Background and Objective: The global population has been heavily impacted by the COVID-19 pandemic of coronavirus. Infections are spreading quickly around the world, and new spikes (Delta, Delta Plus, and Omicron) are still being made. The real-time reverse transcription-polymerase chain reaction (RT-PCR) is the method most often used to find viral RNA in a nasopharyngeal swab. However, these diagnostic approaches require human involvement and consume more time per prediction. Moreover, the existing conventional test mainly suffers from false negatives, so there is a chance for the virus to spread quickly. Therefore, a rapid and early diagnosis of COVID-19 patients is needed to overcome these problems. Methods: Existing approaches based on deep learning for COVID detection are suffering from unbalanced datasets, poor performance, and gradient vanishing problems. A customized skip connection-based network with a feature union approach has been developed in this work to overcome some of the issues mentioned above. Gradient information from chest X-ray (CXR) images to subsequent layers is bypassed through skip connections. In the script’s title, ‘‘SCovNet” refers to a skip-connection-based feature union network for detecting COVID-19 in a short notation. The performance of the proposed model was tested with two publicly available CXR image databases, including balanced and unbalanced datasets. Results: A modified skip connection-based CNN model was suggested for a small unbalanced dataset (Kaggle) and achieved remarkable performance. In addition, the proposed model was also tested with a large GitHub database of CXR images and obtained an overall best accuracy of 98.67% with an impressive low false-negative rate of 0.0074. Conclusions: The results of the experiments show that the proposed method works better than current methods at finding early signs of COVID-19. As an additional point of interest, we must mention the innovative hierarchical classification strategy provided for this work, which considered both balanced and unbalanced datasets to get the best COVID-19 identification rate.
2
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
With the onset of the COVID-19 pandemic, the automated diagnosis has become one of the most trending topics of research for faster mass screening. Deep learning-based approaches have been established as the most promising methods in this regard. However, the limitation of the labeled data is the main bottleneck of the data-hungry deep learning methods. In this paper, a two-stage deep CNN based scheme is proposed to detect COVID-19 from chest X-ray images for achieving optimum performance with limited training images. In the first stage, an encoder-decoder based autoencoder network is proposed, trained on chest X-ray images in an unsupervised manner, and the network learns to reconstruct the X-ray images. An encoder-merging network is proposed for the second stage that consists of different layers of the encoder model followed by a merging network. Here the encoder model is initialized with the weights learned on the first stage and the outputs from different layers of the encoder model are used effectively by being connected to a proposed merging network. An intelligent feature merging scheme is introduced in the proposed merging network. Finally, the encoder-merging network is trained for feature extraction of the X-ray images in a supervised manner and resulting features are used in the classification layers of the proposed architecture. Considering the final classification task, an EfficientNet-B4 network is utilized in both stages. An end to end training is performed for datasets containing classes: COVID-19, Normal, Bacterial Pneumonia, Viral Pneumonia. The proposed method offers very satisfactory performances compared to the state of the art methods and achieves an accuracy of 90:13% on the 4-class, 96:45% on a 3-class, and 99:39% on 2-class classification.
3
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
Corona virus disease-2019 (COVID-19) is a pandemic caused by novel coronavirus. COVID-19 is spreading rapidly throughout the world. The gold standard for diagnosing COVID-19 is reverse transcription-polymerase chain reaction (RT-PCR) test. However, the facility for RT-PCR test is limited, which causes early diagnosis of the disease difficult. Easily available modalities like X-ray can be used to detect specific symptoms associated with COVID-19. Pre-trained convolutional neural networks are widely used for computer-aided detection of diseases from smaller datasets. This paper investigates the effectiveness of multi-CNN, a combination of several pre-trained CNNs, for the automated detection of COVID-19 from X-ray images. The method uses a combination of features extracted from multi-CNN with correlation based feature selection (CFS) technique and Bayesnet classifier for the prediction of COVID-19. The method was tested using two public datasets and achieved promising results on both the datasets. In the first dataset consisting of 453 COVID-19 images and 497 non-COVID images, the method achieved an AUC of 0.963 and an accuracy of 91.16%. In the second dataset consisting of 71 COVID-19 images and 7 non-COVID images, the method achieved an AUC of 0.911 and an accuracy of 97.44%. The experiments performed in this study proved the effectiveness of pre-trained multi-CNN over single CNN in the detection of COVID-19.
