation of 0.92. The results illustrate the suitability of this CNN-based approach for detecting hot-spots areas of invasive breast cancer in WSI.

Improving the accuracy of gastrointestinal neuroendocrine tumor grading with deep learning

The Ki-67 index is an established prognostic factor in gastrointestinal neuroendocrine tumors (GI-NETs) and defines tumor grade. It is currently estimated by microscopically examining tumor tissue single-immunostained (SS) for Ki-67 and counting the number of Ki-67-positive and Ki-67-negative tumor cells within a subjectively picked hot-spot. Intraobserver variability in this procedure as well as difficulty in distinguishing tumor from non-tumor cells can lead to inaccurate Ki-67 indices and possibly incorrect tumor grades. We introduce two computational tools that utilize Ki-67 and synaptophysin double-immunostained (DS) slides to improve the accuracy of Ki-67 index quantitation in GI-NETs: (1) Synaptophysin-KI-Estimator (SKIE), a pipeline automating Ki-67 index quantitation via whole-slide image (WSI) analysis and (2) deep-SKIE, a deep learner-based approach where a Ki-67 index heatmap is generated throughout the tumor. Ki-67 indices for 50 GI-NETs were quantitated using SKIE and compared with DS slide assessments by three pathologists using a microscope and a fourth pathologist via manually ticking off each cell, the latter of which was deemed the gold standard (GS). Compared to the GS, SKIE achieved a grading accuracy of 90% and substantial agreement (linear-weighted Cohen’s kappa 0.62). Using DS WSIs, deep-SKIE displayed a training, validation, and testing accuracy of 98.4%, 90.9%, and 91.0%, respectively, significantly higher than using SS WSIs. Since DS slides are not standard clinical practice, we also integrated a cycle generative adversarial network into our pipeline to transform SS into DS WSIs. The proposed methods can improve accuracy and potentially save a significant amount of time if implemented into clinical practice.

Development and Validation of an Image Analysis System for the Measurement of Cell Proliferation in Mammary Glands of Rats

Reliable detection and measurement of cell proliferation are essential in the preclinical assessment of carcinogenic risk of therapeutics. In this context, the assessment of mitogenic potential on mammary glands is crucial in the preclinical safety evaluation of novel insulins. The existing manual counting is time-consuming and subject to operator bias. To standardize the processes, make it faster, and resistant to errors, we developed a semiautomated image analysis system (CEPA software, which is open-source) for counting of proliferating cells in photomicrographs of mammary gland sections of rats labeled with Ki-67. We validated the software and met the predefined targets for specificity, accuracy, and reproducibility. In comparison to manual counting, the respective mean differences in absolute labeling indices (LIs) for CEPA software were 3.12% for user 1 and 3.05% for user 2. The respective regression analysis revealed a good correlation between the CEPA software user and manual counting. Moreover, the CEPA software showed enhanced reproducibility between independent users. The interuser variability is centered around 0 and the absolute difference was about 0.53% LI. Based on validation data, our software has superiority to the manual counting and is a valid and reliable tool for the routine analysis of cell proliferation in mammary glands from rats exposed to insulin analogs.

Molecular profiling predicts meningioma recurrence and reveals loss of DREAM complex repression in aggressive tumors

Meningiomas account for one-third of all primary brain tumors. Although typically benign, about 20% of meningiomas are aggressive, and despite the rigor of the current histopathological classification system there remains considerable uncertainty in predicting tumor behavior. Here, we analyzed 160 tumors from all 3 World Health Organization (WHO) grades (I through III) using clinical, gene expression, and sequencing data. Unsupervised clustering analysis identified 3 molecular types (A, B, and C) that reliably predicted recurrence. These groups did not directly correlate with the WHO grading system, which classifies more than half of the tumors in the most aggressive molecular type as benign. Transcriptional and biochemical analyses revealed that aggressive meningiomas involve loss of the repressor function of the DREAM complex, which results in cell-cycle activation; only tumors in this category tend to recur after full resection. These findings should improve our ability to predict recurrence and develop targeted treatments for these clinically challenging tumors.

Histological score for degrees of severity in an implant-associated infection model in mice

INTRODUCTION

Several scores were introduced to diagnose and to classify osteomyelitis in practice. Mouse models are often used to study the pathophysiology of bone infection and to test therapeutic strategies. Aim of the present study was to design a score to diagnose and quantify implant-associated infection in a murine experimental model.

MATERIALS AND METHODS

Four independent parameters were developed: existence of callus, consolidation of the fracture, structural changes of the medullary cavity and number of bacteria. The score was assessed in a standardized implant-associated mouse model with 35 BALB/c-mice. The left femur was osteotomized, fixed by a titanium locking plate and infection was induced by inoculation of Staphylococcus aureus into the fracture gap. For the sham group, the procedure was performed without inoculation of bacteria. The score was assessed on days 7, 14 and 28. Each item of the score showed lower values for the infection group compared to the controls after 4 weeks.

