OBJECTIVE We upgraded our own original colour filtration pixel-by-pixel (CFPP) method [1] to enab... more OBJECTIVE We upgraded our own original colour filtration pixel-by-pixel (CFPP) method [1] to enable not only automatic and rapid assessment of a tumor's proliferation index but also quick automatic location of hot-spots (Regions of Interest, ROIs) in immunohistochemically stained microscopic images of tumors. APPROACH Neoplastic cells get stained differently than normal cells. By counting in the given window the numbers of pixels belonging to the given subspaces of the space of colours, (R,G,B) , which correspond, respectively, to proliferating cells (that are mostly neoplastic) and non-proliferating cells (that are mostly normal) we calculate local proliferation index in this window. And the window is moved all around the whole histopathological virtual slide (WSI) or around a chosen part of WSI. By adding the respective numbers calculated for all the windows covering the WSI or the chosen part of it one easily calculates the global proliferation index. MAIN RESULTS The method is rapid and does not require the time consuming step of selecting ROIs manually nor it needs 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 diffuse large B-cell lymphoma (DLBCL) slide images. SIGNIFICANCE By appropriate changes in the colours' R, G, and B 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 be applied for analysis of series of images to assess local dynamics of image complexity in application in Network Physiology.
The aim of this study is the comparison of the movement of various types of human dermal fibrobla... more The aim of this study is the comparison of the movement of various types of human dermal fibroblasts with and without genetic modification. Three groups of fibroblasts were cultured and monitored: (1) control group of fibroblasts isolated from human skin (not transduced); (2) transduced with lentivirus bearing EGFP fluorescent marker; (3) transduced with lentivirus bearing DsRed2 fluorescent marker. The experimental sequences of images documenting cells’ movement have been analyzed using image processing methods. The cells’ movement was described by: (1) distance between each two consecutive images/frames, (2) distance and displacement covered by each fibroblast in 30 minutes (3) length of cell crawling cycle, (4) the movement tortuosity coefficient for all fibroblasts in each group. Also shape features such as: area, perimeter, eccentricity, length of the major axis have been analyzed. It appears that the general movement behavior is not changed by the process of transduction but some of its aspects are modified. The efficiency of movement in the sense of distance covered and region penetration is decreased because of changes in cells’ morphology. Transduced cells are less polarized and develop extra podia during their crawling.
Advances in intelligent systems and computing, Nov 1, 2018
Quantitative analysis of histopathological sections can be used to support the diagnosis and eval... more Quantitative analysis of histopathological sections can be used to support the diagnosis and evaluate the disease progression by pathologists. The use of computer-aided diagnosis in pathology can substantially enhance the efficiency and accuracy of pathologists decisions, and overall benefit the patient. The evaluation of the shape of specific types of cell nuclei plays an important role in histopathological examination in various types of cancer. In this study we try to verify how much the results of segmentation could be improved with applying boundary refinement algorithm to thresholded histopathological image. In this paper we studied 5 methods based on various approaches: active contour, k-means clustering, and region-growing. For evaluation purposes, ground truth templates were generated by manual annotation of images. The performance is evaluated using pixel-wise sensitivity and specificity metrics. It appears that satisfactory results were achieved only by two algorithms based on active contour. By applying methodology based on active contour algorithm we managed to achieve sensitivity of about 93% and specificity of over 99%. To sum up, thresholding algorithms produce results that almost never perfectly fit to real object’s boundary, but this initial detection of objects followed by boundary refinement results in more accurate segmentation.
Histopathological sections allow pathologists to evaluate a wide range of specimens, including br... more Histopathological sections allow pathologists to evaluate a wide range of specimens, including breast cancer, obtained from biopsies and surgical procedures. The accuracy of the employed automated cell detection technique is critical in obtaining efficient diagnostic performance. In this paper we investigate 18 different adaptive threshold methods based on various approaches. We validate the methods on a set of histopathological images of breast cancer, where immunohistochemical staining of FOXP3 was performed with 3,3’-diaminobenzidine and hematoxylin. The thresholding is performed on monochromatic images derived from original images: separate channels of Red-Green-Blue and Hue-Saturation-Value, layers of results of color deconvolution, ‘brown’ channel, ‘blue-ratio’ layer. The main objective of the evaluation is to determine if the detected objects obtained by the tested methods of thresholding are consistent with the manually labeled ones. The performance is evaluated using precision, sensitivity and F1 score measures. It appears that satisfactory results were achieved only by 6 methods. It was found that bradley method is the best performing method for nuclei detection in this type of stained tissue samples. It has best sensitivity value for images after color deconvolution and Value layer (of Hue-Saturation-Value color space), 0.970 and 0.975 respectively. As a result, we recommend a most efficient local threshold technique in the case of nuclei detection in digitized immunohistochemically stained tissue sections. This initial detection of objects followed by texture, size and shape analysis will give a collection of cells’ nuclei to perform further accurate segmentation. The proposed detection method will be used in a framework focused on computer-aided diagnosis.
