UBNet: Deep learning-based approach for automatic X-ray image detection of pneumonia and COVID-19 patients

Widodo, Chomsin S.; Naba, Agus; Mahasin, Muhammad M; Yueniwati, Yuyun; Putranto, Terawan Agus; Patra, Pangeran I

doi:10.3233/xst-211005

Cited by 10 publications

(6 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This research develops an image segmentation architecture to obtain a lighter but more accurate image segmentation architecture. The UNet image segmentation architecture is modified by changing the layer arrangement to resemble UBNet [14]. UBNet is a CNN architecture developed explicitly to process chest X-ray images of pneumonia and Covid-19 patients.…”

Section: Discussionmentioning

confidence: 99%

“…This study developed a deep learning model for opacity segmentation in X-ray images of Covid-19 patients by modifying the UNet CNN architecture [13]. The modification is done by rebuilding the UNet architecture based on the lightweight UBNet image classification architecture [14]. UBNet is explicitly developed to process lightweight X-ray images in the case of image classification.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Analysis of Pneumonia Area Development in Covid-19 Patients based on Sequential X-Ray Images Using the Seq_UB Architecture

Widodo,

Naba,

Yuewniwati

et al. 2023

Trends Sci

View full text Add to dashboard Cite

Introduction During the covid pandemic, radiologists diagnosed covid patients by analyzing chest X-ray images. The existence of a quantitative system can help visual observation of lung conditions by radiologists. Purpose of research This research aims to develop an image segmentation algorithm, Seq_UB, which can monitor the development of the lung condition of Covid-19 patients periodically. The Seq_UB algorithm was developed by modifying the UNet system using different CNN layers according to the UBNet v1 algorithm. Materials and methods The dataset for training uses the Montgomery USA Dataset, which contains 138 CXRs with a resolution of 4,020×4,892, of which there are 80 normal CXR images and 58 CXR images of patients identified with tuberculosis; the Shenzhen Hospital Dataset consists of 662 CXR images, of which there are 336 abnormal CXR images. Result and discussion The results show that reducing the input image size does not significantly affect the accuracy of the segmentation results. However, reducing the input image size will affect the resolution of the segmentation results. Where the smaller the input image size, the lower the resolution obtained. This will have an impact on the final interpretation of the segmentation results. Research shows that an input image size of 512 is the best because the resolution of the segmentation results is still very accurate. This study shows that the Seq_UB architecture can perform X-ray image segmentation with relatively stable accuracy and lower computational burden. An interesting pattern was found, where Covid-19 patients quantitatively experienced fluctuations in image segmentation size. Conclusions The Seq_UB system can perform well with a segmentation accuracy of 96 %, and the processing speed takes 0.91 s. A desktop GUI was designed to segment X-ray images more effectively. HIGHLIGHTS Development of the Seq_UB algorithm: This research presents the development of an X-ray image segmentation algorithm with a computational workload that is faster, more accurate, and lighter Quantitative estimation of the lung area percentage of Covid-19 patients based on segmentation with the Seg_UB model Monitoring the development of lung conditions of Covid-19 patients quantitatively by comparing it with normal CXR using the Seg_UB model GRAPHICAL ABSTRACT

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Analysis of Pneumonia Area Development in Covid-19 Patients based on Sequential X-Ray Images Using the Seq_UB Architecture

Widodo,

Naba,

Yuewniwati

et al. 2023

Trends Sci

View full text Add to dashboard Cite

show abstract

“…The authors concluded that DenseNet-169 when combined with ideal SVM RBF kernel hyper-parameter values, outperformed all other models tested. The from-scratch techniques were employed [ 76 , 77 ]. The researchers developed a novel model for pneumonia detection; however, the model’s performance was poor, with accuracy, precision, recall, and F1_score all falling below 90%.…”

