Chest Radiographic Images Research Articles

Evaluating newly approved drugs in combination regimens for multidrug-resistant tuberculosis with fluoroquinolone resistance (endTB-Q): study protocol for a multi-country randomized controlled trial

BackgroundTreatment for fluoroquinolone-resistant multidrug-resistant/rifampicin-resistant tuberculosis (pre-XDR TB) often lasts longer than treatment for less resistant strains, yields worse efficacy results, and causes substantial toxicity. The newer anti-tuberculosis drugs, bedaquiline and delamanid, and repurposed drugs clofazimine and linezolid, show great promise for combination in shorter, less-toxic, and effective regimens. To date, there has been no randomized, internally and concurrently controlled trial of a shorter, all-oral regimen comprising these newer and repurposed drugs sufficiently powered to produce results for pre-XDR TB patients.MethodsendTB-Q is a phase III, multi-country, randomized, controlled, parallel, open-label clinical trial evaluating the efficacy and safety of a treatment strategy for patients with pre-XDR TB. Study participants are randomized 2:1 to experimental or control arms, respectively. The experimental arm contains bedaquiline, linezolid, clofazimine, and delamanid. The control comprises the contemporaneous WHO standard of care for pre-XDR TB. Experimental arm duration is determined by a composite of smear microscopy and chest radiographic imaging at baseline and re-evaluated at 6 months using sputum culture results: participants with less extensive disease receive 6 months and participants with more extensive disease receive 9 months of treatment. Randomization is stratified by country and by participant extent-of-TB-disease phenotype defined according to screening/baseline characteristics. Study participation lasts up to 104 weeks post randomization. The primary objective is to assess whether the efficacy of experimental regimens at 73 weeks is non-inferior to that of the control. A sample size of 324 participants across 2 arms affords at least 80% power to show the non-inferiority, with a one-sided alpha of 0.025 and a non-inferiority margin of 12%, against the control in both modified intention-to-treat and per-protocol populations.DiscussionThis internally controlled study of shortened treatment for pre-XDR TB will provide urgently needed data and evidence for clinical and policy decision-making around the treatment of pre-XDR TB with a four-drug, all-oral, shortened regimen.Trial registrationClinicalTrials.Gov NCT03896685. Registered on 1 April 2018; the record was last updated for study protocol version 4.3 on 17 March 2023.

SHAMSUL: Systematic Holistic Analysis to investigate Medical Significance Utilizing Local interpretability methods in deep learning for chest radiography pathology prediction

The interpretability of deep neural networks has become a subject of great interest within the medical and healthcare domain. This attention stems from concerns regarding transparency, legal and ethical considerations, and the medical significance of predictions generated by these deep neural networks in clinical decision support systems. To address this matter, our study delves into the application of four well-established interpretability methods: Local Interpretable Model-agnostic Explanations (LIME), Shapley Additive exPlanations (SHAP), Gradient-weighted Class Activation Mapping (Grad-CAM), and Layer-wise Relevance Propagation (LRP). Leveraging the approach of transfer learning with a multi-label-multi-class chest radiography dataset, we aim to interpret predictions pertaining to specific pathology classes. Our analysis encompasses both single-label and multi-label predictions, providing a comprehensive and unbiased assessment through quantitative and qualitative investigations, which are compared against human expert annotation. Notably, Grad-CAM demonstrates the most favorable performance in quantitative evaluation, while the LIME heatmap score segmentation visualization exhibits the highest level of medical significance. Our research underscores both the outcomes and the challenges faced in the holistic approach adopted for assessing these interpretability methods and suggests that a multimodal-based approach, incorporating diverse sources of information beyond chest radiography images, could offer additional insights for enhancing interpretability in the medical domain.