RESULTS

Regardless of the assessed time point, the overall total score was significantly higher in the control group compared to the infection group (p < 0.0001). Analysis revealed a sensitivity of 0.85, specificity of 1.0, negative predictive value of 0.67 and positive predictive value of 1.0.

CONCLUSION

The proposed score assessing severity of fracture-related infection in an implant-associated murine model was easy to access, feasible to diagnose and estimate bone healing and infection in a murine bone infection with a high sensitivity. Therefore, this score might be a useful tool to quantify infection-related changes after fracture in further future preclinical studies.

Computer analysis of histopathological images for tumor grading. 2

OBJECTIVE

We have upgraded our own original color filtration pixel-by-pixel (CFPP) method (Klonowski et al 2018a Physiol. Meas. 39 034002) to enable not only automatic and rapid assessment of the proliferation index of a tumor or neoplasm but also quick automatic location of hot-spots (regions of interest, ROIs) in immunohistochemically stained microscopic images of neoplasms and tumors.

APPROACH

Neoplastic cells stain differently from normal cells. By counting in a given window the number of pixels belonging to the given subspaces of (R,G,B) color space which correspond, respectively, to proliferating cells (which are mostly neoplastic) and non-proliferating cells (which are mostly normal) we calculate the local proliferation index in this window. The window is moved all around the whole histopathological virtual slide (WSI) or around a chosen part of the WSI. By adding the respective numbers calculated for all the windows covering the WSI or the chosen part of it one can easily calculate the global proliferation index.

MAIN RESULTS

The method is rapid and does not require the time-consuming step of selecting ROIs manually nor does it need computationally complicated detection of hot-spots, both of which attempt to emulate a pathologist's way of thinking. We apply our method to a set of slide images of diffuse large B-cell lymphoma.

SIGNIFICANCE

By appropriate changes in the (R,G,B) color filtration thresholds, our method may be adapted to the analysis of other types of tumors. It may also be adapted for analysis of microscopic images in neuropathology. Because of its rapidity and simplicity it may also used for analysis of series of images to assess local dynamics of image complexity in network physiology applications.

Evaluation of the proliferation marker Ki-67 in gliomas: Interobserver variability and digital quantification

BACKGROUND

The Ki-67 Labelling Index (LI) is used as an ancillary tool in glioma diagnostics. Interobserver variability has been reported and no precise guidelines are available. Nor is it known whether novel digital approaches would be an advantage. Our aim was to evaluate the inter- and intraobserver variability of the Ki-67 LI between two pathologists and between pathologists and digital quantification both in whole tumour slides and in hot spots using narrow but diagnostically relevant intervals.

METHODS

In samples of 235 low and high grade gliomas, two pathologists (A and B) estimated the Ki-67 LI (5-10% intervals) for whole tumour slides and for hot spots. In 20 of the cases intraobserver variability was evaluated. For digital quantification (C) slides were scanned with subsequent systematic random sampling of viable tumour areas. A software classifier trained to identify positive and negative nuclei calculated the Ki-67 LI. The interobserver agreements were evaluated using kappa (κ) statistics.

RESULTS

The observed proportions of agreement and κ values for Ki-67 LI for whole tumour slides were: A/B: 46% (κ = 0.32); A/C: 37% (κ = 0.26); B/C: 37% (κ = 0.26). For hot spots equivalent values were: A/B: 14% (κ = 0.04); A/C: 18% (κ = 0.09); B/C: 31% (κ = 0.21).

CONCLUSIONS

Interobserver variability was pronounced between pathologists and for pathologists versus digital quantification when attempting to estimate a precise value of the Ki-67 LI. Ki-67 LI should therefore be used with caution and should not be over interpreted in the grading of gliomas. Digital quantification of Ki-67 LI in gliomas was feasible, but intra- and interlaboratory robustness need to be determined.

Computer analysis of histopathological images for tumor grading

OBJECTIVE

We developed a new method that enables automatic and rapid assessment of a tumor's proliferation index from immunohistochemically (IHC) stained microscopic images.

APPROACH

The method is based on computer-aided analysis of images - color filtration pixel-by-pixel (CFPP method) of the whole histopathological virtual slides.

MAIN RESULTS

The method is simple, rapid, and does not require the time consuming step of selecting manually areas of interest nor the need for computationally complicated detection of hot-spots, both of which attempt to emulate a pathologist's way of estimating a proliferation index. We apply our method to a set of diffuse large B-cell lymphoma (DLBCL) slide images.

SIGNIFICANCE

By appropriate changes in the color filtration thresholds, our method may be adapted to the analysis of other types of tumors. It may also be adapted for analysis of microscopic images in neuropathology, like biopsies of dystrophic muscles. Because of its simplicity and rapidity it may also be applied for analysis of series of images to assess dynamics of image complexity in network physiology.

