Ograniczanie wyników
Czasopisma help
Autorzy help
Lata help
Preferencje help
Widoczny [Schowaj] Abstrakt
Liczba wyników
Powiadomienia systemowe
  • Sesja wygasła!

Znaleziono wyników: 23

Liczba wyników na stronie
first rewind previous Strona / 2 next fast forward last
Wyniki wyszukiwania
Wyszukiwano:
w słowach kluczowych:  computer-aided diagnosis
help Sortuj według:

help Ogranicz wyniki do:
first rewind previous Strona / 2 next fast forward last
EN
Mammography is the primary imaging modality used for early detection and diagnosis of breast cancer. X-ray mammogram analysis mainly refers to the localization of suspicious regions of interest followed by segmentation, towards further lesion classification into benign versus malignant. Among diverse types of breast abnormalities, masses are the most important clinical findings of breast carcinomas. However, manually segmenting breast masses from native mammograms is time-consuming and error-prone. Therefore, an integrated computer-aided diagnosis system is required to assist clinicians for automatic and precise breast mass delineation. In this work, we present a two-stage multiscale pipeline that provides accurate mass contours from high-resolution full mammograms. First, we propose an extended deep detector integrating a multi-scale fusion strategy for automated mass localization. Second, a convolutional encoder-decoder network using nested and dense skip connections is employed to fine-delineate candidate masses. Unlike most previous studies based on segmentation from regions, our framework handles mass segmentation from native full mammograms without any user intervention. Trained on INbreast and DDSM-CBIS public datasets, the pipeline achieves an overall average Dice of 80.44% on INbreast test images, outperforming state-of-the-art. Our system shows promising accuracy as an automatic full-image mass segmentation system. Extensive experiments reveals robustness against the diversity of size, shape and appearance of breast masses, towards better interaction-free computer-aided diagnosis.
EN
Purpose: In clinical practice, motor development in infants is assessed subjectively. Many researchers propose objective methods, which have numerous limitations, by attaching markers or sensors to the child’s limbs. The purpose of this study is to attempt to develop objectified numerical indices to describe the limb movements of infants without interfering with spontaneous activity. Methods: 20-minute video recordings of three infants’ movements who were purposively selected from 51 subjects were included in the study. The procedure of automatic calculation of head position time in 3 positions was applied. Movement features were determined to allow for the delineation of coefficients describing the movement in numerical values. Results: Presented parameters describe three infant’s movement aspects: quality (strength), distribution of postural tonus and asymmetry in relation to head position, described as four independent values. Estimated parameters variability over time was weighted up according to expert observations. The presented method is a direct reflection of infants' observation, currently performed by highly educated and experienced therapists. Conclusions: The interpretability and usefulness of the presented parameters were proved. All parameters estimation is fully automated. The conducted research is a prelude to future work related to creating an objective and repeatable tool, initially monitoring and ultimately supporting early diagnosis for differentiating normal and abnormal motor development.
EN
(Aim) Abnormal breast can be diagnosed using the digital mammography. Traditional manual interpretation method cannot yield high accuracy. (Method) In this study, we proposed a novel computer-aided diagnosis system for detecting abnormal breasts in mammogram images. First, we segmented the region-of-interest. Next, the weighted-type fractional Fourier transform (WFRFT) was employed to obtain the unified time-frequency spectrum. Third, principal component analysis (PCA) was introduced and used to reduce the spectrum to only 18 principal components. Fourth, feed-forward neural network (FNN) was utilized to generate the classifier. Finally, a novel algorithm-specific parameter free approach, Jaya, was employed to train the classifier. (Results) Our proposed WFRFT + PCA + Jaya-FNN achieved sensitivity of 92.26% ± 3.44%, specificity of 92.28% ± 3.58%, and accuracy of 92.27% ± 3.49%. (Conclusions) The proposed CAD system is effective in detecting abnormal breasts and performs better than 5 state-of-the-art systems. Besides, Jaya is more effective in training FNN than BP, MBP, GA, SA, and PSO.