Development of Pericardial Fat Count Images Using a Combination of Three Different Deep-Learning Models: Image Translation Model From Chest Radiograph Image to Projection Image of Three-Dimensional Computed Tomography

Pericardial fat (PF)-the thoracic visceral fat surrounding the heart-promotes the development of coronary artery disease by inducing inflammation of the coronary arteries. To evaluate PF, we generated pericardial fat count images (PFCIs) from chest radiographs (CXRs) using a dedicated deep-learning model. We reviewed data of 269 consecutive patients who underwent coronary computed tomography (CT). We excluded patients with metal implants, pleural effusion, history of thoracic surgery, or malignancy. Thus, the data of 191 patients were used. We generated PFCIs from the projection of three-dimensional CT images, wherein fat accumulation was represented by a high pixel value. Three different deep-learning models, including CycleGAN were combined in the proposed method to generate PFCIs from CXRs. A single CycleGAN-based model was used to generate PFCIs from CXRs for comparison with the proposed method. To evaluate the image quality of the generated PFCIs, structural similarity index measure (SSIM), mean squared error (MSE), and mean absolute error (MAE) of (i) the PFCI generated using the proposed method and (ii) the PFCI generated using the single model were compared. The mean SSIM, MSE, and MAE were 8.56×10-1, 1.28×10-2, and 3.57×10-2, respectively, for the proposed model, and 7.62×10-1, 1.98×10-2, and 5.04×10-2, respectively, for the single CycleGAN-based model. PFCIs generated from CXRs with the proposed model showed better performance than those generated with the single model. The evaluation of PF without CT may be possible using the proposed method.

Exploring transfer learning in chest radiographic images within the interplay between COVID-19 and diabetes.

The intricate relationship between COVID-19 and diabetes has garnered increasing attention within the medical community. Emerging evidence suggests that individuals with diabetes may experience heightened vulnerability to COVID-19 and, in some cases, develop diabetes as a post-complication following the viral infection. Additionally, it has been observed that patients taking cough medicine containing steroids may face an elevated risk of developing diabetes, further underscoring the complex interplay between these health factors. Based on previous research, we implemented deep-learning models to diagnose the infection via chest x-ray images in coronavirus patients. Three Thousand (3000) x-rays of the chest are collected through freely available resources. A council-certified radiologist discovered images demonstrating the presence of COVID-19 disease. Inception-v3, ShuffleNet, Inception-ResNet-v2, and NASNet-Large, four standard convoluted neural networks, were trained by applying transfer learning on 2,440 chest x-rays from the dataset for examining COVID-19 disease in the pulmonary radiographic images examined. The results depicted a sensitivity rate of 98 % (98%) and a specificity rate of almost nightly percent (90%) while testing those models with the remaining 2080 images. In addition to the ratios of model sensitivity and specificity, in the receptor operating characteristics (ROC) graph, we have visually shown the precision vs. recall curve, the confusion metrics of each classification model, and a detailed quantitative analysis for COVID-19 detection. An automatic approach is also implemented to reconstruct the thermal maps and overlay them on the lung areas that might be affected by COVID-19. The same was proven true when interpreted by our accredited radiologist. Although the findings are encouraging, more research on a broader range of COVID-19 images must be carried out to achieve higher accuracy values. The data collection, concept implementations (in MATLAB 2021a), and assessments are accessible to the testing group.

Joint classification and segmentation for an interpretable diagnosis of acute respiratory distress syndrome from chest x-rays.