4
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
Rapid and accurate detection of COVID-19 coronavirus is necessity of time to prevent and control of this pandemic by timely quarantine and medical treatment in absence of any vaccine. Daily increase in cases of COVID-19 patients worldwide and limited number of available detection kits pose difficulty in identifying the presence of disease. Therefore, at this point of time, necessity arises to look for other alternatives. Among already existing, widely available and low-cost resources, X-ray is frequently used imaging modality and on the other hand, deep learning techniques have achieved state-of-the-art performances in computer-aided medical diagnosis. Therefore, an alternative diagnostic tool to detect COVID-19 cases utilizing available resources and advanced deep learning techniques is proposed in this work. The proposed method is implemented in four phases, viz., data augmentation, preprocessing, stage-I and stage-II deep network model designing. This study is performed with online available resources of 1215 images and further strengthen by utilizing data augmentation techniques to provide better generalization of the model and to prevent the model overfitting by increasing the overall length of dataset to 1832 images. Deep network implementation in two stages is designed to differentiate COVID-19 induced pneumonia from healthy cases, bacterial and other virus induced pneumonia on X-ray images of chest. Comprehensive evaluations have been performed to demonstrate the effectiveness of the proposed method with both (i) training-validation-testing and (ii) 5- fold cross validation procedures. High classification accuracy as 97.77%, recall as 97.14% and precision as 97.14% in case of COVID-19 detection shows the efficacy of proposed method in present need of time. Further, the deep network architecture showing averaged accuracy/ sensitivity/specificity/precision/F1-score of 98.93/98.93/98.66/96.39/98.15 with 5-fold cross validation makes a promising outcome in COVID-19 detection using X-ray images.
5
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
This paper aims to propose a deep learning model based on big data for the healthcare system to predict social network data. Social network users post large amounts of healthcare information on a daily basis and at the same time hospitals and medical laboratories store very large amounts of healthcare data, such as X-rays. The authors provide an architecture that can integrate deep learning, social networks, and big data. Deep learning is one of the most challenging areas of research and is becoming increasingly popular in the health sector. It uses deep analysis to extract knowledge with optimum precision. The proposed architecture consists of three layers: the deep learning layer, the big data layer, and the social networks layer. The big data layer includes data for health care, such as X-ray images. For the deep learning layer, three Convolution Neuronal Network models are proposed for X-ray image classification. As a result, social network layer users can access the proposed system to predict their X-ray image posts.
Wprowadzenie: Co łączy ze sobą światowe sławy: piłkarza Roberta Lewandowskiego, kolarkę górską Maję Włoszczowską, tenisistkę Urszulę Radwańską, skoczka narciarskiego Kamila Stocha i lekkoatletkę Anitę Włodarczyk? Cała wyżej wymieniona grupa sportowców ma na swoim koncie urazy stawu skokowego, bo urazy tej okolicy dotykają przede wszystkim ludzi aktywnych fizycznie. Cel: Celem pracy była weryfikacja prawidłowości wykonywania zdjęć rentgenowskich stawu skokowego oraz analiza danych zawartych w skierowaniach przez dwie grupy badaczy: techników elektroradiologii i studentów kierunku elektroradiologia. Materiał i metody: Analizie retrospektywnej poddano skierowania i zdjęcia rentgenowskie stawu skokowego wykonane 105 pacjentom w dwóch projekcjach: przednio-tylnej (AP) i bocznej (LAT) w Zakładzie Radiologii Uniwersyteckiego Szpitala Klinicznego w Białymstoku. Ocena została dokonana przez trzech techników elektroradiologii oraz trzech studentów kierunku elektroradiologia. Analizie poddano również wiek i płeć pacjentów oraz rozpoznania. Dane poddano analizie statystycznej przy użyciu programu Statistica 13.1 oraz MS Excel 2010. Wyniki: Spośród 105 radiogramów objętych analizą 58 (55,24%) wykonano kobietom. Wśród radiogramów wykonanych w projekcji bocznej brak było określenia strony badanej pacjenta w 47 (44,76%) przypadkach. W sposób istotny statystycznie różniły się od siebie oceny zakresu (p = 0,01) i poprawności pozycjonowania (p = 0 ,01) radiogramów w ykonanych w projekcji A P. Istnieje istotny statystycznie związek pomiędzy zastosowaniem u pacjenta opatrunku gipsowego a oceną przez techników poprawności projekcji przednio-tylnej stawu skokowego (p = 0,01). W grupie studentów wykazano związek pomiędzy opatrunkiem gipsowym a oceną zakresu radiogramów w projekcji bocznej (p = 0,04) oraz zgodnością projekcji bocznej (p = 0,01). Wnioski: 1. Wykazano różnice w ocenie radiogramów pod kątem zakresu oraz pozycjonowania pomiędzy grupą techników elektroradiologii i studentów. 2. Studenci rzadziej akceptowali zakres w ocenianych zdjęciach szczególnie w projekcji przednio-tylnej. 3. Technicy częściej akceptowali poprawność projekcji przednio-tylnej u pacjentów z opatrunkiem gipsowym.