PL
Starcze zwyrodnienie plamki żółtej (AMD) jest chorobą cywilizacyjną XXI wieku. Charakteryzuje się tworzeniem nowych patologicznych naczyń krwionośnych, jak również ucieczką elementów morfotycznych i białek poza naczynia już istniejące, co skutkuje zapoczątkowaniem procesu zapalnego. Nieleczenie tej choroby może prowadzić do ślepoty. Ze względu na częste występowanie AMD i brak skutecznego leczenia farmakologicznego prowadzone są liczne badania w celu ulepszenia istniejących i wynalezienia nowych metod diagnostyki i leczenia tej choroby. Przeprowadzone badania kliniczne potwierdziły, że angiografia fluoresceinowa umożliwia obserwowanie postępu AMD. Dostrzegając możliwości komputerowej analizy wyników angiografii, w ramach niniejszej pracy zaprojektowano program komputerowy, który daje możliwość analizy wyników badań wykonanych za pomocą angiografii fluoresceinowej oraz testu Amslera. Szybki dostęp do bazy danych chorych na AMD ułatwi lekarzom pracę z pacjentem i zaoszczędzi cenny czas. Istotnym elementem zaproponowanego rozwiązania jest możliwość gromadzenia, przechowywania oraz analizy wyników badań pacjentów chorych na AMD.
EN
Age-related macular degeneration (AMD) is a twenty-first century civilization disease. It is characterized by pathological formation of new blood vessels, as well as an escape of blood cells and proteins beyond existing vessels. If untreated, the disease can lead to blindness. Due to the frequent occurrence of AMD and the lack of effective pharmacological treatment, series of studies are conducted to improve the existing and invent new methods of diagnosis and treatment. Clinical trials have confirmed that fluorescein angiography allows observation of AMD progress. A computer program was designed exploiting computer analysis of angiography result. The program allows analyzing the results of tests carried out by fluorescein angiography and Amsler test. Quick access to a database of patients with AMD will help physicians work with the patient and save valuable time. An important element of the proposed solution is the ability to collect, store, and analyze the results of studies of patients with AMD.
5
PL
W artykule przedstawiono koncepcję komputerowego rozumienia obrazów medycznych, wykorzystującą tzw. seony do tworzenia modelu treści obrazowej istotnej diagnostycznie. Zakłada ona modelowanie efektu ‘umysłowego poznania’ na bazie percepcji informacji obrazowej w kontekście określonej wiedzy dziedzinowej (diagnostyka określonego schorzenia) oraz sytuacyjnej (wyniki badań ogólnych, specjalistycznych, innych diagnostycznych, czynniki ryzyka itp.). Celem jest integracja numerycznej reprezentacji obrazowej informacji diagnostycznej ze sformalizowanym modelem wiedzy danej dziedziny (np. w formie ontologii) oraz modelem subiektywnego procesu poznania obrazowanej rzeczywistości przez ekspertów. Wszystkie te elementy nawiązują do określonego zadania klinicznego. Taki zintegrowany model jest optymalizowany z kryterium maksymalnej ekstrakcji formalnych komponentów treści (czyli wspomniane seony), które mają decydującą rolę w rozumieniu i interpretacji obrazów. Obliczeniowe komponenty nabierają znaczenia diagnostycznego wskutek weryfikacji ich przydatności w subiektywnych testach klinicznych. Poszukiwane są jednak obliczeniowe wzorce oraz odpowiednie normy i metryki, które pozwolą wykrywać istotne komponenty w przypadku zupełnego braku ich percepcji w klasycznych formach odbioru treści obrazowej. Dostosowano prezentowaną metodę numerycznego rozumienia rejestrowanych danych obrazowych do ogólnie przyjętej metodyki komputerowego wspomagania diagnostyki (CAD) medycznej. Wybrano przy tym istotne zastosowania kliniczne, m.in. wspomaganie diagnostyki wczesnych udarów mózgu na bazie zobrazowań tomografii komputerowej (CT) oraz detekcję symptomów raka sutka w mammogramach.
EN
The concept of computer understanding of medical images was presented. So-called seons were proposed to create model of diagnostically significant image content. The effect of mental cognition was considered with ability of visual perception to recognize information in the context of specific domain knowledge (e.g. conditions of a specific disease) and clinical data (i.e. results of general or specialistic examinations, other diagnosis, risk factors, etc.). Therefore, the research purpose was integration of the numerical representation of imaged diagnostic information with the formal model of domain knowledge (i.e. ontology) and the model of image subjective cognition by experts. All these issues relate to specific clinical tasks. This integrated model is optimized with the criterion of maximum extraction of formal components of content (represented by set of seons), which have a crucial role in the understanding and interpretation of the images. Computational components become diagnostically important due to the verification of their usefulness in the subjective clinical tests. However, computational patterns and relevant norms/metrics that allow to detect the essential components for a complete lack of visual perception according to classical procedures are sought. Moreover, discussed method of understanding the recorded image data was adjusted to methodology of computer-aided diagnosis (CAD) in medicine. Two important clinical applications, including acute stroke diagnosis with the computed tomography imaging (CT), and recognition of breast cancer in mammograms were indicated to illustrate possible applications.