Advances in intelligent systems and computing, Aug 20, 2017
The aim of this study is the comparison of the movement of various types of human dermal fibrobla... more The aim of this study is the comparison of the movement of various types of human dermal fibroblasts with and without genetic modification. Three groups of fibroblasts were cultured and monitored: (1) control group of fibroblasts isolated from human skin (not transduced); (2) transduced with lentivirus bearing EGFP fluorescent marker; (3) transduced with lentivirus bearing DsRed2 fluorescent marker. The experimental sequences of images documenting cells’ movement have been analyzed using image processing methods. The cells’ movement was described by: (1) distance between each two consecutive images/frames, (2) distance and displacement covered by each fibroblast in 30 minutes (3) length of cell crawling cycle, (4) the movement tortuosity coefficient for all fibroblasts in each group. Also shape features such as: area, perimeter, eccentricity, length of the major axis have been analyzed. It appears that the general movement behavior is not changed by the process of transduction but some of its aspects are modified. The efficiency of movement in the sense of distance covered and region penetration is decreased because of changes in cells’ morphology. Transduced cells are less polarized and develop extra podia during their crawling.
Differences between computer-assisted image analysis (CAI) algorithms may cause discrepancies in ... more Differences between computer-assisted image analysis (CAI) algorithms may cause discrepancies in the identification of immunohistochemically stained immune biomarkers in biopsies of breast cancer patients. These discrepancies have implications for their association with disease outcome. This study aims to compare three CAI procedures (A, B and C) to measure positive marker areas in post-neoadjuvant chemotherapy biopsies of patients with triple-negative breast cancer (TNBC) and to explore the differences in their performance in determining the potential association with relapse in these patients. A total of 3304 digital images of biopsy tissue obtained from 118 TNBC patients were stained for seven immune markers using immunohistochemistry (CD4, CD8, FOXP3, CD21, CD1a, CD83, HLA-DR) and were analyzed with procedures A, B and C. The three methods measure the positive pixel markers in the total tissue areas. The extent of agreement between paired CAI procedures, a principal component analysis (PCA) and Cox multivariate analysis was assessed. Comparisons of paired procedures showed close agreement for most of the immune markers at low concentration. The probability of differences between the paired procedures B/C and B/A was generally higher than those observed in C/A. The principal component analysis, largely based on data from CD8, CD1a and HLA-DR, identified two groups of patients with a significantly lower probability of relapse than the others. The multivariate regression models showed similarities in the factors associated with relapse for procedures A and C, as opposed to those obtained with procedure B. General agreement among the results of CAI procedures would not guarantee that the same predictive breast cancer markers were consistently identified. These results highlight the importance of developing additional strategies to improve the sensitivity of CAI procedures.
SummaryEvaluating whole slide images of histological and cytological samples is used in pathology... more SummaryEvaluating 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 interpolat...
Advances in Intelligent Systems and Computing, 2019
The aim of this study is the comparison of the various deep convolutional neural networks for seg... more The aim of this study is the comparison of the various deep convolutional neural networks for segmentation of fibroblast in brightfield microscopic images. This investigation compares two main architectures: Unet and Linknet. Every main architecture is equipped with various ‘backbone’ network creating specific bundle. The experimental dataset consisting of 16 sequences of images of monitored cells’ culture have been split into training and validation set. Then it was analysed and used for validation of the networks to establish the best bundle (net architecture and ‘backbone’). This study proved that trained deep convolutional neural networks could be used as a segmentation tool in this task.
Abstract With the advent and great advances of methods based on deep learning in image analysis, ... more Abstract With the advent and great advances of methods based on deep learning in image analysis, it appears that they can be effective in digital pathology to support the work of pathologists. However, a major limitation in the development of computer-aided diagnostic systems for pathology is the cost of data annotation. Evaluation of tissue (histopathological) and cellular (cytological) specimens seems to be a complex challenge. To simplify the laborious process of obtaining a sufficiently large set of data, a number of different systems could be used for image annotation. Some of these systems are reviewed in this paper with a comparison of their capabilities.