EN
Introduction: What units the most famous sportsmen in the world? They were all subjected to ankle joint injuries, because this trauma is common among physically active people. Purpose: The main goal of this study was the analysis data in referrals and verification of the correctness of making ankle x-rays, performed by 2 groups of researchers: radiographers and radiography students. Materials and methods: Referrals a nd a nkle x-rays i n 2 projections: anterior-posterior (AP) and lateral (LAT) were retrospectively analysed. Radiograms of 105 patients from Radiology Department in Uniwersytecki Szpital Kliniczny in Białystok were used. Evaluation was performed by 3 radiographers and 3 radiology students. Age and sex of the patients and diagnosis were analysed too. Data were statistically analysed with Statistica 13.1 and MS Excel 2010. Results: 58 from 105 analysed x-rays (55,24%) were made among women. The anatomical side was not indicated on lateral radiographs in 47 cases (44,76%). Statistically significant differences were shown in evaluations of anatomical coverage assessment ( p = 0,01) and correctness of positioning ( p = 0,01) in AP x-rays. A statistically significant association between having a plaster cast and evaluation of correctness making anterior- posterior projection on ankle x-ray was observed (p = 0,01). In students group statistically significant association between having a plaster cast and evaluation of anatomical coverage assessment in lateral x-rays (p = 0,04) and compatibility of lateral projection with the guidelines (p = 0,01) was observed. Conclusions: 1. Statistically significant differences in the assessment of radiographs with regard to anatomical coverage and positioning between a group of radiology technicians and students were shown. 2. Students less frequently accepted range in evaluated images, especially in the anterior-posterior projection. 3. Technicians more often accepted the correctness of anterior- posterior projections in patients with a plaster cast.
7
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
The paper presents to aspects of measuring disturbances in the selected measurement systems based on digital radiographic images. Systems using X-ray images of the ankle or hip joint and dedicated programs and measurements algorithms were selected. The author performed an analysis of the standard uncertainty of measurement. Significant differences in uncertainty levels for the various indicators were observed, and must therefore be interpreted independently.
PL
W artykule zostaną przedstawione aspekty zakłóceń pomiarów w wybranych systemach pomiarowych dotyczących zdjęć rentgenowskich. Do badań wybrano systemy wykorzystujące zdjęcia rentgenowskie stawów skokowych i biodrowych oraz własne programy i algorytmy pomiarowe. Przeprowadzono analizę poziomu niepewności standardowej pomiarów. Zaobserwowano znaczne różnice w poziomie niepewności dla różnych wskaźników i dlatego należy je interpretować niezależnie.