PL
W artykule zaprezentowano nowe podejście do automatycznego rozumienia obrazów medycznych na przykładzie zobrazowań unaczynienia wieńcowego uzyskiwanych w trakcie badań spiralną tomografią komputerową (CT). W szczególności przedstawiono próby wykorzystania lingwistycznych metod strukturalnej analizy obrazów w postaci algorytmów grafowych, wykorzystywanych do tworzenia systemów wspomagania diagnostyki medycznej, a także kognitywnej analizy i rozumienia zobrazowań medycznych tętnic wieńcowych serca. Uzyskane wyniki potwierdzają duże znaczenie zaproponowanych rozwiązań w diagnostyce choroby niedokrwiennej serca.
EN
The paper presents a novel approach to analysis of CT (computed tomography) coronary artery images based on automatic image understanding paradigm. In particular there will be presented attempts at using linguistic methods of structural image analysis in the form of graph algorithms to develop a new type of systems for the cognitive analysis and understanding of images. Such methodology will be described on an example of detection of pathological changes in coronary arteries of the heart. The problem undertaken is important because the identification and location of significant stenoses in coronary vessels is a widespread practical task. The first section describes the current state of computer-assisted therapeutic decisions taken by the doctors. The second section shows the difficulties faced by developers of systems supporting the work of diagnosing physicians. The third section describes in detail the next steps in modeling and then searching for lesions in coronary arteries (Fig. 2). The fourth section presents analysis of the effectiveness of the proposed solutions which in the set of imaging data reached about 85%. The summary presents the advantages of this technique, in particular graph languages for describing shape features that can effectively be used for modeling and semantic descriptions of occurring pathological changes. The obtained results confirm the importance of the proposed methods in the diagnosis of coronary heart disease.
EN
This paper presents 15 texture features based on GLCM (Gray-Level Co-occurrence Matrix) and GLRLM (Gray-Level Run-Length Matrix) to be used in an automatic computer system for breast cancer diagnosis. The task of the system is to distinguish benign from malignant tumors based on analysis of fine needle biopsy microscopic images. The features were tested whether they provide important diagnostic information. For this purpose the authors used a set of 550 real case medical images obtained from 50 patients of the Regional Hospital in Zielona Góra. The nuclei were isolated from other objects in the images using a hybrid segmentation method based on adaptive thresholding and kmeans clustering. Described texture features were then extracted and used in the classification procedure. Classification was performed using KNN classifier. Obtained results reaching 90% show that presented features are important and may significantly improve computer-aided breast cancer detection based on FNB images.
PL
W ostatnich latach coraz większy nacisk kładziony jest na poprawę jakości oraz skuteczności opieki medycznej. Aby sprostać temu zadaniu powstaje coraz więcej nowych, bądź nowszej generacji urządzeń obrazowej diagnostyki medycznej. Rosnąca liczba badań wykonywanych dla każdego pacjenta powoduje, że lekarze mają do czynienia z coraz większą liczbą obrazów diagnostycznych skojarzonych z danym pacjentem. Aby sprostać oczekiwaniom zmniejszenia liczby błędów medycznych, poprawy efektywności interpretacji licznych zbiorów danych obrazowych oraz usprawnienia dostępu i wymiany informacji, konieczne jest wykorzystanie zaawansowanych i komputerowych metod wspomagania diagnostyki medycznej. W artykule zaprezentowano autorskie rozwiązania w zakresie wspomaganej komputerowo interpretacji zmian patologicznych, uwidacznianych na obrazach pochodzących z badań diagnostycznych tętnic wieńcowych serca. Wskazano także dalsze kierunki badawcze w tym zakresie, które będą rozwijane w niedalekiej przyszłości.
EN
In recent years, increasing emphasis is put on improving the quality and effectiveness of healthcare. To meet this challenge there are more new, or newer generation of medical diagnostic imaging equipment. A growing number of diagno stic examination causes that doctors have to deal with an increasing number of diagnostic images associated with a given patient. To meet the expectations of reducing medical errors, improving the efficiency of the interpretation of numerous sets of visual data, and improving access, and exchange of information, it is necessary to use computer-aided diagnostic methods. This article presents both original solutions in the field of computer-aided interpretation of pathological changes visible in images obtained from diagnostic examination of coronary arteries, as well as further research directions in this area.