Introduction: We have proposed new quantitative methods for assessment and classification of sele... more Introduction: We have proposed new quantitative methods for assessment and classification of selected tumors to assist oncopathological diagnostics. These methods are based on computer-aided analysis of histopathological images. Materials and Methods: In this paper we propose a simple new methods of quantitative image-based assessment of color tissue histopathology slides. The method is based on color filtration pixel-by-pixel of the whole virtual slides will be called CFPP method. Information contained in such slides that results from staining tissues from biopsies with different dyes helps to reveal details that might otherwise not be apparent. Results: We demonstrated that with appropriate image preprocessing our method helps out in diagnosis of Diffuse Large B-Cell Lymphoma (DLBCL or DLBL). Conclusions: In collaboration with pathologists and oncologists our computer-aided method may be adapted and adjusted to assist in diagnosis of other problems in digital pathology. Key-Words:...
The foremost cause of death of breast cancer (BC) patients is metastasis, and the first site to w... more The foremost cause of death of breast cancer (BC) patients is metastasis, and the first site to which BC predominantly metastasizes is the axillary lymph node (ALN). Thus, ALN status is a key prognostic indicator at diagnosis. The immune system has an essential role in cancer progression and dissemination, so its evaluation in ALNs could have significant applications. In the present study we aimed to investigate the association of clinical-pathological and immune variables in the primary tumour and non-metastatic ALNs (ALNs–) of a cohort of luminal A and triple-negative BC (TNBC) patients with cancer-specific survival (CSS) and time to progression (TTP). We analysed the differences in the variables between patients with different outcomes, created univariate and multivariate Cox regression models, validated them by bootstrapping and multiple imputation of missing data techniques, and used Kaplan–Meier survival curves for a 10-years follow-up. We found some clinical-pathological variables at diagnosis (tumour diameter, TNBC molecular profile and presence of ALN metastasis), and the levels of several immune markers in the two studied sites, to be associated with worse CSS and TTP. Nevertheless, only CD68 and CD83 in ALNs– were confirmed as independent prognostic factors for TTP. The study identified the importance of macrophage and dendritic cell markers as prognostic factors of relapse for BC. We highlight the importance of studying the immune response in ALNs–, which could be relevant to the prediction of BC patients’ outcome.
OBJECTIVE We upgraded our own original colour filtration pixel-by-pixel (CFPP) method [1] to enab... more OBJECTIVE We upgraded our own original colour filtration pixel-by-pixel (CFPP) method [1] to enable not only automatic and rapid assessment of a tumor's proliferation index but also quick automatic location of hot-spots (Regions of Interest, ROIs) in immunohistochemically stained microscopic images of tumors. APPROACH Neoplastic cells get stained differently than normal cells. By counting in the given window the numbers of pixels belonging to the given subspaces of the space of colours, (R,G,B) , which correspond, respectively, to proliferating cells (that are mostly neoplastic) and non-proliferating cells (that are mostly normal) we calculate local proliferation index in this window. And the window is moved all around the whole histopathological virtual slide (WSI) or around a chosen part of WSI. By adding the respective numbers calculated for all the windows covering the WSI or the chosen part of it one easily calculates the global proliferation index. MAIN RESULTS The method is rapid and does not require the time consuming step of selecting ROIs manually nor it needs 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 diffuse large B-cell lymphoma (DLBCL) slide images. SIGNIFICANCE By appropriate changes in the colours' R, G, and B 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 be applied for analysis of series of images to assess local dynamics of image complexity in application in Network Physiology.
The aim of this study is the comparison of the movement of various types of human dermal fibrobla... more The aim of this study is the comparison of the movement of various types of human dermal fibroblasts with and without genetic modification. Three groups of fibroblasts were cultured and monitored: (1) control group of fibroblasts isolated from human skin (not transduced); (2) transduced with lentivirus bearing EGFP fluorescent marker; (3) transduced with lentivirus bearing DsRed2 fluorescent marker. The experimental sequences of images documenting cells’ movement have been analyzed using image processing methods. The cells’ movement was described by: (1) distance between each two consecutive images/frames, (2) distance and displacement covered by each fibroblast in 30 minutes (3) length of cell crawling cycle, (4) the movement tortuosity coefficient for all fibroblasts in each group. Also shape features such as: area, perimeter, eccentricity, length of the major axis have been analyzed. It appears that the general movement behavior is not changed by the process of transduction but some of its aspects are modified. The efficiency of movement in the sense of distance covered and region penetration is decreased because of changes in cells’ morphology. Transduced cells are less polarized and develop extra podia during their crawling.