Acute respiratory distress syndrome (ARDS) is a life-threatening condition that can cause a dramatic drop in blood oxygen levels due to widespread lung inflammation. Chest radiography is widely used as a primary modality to detect ARDS due to its crucial role in diagnosing the syndrome, and the x-ray images can be obtained promptly. However, despite the extensive literature on chest x-ray (CXR) image analysis, there is limited research on ARDS diagnosis due to the scarcity of ARDS-labeled datasets. Additionally, many machine learning-based approaches result in high performance in pulmonary disease diagnosis, but their decisions are often not easily interpretable, which can hinder their clinical acceptance. This work aims to develop a method for detecting signs of ARDS in CXR images that can be clinically interpretable. To achieve this goal, an ARDS-labeled dataset of chest radiography images is gathered and annotated for training and evaluation of the proposed approach. The proposed deep classification-segmentation model, Dense-Ynet, provides an interpretable framework for automatically diagnosing ARDS in CXR images. The model takes advantage of lung segmentation in diagnosing ARDS. By definition, ARDS causes bilateral diffuse infiltrates throughout the lungs. To consider the local involvement of lung areas, each lung is divided into upper and lower halves, and our model classifies the resulting lung quadrants. The quadrant-based classification strategy yields the area under the receiver operating characteristic curve of 95.1% (95% CI 93.5 to 96.1), which allows for providing a reference for the model's predictions. In terms of segmentation, the model accurately identifies lung regions in CXR images even when lung boundaries are unclear in abnormal images. This study provides an interpretable decision system for diagnosing ARDS, by following the definition used by clinicians for the diagnosis of ARDS from CXR images.

Objective Image Quality Optimization in Augmented Reality Using Spatial Frequency Domain Models.

Augmented reality (AR) blends the digital and physical worlds by overlapping a virtual image onto the see-through physical environment. However, contrast reduction and noise superposition in an AR head-mounted display (HMD) can substantially limit image quality and human perceptual performance in both the digital and physical spaces. To assess image quality in AR, we performed human and model observer studies for various imaging tasks with targets placed in the digital and physical worlds. A target detection model was developed for the complete AR system including the optical see-through. Target detection performance using different observer models developed in the spatial frequency domain was compared with the human observer results. The non-prewhitening model with eye filter and internal noise results closely track human perception performance as measured by the area under the receiver operating characteristic curve (AUC), especially for tasks with high image noise. The AR HMD non-uniformity limits the low-contrast target (less than 0.02) observer performance for low image noise. In augmented reality conditions, the detectability of a target in the physical world is reduced due to the contrast reduction by the overlaid AR display image (AUC less than 0.87 for all the contrast levels evaluated). We propose an image quality optimization scheme to optimize the AR display configurations to match observer detection performance for targets in both the digital and physical worlds. The image quality optimization procedure is validated using both simulation and bench measurements of a chest radiography image with digital and physical targets for various imaging configurations.

A Multi-stage framework for COVID-19 detection and severity assessment from chest radiography images using advanced fuzzy ensemble technique

Chest radiography (CXR) is a popular imaging modality for screening lung abnormalities. It plays a vital role in monitoring disease progression, allocating limited medical resources effectively, and planning treatment in clinical practice. However, manual CXR image screening is time-consuming and susceptible to human errors. Therefore, computer-aided diagnosis (CAD) systems can be an additional decision-making tool for radiologists. Although prior research had remarkable success in COVID-19 classification tasks, it is still difficult to accurately localize the infection and quantify its severity, which restricts the generalizability and interpretability of CAD-based systems. This work proposes an end-to-end multi-stage architecture for COVID-19 classification, infection region identification, and disease severity assessment. The first stage of the proposed architecture identifies the relevant lung regions by eliminating the irrelevant portions. The second stage discriminates COVID-19 from pneumonia and normal images. In this study, instead of a single pre-trained model, we used three models and a fuzzy rank-based ensemble method to combine the results. Unlike the simple ensemble techniques, the proposed approach generates the final predictions for the test samples by evaluating the confidence in the predictions of the base models. In the final stage of the network, the infection regions are identified, and the amount of infection is calculated when the image is classified as COVID-19. Subsequently, a severity score ranging from 0 to 3 is assigned based on the level of infection severity. Based on the predicted scores, COVID-19 images are categorized into four severity levels: mild, moderate, severe, and critical. The final ensemble segmentation model exceeds all the individual segmentation models with an IoU score of 98.99% and a dice score of 99.49%. Additionally, the infection segmentation model also outperforms all other segmentation models individually, achieving a dice score of 96.75% and an IoU score of 93.71%. Our proposed segmentation-based classification model achieves a maximum accuracy of 98.05% with an F1-score of 97.76% and outperforms five state-of-the-art (SOTA) CNN models and five existing ensemble techniques in the literature. An external dataset also validated the model, and the results showed an accuracy of 97.63% and an F1-score of 97.22%.

