Paweł Filipczuk

Computer-aided diagnosis of breast cancer based on fine needle biopsy microscopic images

Prompt and widely available diagnostics of breast cancer is crucial for the prognosis of patients. One of the diagnostic methods is the analysis of cytological material from the breast. This examination requires extensive knowledge and experience of the cytologist. Computer-aided diagnosis can speed up the diagnostic process and allow for large-scale screening. One of the largest challenges in the automatic analysis of cytological images is the segmentation of nuclei. In this study, four different clustering algorithms are tested and compared in the task of fast nuclei segmentation. K-means, fuzzy C-means, competitive learning neural networks and Gaussian mixture models were incorporated for clustering in the color space along with adaptive thresholding in grayscale. These methods were applied in a medical decision support system for breast cancer diagnosis, where the cases were classified as either benign or malignant. In the segmented nuclei, 42 morphological, topological and texture features were extracted. Then, these features were used in a classification procedure with three different classifiers. The system was tested for classification accuracy by means of microscopic images of fine needle breast biopsies. In cooperation with the Regional Hospital in Zielona Góra, 500 real case medical images from 50 patients were collected. The acquired classification accuracy was approximately 96-100%, which is very promising and shows that the presented method ensures accurate and objective data acquisition that could be used to facilitate breast cancer diagnosis.

Computer-Aided Breast Cancer Diagnosis Based on the Analysis of Cytological Images of Fine Needle Biopsies

The effectiveness of the treatment of breast cancer depends on its timely detection. An early step in the diagnosis is the cytological examination of breast material obtained directly from the tumor. This work reports on advances in computer-aided breast cancer diagnosis based on the analysis of cytological images of fine needle biopsies to characterize these biopsies as either benign or malignant. Instead of relying on the accurate segmentation of cell nuclei, the nuclei are estimated by circles using the circular Hough transform. The resulting circles are then filtered to keep only high-quality estimations for further analysis by a support vector machine which classifies detected circles as correct or incorrect on the basis of texture features and the percentage of nuclei pixels according to a nuclei mask obtained using Otsus thresholding method. A set of 25 features of the nuclei is used in the classification of the biopsies by four different classifiers. The complete diagnostic procedure was tested on 737 microscopic images of fine needle biopsies obtained from patients and achieved 98.51% effectiveness. The results presented in this paper demonstrate that a computerized medical diagnosis system based on our method would be effective, providing valuable, accurate diagnostic information.

Cytological image analysis with firefly nuclei detection and hybrid one-class classification decomposition

Abstract Recently a great increase of interest in digital pathology and cytology can be observed. Computer-aided diagnosis solutions, developed to assist physicians in the early detection of diseases, can improve accuracy and robustness of the diagnosis. In this paper we present a work in progress on a computer-aided breast cancer diagnosis. We propose an efficient medical decision support framework that allows distinguishing between benign, malignant and fibroadenoma cases. The nuclei detection procedure is based on the firefly algorithm. The procedure generates nuclei markers that are used in marker-controlled watershed segmentation. Image recognition is done by a novel classifier. Instead of using a multi-class approach we decided to implement one-class decomposition strategy, where each of the classes is represented by an ensemble of one-class classifiers. We propose to use a multi-objective memetic algorithm to select the pool of one-class predictors that display at the same time high diversity and consistency. Experiments conducted on a set of 675 real case medical images obtained from patients of the Regional Hospital in Zielona Gora showed that our framework returns highly satisfactory results, outperforming other state-of-the-art methods.

Nuclei segmentation for computer-aided diagnosis of breast cancer

Abstract Breast cancer is the most common cancer among women. The effectiveness of treatment depends on early detection of the disease. Computer-aided diagnosis plays an increasingly important role in this field. Particularly, digital pathology has recently become of interest to a growing number of scientists. This work reports on advances in computer-aided breast cancer diagnosis based on the analysis of cytological images of fine needle biopsies. The task at hand is to classify those as either benign or malignant. We propose a robust segmentation procedure giving satisfactory nuclei separation even when they are densely clustered in the image. Firstly, we determine centers of the nuclei using conditional erosion. The erosion is performed on a binary mask obtained with the use of adaptive thresholding in grayscale and clustering in a color space. Then, we use the multi-label fast marching algorithm initialized with the centers to obtain the final segmentation. A set of 84 features extracted from the nuclei is used in the classification by three different classifiers. The approach was tested on 450 microscopic images of fine needle biopsies obtained from patients of the Regional Hospital in Zielona Góra, Poland. The classification accuracy presented in this paper reaches 100%, which shows that a medical decision support system based on our method would provide accurate diagnostic information.