Advances in intelligent systems and computing, Nov 1, 2018
Quantitative analysis of histopathological sections can be used to support the diagnosis and eval... more Quantitative analysis of histopathological sections can be used to support the diagnosis and evaluate the disease progression by pathologists. The use of computer-aided diagnosis in pathology can substantially enhance the efficiency and accuracy of pathologists decisions, and overall benefit the patient. The evaluation of the shape of specific types of cell nuclei plays an important role in histopathological examination in various types of cancer. In this study we try to verify how much the results of segmentation could be improved with applying boundary refinement algorithm to thresholded histopathological image. In this paper we studied 5 methods based on various approaches: active contour, k-means clustering, and region-growing. For evaluation purposes, ground truth templates were generated by manual annotation of images. The performance is evaluated using pixel-wise sensitivity and specificity metrics. It appears that satisfactory results were achieved only by two algorithms based on active contour. By applying methodology based on active contour algorithm we managed to achieve sensitivity of about 93% and specificity of over 99%. To sum up, thresholding algorithms produce results that almost never perfectly fit to real object’s boundary, but this initial detection of objects followed by boundary refinement results in more accurate segmentation.
Histopathological sections allow pathologists to evaluate a wide range of specimens, including br... more Histopathological sections allow pathologists to evaluate a wide range of specimens, including breast cancer, obtained from biopsies and surgical procedures. The accuracy of the employed automated cell detection technique is critical in obtaining efficient diagnostic performance. In this paper we investigate 18 different adaptive threshold methods based on various approaches. We validate the methods on a set of histopathological images of breast cancer, where immunohistochemical staining of FOXP3 was performed with 3,3’-diaminobenzidine and hematoxylin. The thresholding is performed on monochromatic images derived from original images: separate channels of Red-Green-Blue and Hue-Saturation-Value, layers of results of color deconvolution, ‘brown’ channel, ‘blue-ratio’ layer. The main objective of the evaluation is to determine if the detected objects obtained by the tested methods of thresholding are consistent with the manually labeled ones. The performance is evaluated using precision, sensitivity and F1 score measures. It appears that satisfactory results were achieved only by 6 methods. It was found that bradley method is the best performing method for nuclei detection in this type of stained tissue samples. It has best sensitivity value for images after color deconvolution and Value layer (of Hue-Saturation-Value color space), 0.970 and 0.975 respectively. As a result, we recommend a most efficient local threshold technique in the case of nuclei detection in digitized immunohistochemically stained tissue sections. This initial detection of objects followed by texture, size and shape analysis will give a collection of cells’ nuclei to perform further accurate segmentation. The proposed detection method will be used in a framework focused on computer-aided diagnosis.
Advances in intelligent systems and computing, Aug 20, 2017
The aim of this study is the comparison of the movement of various types of human dermal fibrobla... more The aim of this study is the comparison of the movement of various types of human dermal fibroblasts with and without genetic modification. Three groups of fibroblasts were cultured and monitored: (1) control group of fibroblasts isolated from human skin (not transduced); (2) transduced with lentivirus bearing EGFP fluorescent marker; (3) transduced with lentivirus bearing DsRed2 fluorescent marker. The experimental sequences of images documenting cells’ movement have been analyzed using image processing methods. The cells’ movement was described by: (1) distance between each two consecutive images/frames, (2) distance and displacement covered by each fibroblast in 30 minutes (3) length of cell crawling cycle, (4) the movement tortuosity coefficient for all fibroblasts in each group. Also shape features such as: area, perimeter, eccentricity, length of the major axis have been analyzed. It appears that the general movement behavior is not changed by the process of transduction but some of its aspects are modified. The efficiency of movement in the sense of distance covered and region penetration is decreased because of changes in cells’ morphology. Transduced cells are less polarized and develop extra podia during their crawling.