EN
This work presents some conceptual improvements in assistance of acute stroke diagnosis with Stroke Monitor - computer-aided diagnosis tool developed and elaborated by Telemedicine Group from Institute of Radioelectronics, Warsaw University of Technology. Based on statistical analysis of common error sources we proposed some ideas of improvement capabilities for false positive errors reduction. Simulation and experimental verification confirmed validity of further development directions.
PL
Sformułowano paradygmat wspomagania diagnozy w konwencji GAD, wykorzystujący uproszczone reprezentacje danych obrazowych. Na bazie założeń przyjmujących konieczność wstępnej redukcji różnego typu nadmiarowości, jakie występują w danych źródłowych (nawiązanie do teorii compressive sensing), zaproponowano uproszczone formy analizy i rozumienia danych, prowadzące do przejrzystych efektów ekstrakcji informacji ukrytej. Na przykładzie diagnostyki wczesnego udaru mózgu ukazano przydatność koncepcji upakowanej, rzadkiej reprezentacji danych pozwalającej znacząco uprościć opis informacji.
EN
Paradigm of computer assisted diagnosis according to CAD concept was proposed. Compressive Information extraction was suggested as a result of data representation design adjusted to semantically important signal features. Ali redundancies of the image-oriented diagnostic procedure were ignored in calculated Information description. Further data analysis and understanding to extract subtle or hidden content is simplified and generally more effective. An example of acute stroke diagnosis was used to present capabilities of the paradigm implementation. Sparse OT data representation was used to extract tissue density distribution visualized in simplified forms of semantically diversified regions.
EN
The goal of this article is to present the ways of taking the methodology that has been developed earlier for the needs of automatic understanding of 2D medical visualisations to the level of 3D visualisations. The research will cover the potential of defining, in these visualisations, of elements describing the morphology of pathologic changes and the possibility of applying the linguistic approach as well as cognitive image analysis for the description and modelling of spatially visualised coronary vessels. The considerations will make reference especially to the 3D structure of the coronary vascular tree, which will be described with the use of graph-based languages of shape feature description.
PL
Celem artykułu jest przedstawienie sposobów rozwinięcia metodyki wcześniej opracowanej dla potrzeb automatycznego rozumienia dwuwymiarowych zobrazowań medycznych na zobrazowania 3D. Badana dają możliwości zdefiniowania w tych zobrazowaniach elementów opisu morfologii zmian patologicznych oraz możliwości zastosowania lingwistycznego podejścia, a także kognitywnej analizy obrazów do opisu i modelowania przestrzennych struktur biomedycznych. Rozważania w warstwie szczegółowej nawiązywać będą do trójwymiarowej struktury drzewa naczyń wieńcowych, która opisywana będzie z wykorzystaniem grafowych języków opisu cech kształtów.
PL
Praca dotyczy komputerowych metod wspomagania diagnostyki udaru niedokrwiennego mózgu wykorzystującej badania tomografii komputerowej bez podania środka kontrastowego. Jako najbardziej wiarygodny, czuły i specyficzny wskaźnik obecności tych patologii wy-brano obszar hipodensyjny odpowiadający strefie objętej niedokrwieniem. Przedstawiono koncepcję Monitora Udaru, jako narzędzia wspierającego interpretację badań TK przez: poprawę percepcji subtelnych zmian gęstości w obszarach podatnych na udar, segmentację wybranych struktur diagnostycznych, a także mniej lub bardziej wyrazistą ekstrakcję potencjalnych zmian niedokrwiennych - obszarów udarowych. Proponowana metoda monitora bazuje na wydzieleniu tkanek mózgowia, wyznaczeniu obszarów potencjalnego udaru, analizie wieloskalowej informacji obrazowej ze wzmocnieniem cech użytecznych oraz redukcją szumów i artefaktów, a także dobranej formy wizualizacji przetworzonych obrazów, uwzględniającej znaczenie wydzielonych regionów i określonych cech obrazów. Wyniki przeprowadzonych eksperymentów potwierdzają użyteczność monitora w opisie badań pacjentów w nadostrej fazie udaru.