Impact of Her-2 Overexpression on Survival of Patients with Metastatic Breast Cancer

Introduction: Breast cancer is a complex and heterogeneous disease which is increasingly important as a public health problem. In Brazil, 57,960 new cases have been estimated to be the burden in 2016 and 2017. Despite advances in early diagnosis and therapy, approximately 20-30% of patients, even with early stage lesions, will develop distant metastatic disease. Tumors with similar clinical and pathological presentations may have differing behavior, so it is important to understand specific biological characteristics. Objective: To investigate tumor markers of primary tumors featuring pleural metastasis to identify organ-specific characteristics of metastatic breast cancer. Methods: In a historical cohort study, immunohistochemistry was performed on cell blocks of neoplastic pleural effusions and results were compared with clinicopathological data. Results: The median survival time with Her-2 overexpression in malignant pleural effusions was 2.2 months, whereas cases without overexpression survived, on average, for seven months (p = 0.02). Conclusions: We emphasize that metastases may behave independently of primary tumors, but the present results indicate that therapeutic agents targeting Her-2 overexpression could increase survival in metastatic breast cancer cases.

An Advanced Deep Learning Approach for Ki-67 Stained Hotspot Detection and Proliferation Rate Scoring for Prognostic Evaluation of Breast Cancer

Being a non-histone protein, Ki-67 is one of the essential biomarkers for the immunohistochemical assessment of proliferation rate in breast cancer screening and grading. The Ki-67 signature is always sensitive to radiotherapy and chemotherapy. Due to random morphological, color and intensity variations of cell nuclei (immunopositive and immunonegative), manual/subjective assessment of Ki-67 scoring is error-prone and time-consuming. Hence, several machine learning approaches have been reported; nevertheless, none of them had worked on deep learning based hotspots detection and proliferation scoring. In this article, we suggest an advanced deep learning model for computerized recognition of candidate hotspots and subsequent proliferation rate scoring by quantifying Ki-67 appearance in breast cancer immunohistochemical images. Unlike existing Ki-67 scoring techniques, our methodology uses Gamma mixture model (GMM) with Expectation-Maximization for seed point detection and patch selection and deep learning, comprises with decision layer, for hotspots detection and proliferation scoring. Experimental results provide 93% precision, 0.88% recall and 0.91% F-score value. The model performance has also been compared with the pathologists’ manual annotations and recently published articles. In future, the proposed deep learning framework will be highly reliable and beneficial to the junior and senior pathologists for fast and efficient Ki-67 scoring.

Content-based analysis of Ki-67 stained meningioma specimens for automatic hot-spot selection

Background

Hot-spot based examination of immunohistochemically stained histological specimens is one of the most important procedures in pathomorphological practice. The development of image acquisition equipment and computational units allows for the automation of this process. Moreover, a lot of possible technical problems occur in everyday histological material, which increases the complexity of the problem. Thus, a full context-based analysis of histological specimens is also needed in the quantification of immunohistochemically stained specimens. One of the most important reactions is the Ki-67 proliferation marker in meningiomas, the most frequent intracranial tumour. The aim of our study is to propose a context-based analysis of Ki-67 stained specimens of meningiomas for automatic selection of hot-spots.

Methods

The proposed solution is based on textural analysis, mathematical morphology, feature ranking and classification, as well as on the proposed hot-spot gradual extinction algorithm to allow for the proper detection of a set of hot-spot fields. The designed whole slide image processing scheme eliminates such artifacts as hemorrhages, folds or stained vessels from the region of interest. To validate automatic results, a set of 104 meningioma specimens were selected and twenty hot-spots inside them were identified independently by two experts. The Spearman rho correlation coefficient was used to compare the results which were also analyzed with the help of a Bland-Altman plot.

Results

The results show that most of the cases (84) were automatically examined properly with two fields of view with a technical problem at the very most. Next, 13 had three such fields, and only seven specimens did not meet the requirement for the automatic examination. Generally, the Automatic System identifies hot-spot areas, especially their maximum points, better. Analysis of the results confirms the very high concordance between an automatic Ki-67 examination and the expert’s results, with a Spearman rho higher than 0.95.

Conclusion

The proposed hot-spot selection algorithm with an extended context-based analysis of whole slide images and hot-spot gradual extinction algorithm provides an efficient tool for simulation of a manual examination. The presented results have confirmed that the automatic examination of Ki-67 in meningiomas could be introduced in the near future.

Electronic supplementary material

The online version of this article (doi:10.1186/s13000-016-0546-7) contains supplementary material, which is available to authorized users.

Survey: interpolation methods for whole slide image processing

Evaluating whole slide images of histological and cytological samples is used in pathology for diagnostics, grading and prognosis . It is often necessary to rescale whole slide images of a very large size. Image resizing is one of the most common applications of interpolation. We collect the advantages and drawbacks of nine interpolation methods, and as a result of our analysis, we try to select one interpolation method as the preferred solution. To compare the performance of interpolation methods, test images were scaled and then rescaled to the original size using the same algorithm. The modified image was compared to the original image in various aspects. The time needed for calculations and results of quantification performance on modified images were also compared. For evaluation purposes, we used four general test images and 12 specialized biological immunohistochemically stained tissue sample images. The purpose of this survey is to determine which method of interpolation is the best to resize whole slide images, so they can be further processed using quantification methods. As a result, the interpolation method has to be selected depending on the task involving whole slide images.