Segmentation of Tuberculosis Lungs on Computer Tomography Images
نویسندگان
چکیده
Background. Tuberculosis is a chronic lung disease that occurs due to bacterial infection and one of the top ten causes human death. As part automated diagnostic system, detecting tuberculosis lesions on computed tomograms lungs in automatic mode an urgent task. Objective. We are aimed solve segmentation tuberculosis-affected areas problem computer using digital image processing based U-networks. Methods. The data for training network were provided by specialists National Institute Phthisiology Pulmonology named after F.V. Yanovsky, NAMS Ukraine. performed applying artificial intelligence convolutional neural UNet, which has been developed medical tasks. considered three versions UNet networks with different parameter values. A feature U-Net absence fully connected layers. This example encoder-decoder architecture, shows high results problems semantic segmentation. In last two models, we applied technique early stopping avoids effect overfitting network. number epochs set margin, process parameters stops as soon model performance improving test set. Results. was divided into 320 samples (80%) training, 40 (10%) testing, exam. effectiveness models evaluated parameters: Precision, Recall, Matthews correlation coefficient. final provides exam, such accuracy 0.82, sensitivity 0.75, coefficient 78%. Conclusions. conducted studies allowed us obtain tomography images. proposed will be used further development systems tuberculosis.
منابع مشابه
Automatic Lung Lobe Segmentation and Fissure Extraction Using Itempls Method on Computer Tomography Images
This paper proposes an automatic lung lobe segmentation method in a supervised way. The lung image is given as input and the lung image is segmented using iterative threshold method. Then fissure information is enhanced using log supportive filter. Then Fissure locations are extracted using Vector training and matching process. Finally the lung lobes are segmented and marked as individual lobes...
متن کاملPerformance Analysis of Segmentation of Hyperspectral Images Based on Color Image Segmentation
Image segmentation is a fundamental approach in the field of image processing and based on user’s application .This paper propose an original and simple segmentation strategy based on the EM approach that resolves many informatics problems about hyperspectral images which are observed by airborne sensors. In a first step, to simplify the input color textured image into a color image without tex...
متن کاملFuzzy Modeling of Electrical Impedance Tomography Images of the Lungs
OBJECTIVES Aiming to improve the anatomical resolution of electrical impedance tomography images, we developed a fuzzy model based on electrical impedance tomography's high temporal resolution and on the functional pulmonary signals of perfusion and ventilation. INTRODUCTION Electrical impedance tomography images carry information about both ventilation and perfusion. However, these images ar...
متن کاملPathological evaluation of computed tomography images of lungs.
A method is described which allows the features seen in images generated during computed tomography (CT) of lungs previously removed at necropsy to be compared with those seen in corresponding thin sections made from the same lungs. After removal from the body, lungs were inflation-fixed using warm formalin vapour before being x-rayed and then scanned in the inflated state. Slices corresponding...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Innovative biosystems and bioengineering
سال: 2021
ISSN: ['2616-177X']
DOI: https://doi.org/10.20535/ibb.2021.5.2.233051