Differences between computer-assisted image analysis (CAI) algorithms may cause discrepancies in ... more Differences between computer-assisted image analysis (CAI) algorithms may cause discrepancies in the identification of immunohistochemically stained immune biomarkers in biopsies of breast cancer patients. These discrepancies have implications for their association with disease outcome. This study aims to compare three CAI procedures (A, B and C) to measure positive marker areas in post-neoadjuvant chemotherapy biopsies of patients with triple-negative breast cancer (TNBC) and to explore the differences in their performance in determining the potential association with relapse in these patients. A total of 3304 digital images of biopsy tissue obtained from 118 TNBC patients were stained for seven immune markers using immunohistochemistry (CD4, CD8, FOXP3, CD21, CD1a, CD83, HLA-DR) and were analyzed with procedures A, B and C. The three methods measure the positive pixel markers in the total tissue areas. The extent of agreement between paired CAI procedures, a principal component analysis (PCA) and Cox multivariate analysis was assessed. Comparisons of paired procedures showed close agreement for most of the immune markers at low concentration. The probability of differences between the paired procedures B/C and B/A was generally higher than those observed in C/A. The principal component analysis, largely based on data from CD8, CD1a and HLA-DR, identified two groups of patients with a significantly lower probability of relapse than the others. The multivariate regression models showed similarities in the factors associated with relapse for procedures A and C, as opposed to those obtained with procedure B. General agreement among the results of CAI procedures would not guarantee that the same predictive breast cancer markers were consistently identified. These results highlight the importance of developing additional strategies to improve the sensitivity of CAI procedures.
SummaryEvaluating whole slide images of histological and cytological samples is used in pathology... more SummaryEvaluating 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 interpolat...
Advances in Intelligent Systems and Computing, 2019
The aim of this study is the comparison of the various deep convolutional neural networks for seg... more The aim of this study is the comparison of the various deep convolutional neural networks for segmentation of fibroblast in brightfield microscopic images. This investigation compares two main architectures: Unet and Linknet. Every main architecture is equipped with various ‘backbone’ network creating specific bundle. The experimental dataset consisting of 16 sequences of images of monitored cells’ culture have been split into training and validation set. Then it was analysed and used for validation of the networks to establish the best bundle (net architecture and ‘backbone’). This study proved that trained deep convolutional neural networks could be used as a segmentation tool in this task.
Abstract With the advent and great advances of methods based on deep learning in image analysis, ... more Abstract With the advent and great advances of methods based on deep learning in image analysis, it appears that they can be effective in digital pathology to support the work of pathologists. However, a major limitation in the development of computer-aided diagnostic systems for pathology is the cost of data annotation. Evaluation of tissue (histopathological) and cellular (cytological) specimens seems to be a complex challenge. To simplify the laborious process of obtaining a sufficiently large set of data, a number of different systems could be used for image annotation. Some of these systems are reviewed in this paper with a comparison of their capabilities.
Introduction: We have proposed new quantitative methods for assessment and classification of sele... more Introduction: We have proposed new quantitative methods for assessment and classification of selected tumors to assist oncopathological diagnostics. These methods are based on computer-aided analysis of histopathological images. Materials and Methods: In this paper we propose a simple new methods of quantitative image-based assessment of color tissue histopathology slides. The method is based on color filtration pixel-by-pixel of the whole virtual slides will be called CFPP method. Information contained in such slides that results from staining tissues from biopsies with different dyes helps to reveal details that might otherwise not be apparent. Results: We demonstrated that with appropriate image preprocessing our method helps out in diagnosis of Diffuse Large B-Cell Lymphoma (DLBCL or DLBL). Conclusions: In collaboration with pathologists and oncologists our computer-aided method may be adapted and adjusted to assist in diagnosis of other problems in digital pathology. Key-Words:...
The foremost cause of death of breast cancer (BC) patients is metastasis, and the first site to w... more The foremost cause of death of breast cancer (BC) patients is metastasis, and the first site to which BC predominantly metastasizes is the axillary lymph node (ALN). Thus, ALN status is a key prognostic indicator at diagnosis. The immune system has an essential role in cancer progression and dissemination, so its evaluation in ALNs could have significant applications. In the present study we aimed to investigate the association of clinical-pathological and immune variables in the primary tumour and non-metastatic ALNs (ALNs–) of a cohort of luminal A and triple-negative BC (TNBC) patients with cancer-specific survival (CSS) and time to progression (TTP). We analysed the differences in the variables between patients with different outcomes, created univariate and multivariate Cox regression models, validated them by bootstrapping and multiple imputation of missing data techniques, and used Kaplan–Meier survival curves for a 10-years follow-up. We found some clinical-pathological variables at diagnosis (tumour diameter, TNBC molecular profile and presence of ALN metastasis), and the levels of several immune markers in the two studied sites, to be associated with worse CSS and TTP. Nevertheless, only CD68 and CD83 in ALNs– were confirmed as independent prognostic factors for TTP. The study identified the importance of macrophage and dendritic cell markers as prognostic factors of relapse for BC. We highlight the importance of studying the immune response in ALNs–, which could be relevant to the prediction of BC patients’ outcome.
Uploads
Papers by Anna Korzynska