8
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
Wprowadzenie. Zdjęcie RTG porównawcze rąk jest podstawowym badaniem obrazowym w rozpoznaniu i ocenie procesu leczenia reumatoidalnego zapalenia stawów rąk. Objawy radiologiczne często są bardzo dyskretne, co wymaga bardzo dobrej jakości obrazów w zakresie kontrastu. Kontrastowość obrazu rentgenowskiego związana jest z niejednorodnością wiązki promieniowania emitowanego przez lampę rentgenowską. Cel. Celem badania była retrospektywna analiza anodowego efektu osłabienia promieniowania rentgenowskiego na radiogramach porównawczych rąk. Materiał i metoda. Materiał badawczy stanowiły 154 zdjęcia rentgenowskie porównawcze rąk. Dokonano analizy retrospektywnej radiogramów w zakresie oceny wieku i płci pacjentów, pozycjonowanie rąk względem osi lampy RTG, wielkości pola kolimacji, wartości Dose Area Product, jasności pikseli tła w linii pośrodkowej radiogramu. Do analizy między zmiennymi zastosowano test U Manna-Whitneya oraz test korelacji rang Spearmana, poziom istotności p < 0,001. Wyniki. Najczęściej radiogramy były wykonane kobietom (84%) w wieku od 22 do 82 lat. Tylko 4 (2,6%) radiogramy zostały wykonane bez zachowania schematu katoda-nadgarstek/anoda-palce. Najczęściej radiogramy były wykonane przy polu kolimacji w przedziale 701-800 cm2 oraz DAP 2,6-3 cGy·cm2. Udowodniono silną zależność (R = 0,63) pomiędzy wymiarem długości boku pola kolimacji a różnicą jasności pikseli mierzoną w ustalonych punktach pomiaru oraz przeciętną zależność (R = 0,43) pomiędzy polem kolimacji a różnicą jasności pikseli.
EN
Introduction and aim of the study. X-ray examination is the primary imaging technique in the diagnosis and evaluation of the treatment of rheumatoid arthritis. Radiological symptoms are often very discreet, what requires a very good image quality in terms of contrast. X-ray contrast image is related to the inhomogeneity of the X-ray beam. Aim. The aim of the study was a retrospective analysis of the heel effect on X-ray hands. Material and methods. Material consisted of 154 radiographs. The retrospective analysis of radiographs was made, within the range of age and sex of patients, the positioning of hands relative to the axis of the X-ray tube, size of the collimation, Dose Area Product, brightness of the background pixels in the midline radiograph. The Mann-Whitney U-test and Spearman’s rank correlation (p < 0.001), was used in analysis. Results. Most frequently, the radiographs were made of women (84%) between the ages of 22 to 82 years. Only 4 (2,6%) radiographs were taken without pattern „wrist-cathode/fingers-anode”. Most frequently, the radiographs were made with the collimation in the range of 701-800 cm2 and DAP 2,6-3 cGy·cm2. The strong correlation (R = 0.63) between the side length of the collimation field and the difference pixel brightness, and the average correlation (R = 0.43) between the size collimation field and the difference pixel brightness, were proven.
9
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
In the paper is presented methodology of the X-ray image processing application to investigate gravitational flow in rectangular silo model. The proposed normalization procedure of X-ray data allows to visualize the changes of the volume fraction of sand during silo discharging process. The applied procedure of image processing, in contrast to the previously author works, allows to obtained more accurate information about the changes of material distribution level during process. The conducted image analysis simplifies the investigation of mass flow in various area of silo. The obtained results show the different particle behaviour in centre and at silo wall area. The experiments were conducted for different initially level of sand densities and roughness of the silo wall. Visualization of dissimilarity in interaction between the particles and particles, and between particles and the silo walls, even for smooth wall, was the main result of the Xray image analysis, especially for shear zone visualization.
For over a century, X-ray radiation has played an important role in the area of the conservation and restoration of cultural heritage objects. X-ray techniques are amongst the most fundamental and helpful methods used in the investigation of art works. This paper reviews the application of traditional radiography, X-ray dual source computed tomography (DSCT) and scanning electron microscopy combined with energy dispersed X-ray spectroscopy (SEM-EDX) to the investigation of a wooden, Gothic sculpture, The risen Christ. Thanks to the properties of X-ray radiation (different absorption by various materials) first two methods allow the assessment of the preservation state and the observation of the internal structure of an object in 3-D. While SEM-EDX analysis permits the elemental analysis of the polychrome layers. As a result 2-D and 3-D images, permitting the full volume inspection of an object, were taken in a totally non-destructive way. The morphological and physical information about the inner structure of the investigated wooden sculpture was obtained, revealing changes related to previous restorations, as well as ageing effects. Employing the SEM-EDX, painting materials (pigments and filers), were identified. Gained data is essential for restorers to understand the whole structure of the studied object and to decide which further investigation and restoration steps have to be undertaken.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.