Radiation dose evaluation to organs in neonatal patients by field size during potable X-ray examination in incubators: A Monte Carlo simulation study.

Neonatal patients located in incubators are exposed to as many as 159 radiographs until discharge. To reduce the dose exposed to the patient, factors that may cause unnecessary exposure to the patient were judged. When conducting portable X-rays of neonatal patients located in an incubator, it is not easy to determine the exact field size because collimation light is exposed on the acrylic plate, an incubator canopy, and the resulting shadow is reflected on the patient's body. This study aims to measure the organ dose exposed to the patient according to the field size when a portable radiograph is given to a neonatal patient in a neonatal intensive care unit (NICU) incubator. To identify the absorbed organ dose depending on the radiation field size during portable X-ray examination of neonatal patient, a Monte Carlo N-Particle (MCNP) simulation, a SpeckCalc program, and a neonatal phantom from the ICRP 89 are applied for the calculation. According to the minimal field size (MinFS) standards of the European Commission (EC), the smaller field size is intended to measure tightly from the top of the lung apices to the bottom of the genitals; a larger field size is also calculated by adding 6 cm in width and length. Compared to the hospital C condition from the previous study, the larger and smaller field sizes are decreased by an average of 45% and 67%, respectively. Study results also show a 42% reduction in smaller field size compared to the larger field size. When taking chest and abdomen radiographic images of neonatal patients in incubators, appropriate field sizes are required to prevent inappropriate dose absorption for non-thoracic organs.

Deep Learning for Pneumonia Classification in Chest Radiography Images using Wavelet Transform

Chronic respiratory diseases constitute a prognostic severity factor for some respiratory illnesses. A case in point is pneumonia, a lung infection, whose effective management requires highly accurate diagnosis and precise treatment. Categorizing pneumonia as positive or negative does go through a process of classifying chest radiography images. This task plays a crucial role in medical diagnostics as it facilitates the detection of pneumonia and helps in making timely treatment decisions. Deep learning has shown remarkable effectiveness in various medical imaging applications, including the recognition and categorization of pneumonia in chest radiography images. The main aim of this research is to compare the efficacy of two convolutional neural network models for classifying pneumonia in chest radiography images. The first model was directly trained on the original images, achieving a training accuracy of 0.9266, whereas the second model was trained on images transformed using wavelets and achieved a training accuracy of 0.94. The second model demonstrated significantly superior results in terms of accuracy, sensitivity, and specificity.

Comparative Review of Computed Tomography Scans and Radiography in Evaluating Blunt Chest Trauma

Previous comparative studies revealed that a substantial proportion of blunt chest trauma often goes unnoticed when evaluated \with chest X-rays. The computed tomography scan, therefore, could play an important role in the initial assessment of patients following blunt chest trauma. This review aims to highlight the findings of chest computed tomography scans and compare them to chest radiography in patients with thoracic trauma resulting from blunt trauma. A systematic search across diverse online databases was performed in PubMed, MEDLINE, and Embase databases. The study selection process was guided by inclusion criteria that focused on individuals afflicted with chest injuries following accidents, engaged in a direct comparison of the diagnostic efficacy of chest CT scans and conventional chest X-rays, and was available in English. In the medical literature, chest radiography and CT imaging provide essential insights to guide clinical decision-making. The precise ability to distinguish between typical and atypical indications of damage across both superficial and deeper anatomical structures holds considerable importance during the initial assessment of such cases. A systematic approach, exemplified through the proposed "outside-in" sequence, offers a proficient strategy to evaluate all potential injuries comprehensively. This approach streamlines the evaluation process, ultimately empowering the medical team to promptly and effectively identify and manage these injuries.

Complex network-based classification of radiographic images for COVID-19 diagnosis.