EN
This paper presents paradigm of stroke monitor aimed to make visible acute ischemic signs in unenhanced CT brain imaging. The importance of as early as possible brain infarct detection to make possible successful thrombolytic therapy was underlined. The rationales to make possible the hyperacute hypodensity differentiation were analyzed. Subtle tissue attenuation changes were investigated, denoised, extracted and visualized. Three-staged algorithm was based on soft tissue extraction and segmentation of stroke-susceptible regions, multiscale image processing with noise/artifacts reduction and local contrast enhancement followed by adaptive visualization of extracted structures. Flexible and useful interface with possible extensions was proposed to manage examination protocol. Experimental verification confirmed usefulness of stroke monitor for educational and acute diagnosis purposes.
EN
Although mammography is a standard of reference for detection of early breast cancer, as many as 25% of breast cancers may be missed. To reduce the possibility of missing a cancer, the following methods and tools have been proposed: continuing education and training, prospective double reading, retrospective evaluation of missed cases, and use of computer-aided detection (CAD). The purpose of the reported work was to evaluate the usefulness and the potential of our aiding tools: an ontology driven editor for mammographic lesion description (MammoEdit) and a CAD-tool (Mammo Viewer) to enhance radiologist's diagnostic performance. To this end test sample of mammograms was analyzed twice, without and with aiding tools. The obtained data were analyzed using (ROC) analysis and Kappa statistics. Statistical analysis of the test data demonstrated potential of both tools to enhance radiologist's diagnostic performance.
PL
W pracy przedstawiono nową metodę opisu tekstur, przystosowaną do analizy grupy obrazów, przedstawiających na różne sposoby ten sam fragment organu. Charakteryzując obszary zainteresowania, uwzględniono nie tylko cechy teksturalne wyliczone na ich podstawie, ale również ich zależność od warunków pozyskiwania obrazów. Zaproponowano kilka sposobów konstrukcji przestrzeni parametrów odzwierciedlających zmianę tekstury, która zachodzi pod wpływem zmian warunków akwizycji. Proponowaną metodę zweryfikowano doświadczalnie w klasyfikacji obrazów tomograficznych wątroby. Rozpoznawano cztery typy tkanki, dla każdego przypadku rozważono trzy momenty akwizycji, związane z obecnością i propagacją środka kontrastującego. Wyniki uzyskane przy użyciu różnych zestawów cech teksturalnych i klasyfikatora w postaci dipolowych drzew decyzyjnych pokazują, że uwzględnienie zmian tekstury pod wpływem propagacji środka kontrastującego znacznie poprawia diagnozę.
EN
In the work, a new method of texture characterization from multiple scan series is presented. Images with the same slice position, acquired at different conditions, are analyzed simultaneously. Thereby not only texture characteristics of the considered region of interest are taken into account, but also their variations over the different acquisition moments. A few approaches to description of these variations were proposed. They were applied in recognition of four types of hepatic tissue. Liver CT images were acquired during the three typical phases related to presence and propagation of contrast material. Experiments with various sets of texture parameters and dipolar decision tree as a classifier showed that simultaneous analysis of texture features derived from three subsequent acquisition moments could considerably improve the classification accuracy.
15
Content available remote Computer-aided interpretation of medical images: mammography case study
EN
This paper presents the current limitations and challenges of computer-aided interpretation of radiological examinations. The analysis and the proposed improvements in interpretation arose from our experience, knowledge and observations with the collected suggestions and conclusions. The emphasized topics are as follows: computer understanding of human determinants of diagnosis, characteristics and enhancement of observer performance, diagnostic accuracy measures of image examinations, computer-aided diagnosis (CAD) systems, and numerical description of medical image-based content. All of these diagnosis support concepts can be integrated into an intelligent diagnosis interface and enhanced, basing on a formal description of semantic image content, i.e. ontology implied as a reliable, dynamic platform of medical knowledge, useful for diagnosis. CAD for mammography and content-based image indexing supported by the ontology were integrated for the needs of an enhanced diagnostic workstation applied in tele-information medical systems. A design of an effective human-machine interface has arisen as the leading problem of the current challenges.
EN
Although mammography is the standard of reference for the detection of early breast cancer, as many as 25% of breast cancers may be missed. To reduce the possibility of missing a cancer, the following methods and tools has been proposed: continuing education and training, prospective double reading, retrospective evaluation of missed cases, and use of computer-aided detection (CAD). In the presented paper we report on preliminary results of reducing the number of false-negative cases in mammograms interpretation by using ontology-driven editor for mammograms description, and MammoViewer, a CAD tool for radiologists' perception improvement. The use of editor resulted in reduction of interpretation errors and improved consistency of diagnosis. Computerized image processing methods make the signs of pathologies more conspicuous and so resulted in improvement of lesion perception.