Automatic Breast Cancer Diagnosis Based on K-Means Clustering and Adaptive Thresholding Hybrid Segmentation

The paper presents k-means based hybrid segmentation method for breast cancer diagnosis problem. It is part of the computer system to support diagnosis based on microscope images of the fine needle biopsy. The system assumes distinguishing malignant from benign cases. Described method is an alternative to the previously presented algorithms based on fuzzy c-means clustering and competitive neural networks. However, it uses similar idea of combining clustering in RGB space with adaptive thresholding. At first, thresholding reveals objects on background. Then image is clustered with k-means algorithm to distinguish nuclei from red blood cells and other objects. Correct segmentation is crucial to obtain good quality features measurements and consequently successful diagnosis. The system of malignancy classification was tested on a set of real case medical images with promising results.

Fuzzy Clustering and Adaptive Thresholding Based Segmentation Method for Breast Cancer Diagnosis

The paper provides a preview of some work in progress on the computer system to support breast cancer diagnosis. The approach is based on microscope images of the FNB (Fine Needle Biopsy) and assumes distinguishing malignant from benign cases. Research is focused on two different problems. The first is segmentation and extraction of morphometric parameters of nuclei present on cytological images. The second concentrates on breast cancer classification using selected features. Studies in both areas are conducted in parallel. This work is mainly devoted to the problem of image segmentation in order to obtain good quality features measurements. Correct segmentation is crucial for successful diagnosis. The paper describes hybrid segmentation algorithm based on fuzzy clustering and adaptive thresholding. The automatic system of malignancy classification was applied on a set of medical images with promising results.

Classifier ensemble for an effective cytological image analysis

Breast cancer is the most common type of cancer among women. As early detection is crucial for the patients health, much attention has been paid to the development of tools for effective recognition of this disease. This article presents an application of image analysis and classification methods for fine needle biopsy. In our approach, each patient is described by nine microscopic images taken from the biopsy sample. The images are related to regions of the biopsy that seem interesting to the physician who selects them arbitrarily. We propose four different hybrid segmentation algorithms dedicated to processing these images and examine their effectiveness for the nuclei feature extraction task. Classification is carried out with the usage of a classifier ensemble based on the Random Subspaces approach. To boost its effectiveness, we use a linear combination of the support functions returned by the individual classifiers in the ensemble. In the proposed medical support system, the final decision about the patient is delivered after a fusion of nine separate outputs of the classifier - each for a different image. Experimental results carried out on a diverse dataset collected by the authors prove that the proposed solution outperforms state-of-the-art classifiers and shows itself to be a valuable tool for supporting day-to-day cytologists routine.

Automatic nuclei detection on cytological images using the firefly optimization algorithm

The firefly algorithm is a powerful optimization method inspired by the flashing behavior of fireflies. In our work on computer aided breast cancer diagnosis we met a problem of automatic marking of nuclei. Our system is based on analysis of microscopic images of fine needle biopsy material. The task of the system is to identify benign and malignant lesions (optionally it can also distinguish fibroadenoma). For this purpose it extracts nuclei from cytological images in segmentation phase, then it determines their morphometric features and finally classifies the case. Some segmentation methods require a preliminary selection of objects on the image. We have adapted the firefly algorithm to this task. We have also proposed an initialization procedure. The method was experimentally shown to be satisfactorily effective. The approach was tested with real case medical data collected from patients of the Regional Hospital in Zielona Gora.

Swarm Intelligence Algorithms for Multi-level Image Thresholding

The multi-level thresholding is one of the most important issues in image segmentation. It is a time consuming problem, i.e. finding appropriate threshold values could take exceptionally long computational time. In this paper we evaluate and compare three meta-heuristic techniques tackling to this problem: ant colony, fireflies and honey bee mating. The proposed methods are validated by illustrative examples. The results show that the investigated methods do not only provide good segmentation results but also their computational effort makes them very efficient approaches.

Multi-label fast marching and seeded watershed segmentation methods for diagnosis of breast cancer cytology

Digital cytology plays an increasingly important role in breast cancer diagnosis. However, analysis of cytologic images is a very difficult task. Especially, nuclei segmentation is extremely challenging. In our work on fully automated medical diagnosis system we encountered the problem of densely clustered nuclei. We decided to use a segmentation algorithm that is rather rarely found in the literature. Multi-label fast marching was applied and compared to well-known and extensively used seeded watershed algorithm. In both methods, it is critical to determine the appropriate starting points (seeds). The seeds were determined using a combination of adaptive thresholding in grayscale, clustering in color space and conditional erosion. The proposed segmentation procedure was tested for suitability for diagnosis of the cancer. Experiments were conducted on a set of 450 microscopic images of fine needle biopsies obtained from patients of the Regional Hospital in Zielona Góra, Poland. The images were classified as either benign or malignant using 84 features extracted from isolated nuclei. Both methods gave very promising results and showed that our method is effective and can be successfully applied for computer-aided diagnosis system.