Section: Discussionmentioning

confidence: 99%

Automated Lung-Related Pneumonia and COVID-19 Detection Based on Novel Feature Extraction Framework and Vision Transformer Approaches Using Chest X-ray Images

et al. 2022

View full text Add to dashboard Cite

According to research, classifiers and detectors are less accurate when images are blurry, have low contrast, or have other flaws which raise questions about the machine learning model’s ability to recognize items effectively. The chest X-ray image has proven to be the preferred image modality for medical imaging as it contains more information about a patient. Its interpretation is quite difficult, nevertheless. The goal of this research is to construct a reliable deep-learning model capable of producing high classification accuracy on chest x-ray images for lung diseases. To enable a thorough study of the chest X-ray image, the suggested framework first derived richer features using an ensemble technique, then a global second-order pooling is applied to further derive higher global features of the images. Furthermore, the images are then separated into patches and position embedding before analyzing the patches individually via a vision transformer approach. The proposed model yielded 96.01% sensitivity, 96.20% precision, and 98.00% accuracy for the COVID-19 Radiography Dataset while achieving 97.84% accuracy, 96.76% sensitivity and 96.80% precision, for the Covid-ChestX-ray-15k dataset. The experimental findings reveal that the presented models outperform traditional deep learning models and other state-of-the-art approaches provided in the literature.

show abstract

“…Unfortunately, the number of annotated image datasets is often limited since reliable manual annotation is exceptionally time-consuming and needs experienced clinicians [4]. Hence, most of former studies in developing DL models have focused on classifying between COVID-19 and normal tissue or other forms of pneumonia cases [9][10][11][12], which has not been clinically effective [13]. Therefore, in order to overcome the difficulties of previous "black-box" type DL models, developing new CAD based on DL models to automatically segment COVID-19 infected areas with an interactive graphical user interface (GUI) is essential to increase transparency and allow radiologists to visually examine the segmented infected lesions, which can lead to quantifying the severity and spread of COVID-19 infection more accurately and robustly.…”

Section: Introductionmentioning

confidence: 99%

An observer comparison study to evaluate a machine learning model to quantify the infected pneumonia on lung CT images

Abdihamzehkolaei

Mirniaharikandehei

Choquehuanca

et al. 2023

Medical Imaging 2023: Image Perception, Observer Performance, and Technology Assessment

View full text Add to dashboard Cite

Quantification of infected lung volume using computed tomography (CT) images can play a critical role in predicting the severity of pulmonary infectious disease. Manual segmentation of infected areas from several CT image slices, however, is not efficient and viable in clinical practice. To assist clinicians in overcoming this challenge, we developed a new method to automatically segment and quantify the percentage of the infected lung volume. First, we used a public dataset of 20 COVID-19 patients, which consists of manually annotated lung and infection masks, to train a new joint deep learning (DL) model for lung and infection segmentation. As for lung segmentation, a Mask-RCNN model was applied to the lung volume with a novel postprocessing technique. Following that, an ensemble model with a customized residual attention UNet model and feature pyramid network (FPN) models was employed for infection segmentation. Next, we assembled another set of 80 CT scans of Covid-19 patients. Two chest radiologists manually evaluated each CT scan and reported the infected lung volume percentage using a customized graphical user interface (GUI). The developed DL-model was also employed to process these CT images. Then, we compared the agreement between the radiologist (manual) and model-based (automated) percentages of diseased regions. Additionally, the GUI was used to let radiologists rate acceptance of the DL-model generated segmentation results. Analyzing the results demonstrate that the agreement between manual and automated segmentation is >95% in 28 testing cases. Furthermore, >53% of testing cases received the top assessment rating scores from two radiologists (between 4-5 score). Thus, this study illustrates the feasibility of developing a DL-model based automated tool to effectively provide quantitative evaluation of infected lung regions to assist in improving the efficiency of radiologists in infection diagnosis.

show abstract

UBNet: Deep learning-based approach for automatic X-ray image detection of pneumonia and COVID-19 patients

Cited by 10 publications

References 15 publications

Analysis of Pneumonia Area Development in Covid-19 Patients based on Sequential X-Ray Images Using the Seq_UB Architecture

Analysis of Pneumonia Area Development in Covid-19 Patients based on Sequential X-Ray Images Using the Seq_UB Architecture

Automated Lung-Related Pneumonia and COVID-19 Detection Based on Novel Feature Extraction Framework and Vision Transformer Approaches Using Chest X-ray Images

An observer comparison study to evaluate a machine learning model to quantify the infected pneumonia on lung CT images

Contact Info

Product

Resources

About