PL
Komputerowe wspomaganie diagnostyki (CAD) medycznej w zakresie analizy i oceny zmian patologicznych widocznych w obrazach klinicznych staje się niezbędne w codziennej praktyce radiologicznej. Ilość danych oraz tryb dostępu do nich zwiększają zapotrzebowanie na komputerowe wsparcie procesu diagnostycznego. Różnorodność procesów diagnostycznych wyznacza różne konfiguracje, w których wykorzystywane są aplikacje analizy obrazu i klasyfikacji wzorca. Autonomiczna stacja CAD jest rozbudowana do systemu CAD, w którym dane oraz aplikacja analizy obrazów dostępne są lokalnie lub zdalnie. Integracja z systemem archiwizacji i transmisji obrazów oraz ze szpitalnym systemem informacyjnym umożliwia wykorzystanie zgromadzonych w nim danych oraz. infrastruktury informatycznej do transmisji informacji.
EN
Computer aided diagnosis (CAD) in an analysis and assessment of pathological abnormalities in medical imaging is a common procedure in clinical environment. The amount of data and the access mode increase the needs for a computer support at the diagnostic procedure. The variability of diagnostic procedures sets the requirements for the informatics infrastructure in which the application program, responsible for the image analysis and pattern recognition, are implemented. A plug-in CAD work-station is extended to a CAD system in which the database as well as the application software are accessed locally or remotely. The integration with clinical Picture Archiving and Communication System (PACS) lakes the advantages of a large image database for testing and evaluation of the application software and the informatics infrastructure is employed for data communication. Implementation of Web technology permits a CAD server to be connected with clinical workstations used by physicians and accessed remotely.
18
Content available remote Picture languages in machine understanding of medical vizualization
EN
This paper presents theoretical fundamentals and application of context-free and graph languages for cognitive analysis of selected medical visualization. It shows new opportunities for applying these methods automatic understanding of semantic contents of images in intelligent medical information system. A successful extraction of the crucial semantic content of medical image may contribute considerably to the creation of new intelligent cognitive systems, or medical computer vision systems. Thanks to the new idea of cognitive resonance between a stream of the data extracted from the image using linguistic methods, and expectations following from the language representation of the medical knowledge, it is possible to understand the subject-oriented content of the visual data. This article shows that structural techniques of soft-computing may be applied in automatic classification and machine perception based on semantic pattern content in order to determine the semantic meaning of the patterns.
19
Content available remote Picture Languages in Automatic Radiological Palm Interpretation
EN
The paper presents a new technique for cognitive analysis and recognition of pathological wrist bone lesions. This method uses AI techniques and mathematical linguistics allowing us to automatically evaluate the structure of the said bones, based on palm radiological images. Possibilities of computer interpretation of selected images, based on the methodology of automatic medical image understanding, as introduced by the authors, were created owing to the introduction of an original relational description of individual palm bones. This description was built with the use of graph linguistic formalisms already applied in artificial intelligence. The research described in this paper demonstrates that for the needs of palm bone diagnostics, specialist linguistic tools such as expansive graph grammars and EDT-label graphs are particularly well suited. Defining a graph image language adjusted to the specific features of the scientific problem described here permitted a semantic description of correct palm bone structures. It also enabled the interpretation of images showing some in-born lesions, such as additional bones or acquired lesions such as their incorrect junctions resulting from injuries and synostoses.
EN
Microcalcifications are one of more important signs enabling detection of breast cancer at an early stage. The main goal of the research was designing and realization of a system for automatic detection and classification of microcalcifications, taking advantage of the proposed automatic feature selection algorithm. The first step of the detection algorithm is to segment the individual objects : potential microcalcifications. This is achieved by applying opening by reconstruction top-hat technique and image thresholding based on approximation of an image local histogram with a probability density function of Gauss distribution. Selected features of the segmented objects are used as inputs to neural networks. The first classifier verifies the initial detection and the others assess a diagnosis of the input objects. The algorithm results are locations of suggested microcalcifications and optionally automatic diagnosis. The presented form of the system was verified in clinical tests using diagnosed databases (DDSM from the University of South Florida and own digitised database of mammograms). The achieved results are promising and comparable with other known systems. Efficiency of microcalcifications detection was up to 90%.
first rewind previous Strona / 2 next fast forward last
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.