In this work, we present a network-based technique for chest X-ray image classification to help the diagnosis and prognosis of patients with COVID-19. From visual inspection, we perceive that healthy and COVID-19 chest radiographic images present different levels of geometric complexity. Therefore, we apply fractal dimension and quadtree as feature extractors to characterize such differences. Moreover, real-world datasets often present complex patterns, which are hardly handled by only the physical features of the data (such as similarity, distance, or distribution). This issue is addressed by complex networks, which are suitable tools for characterizing data patterns and capturing spatial, topological, and functional relationships in data. Specifically, we propose a new approach combining complexity measures and complex networks to provide a modified high-level classification technique to be applied to COVID-19 chest radiographic image classification. The computational results on the Kaggle COVID-19 Radiography Database show that the proposed method can obtain high classification accuracy on X-ray images, being competitive with state-of-the-art classification techniques. Lastly, a set of network measures is evaluated according to their potential in distinguishing the network classes, which resulted in the choice of communicability measure. We expect that the present work will make significant contributions to machine learning at the semantic level and to combat COVID-19.

Comparative Analysis of MFO, GWO and GSO for Classification of Covid-19 Chest X-Ray Images

Medical images play a crucial role in the classification of various diseases and conditions. One of the imaging modalities is X-rays which provide valuable visual information that helps in the identification and characterization of various medical conditions. Chest radiograph (CXR) images have long been used to examine and monitor numerous lung disorders, such as tuberculosis, pneumonia, atelectasis, and hernia. COVID-19 detection can be accomplished using CXR images as well. COVID-19, a virus that causes infections in the lungs and the airways of the upper respiratory tract, was first discovered in 2019 in Wuhan Province, China, and has since been thought to cause substantial airway damage, badly impacting the lungs of affected persons. The virus was swiftly gone viral around the world and a lot of fatalities and cases growing were recorded on a daily basis. CXR can be used to monitor the effects of COVID-19 on lung tissue. This study examines a comparison analysis of k-nearest neighbors (KNN), Extreme Gradient Boosting (XGboost), and Support-Vector Machine (SVM) are some classification approaches for feature selection in this domain using The Moth-Flame Optimization algorithm (MFO), The Grey Wolf Optimizer algorithm (GWO), and The Glowworm Swarm Optimization algorithm (GSO). For this study, researchers employed a data set consisting of two sets as follows: 9,544 2D X-ray images, which were classified into two sets utilizing validated tests: 5,500 images of healthy lungs and 4,044 images of lungs with COVID-19. The second set includes 800 images, 400 of healthy lungs and 400 of lungs affected with COVID-19. Each image has been resized to 200x200 pixels. Precision, recall, and the F1-score were among the quantitative evaluation criteria used in this study.

The Most Valuable Predictive Factors for Bronchopulmonary Dysplasia in Very Preterm Infants.

It is urgent to make a rapid screening of infants at the highest risk for bronchopulmonary dysplasia (BPD) via some succinct postnatal biomarkers, such as Ureaplasma Urealyticum (UU) infection and chest radiograph images. A retrospective study was performed. Moderate to severe BPD or death was set as the main outcome. The association between putative variables and the main outcome were assessed by bivariate analyses and logistic regression. A total of 134 infants were enrolled. Bivariate analyses showed the gestational age, birth weight, appearances of diffuse opacities or grid shadows/interstitial opacities or mass opacities or cystic lucencies on chest radiographic images, a ductal diameter ≥1.5 mm and whether UU infection was associated with BPD. After adjustment by logistic regression, the risk of BPD with gestational age, sex and specific chest-radiographic manifestations remained significant. Chest radiograph images (appearance of diffuse opacities or grid shadows/interstitial opacities or mass opacities or cystic lucencies) could provide a quick prediction of developing BPD in clinical practice, in addition to gestational age and sex. UU infection was not an independent risk factor for BPD.

Online diagnosis of COVID-19 from chest radiography images by using deep learning algorithms

Online diagnosis of COVID-19 from chest radiography images by using deep learning algorithms

Neural Networks for the Detection of COVID-19 and Other Diseases: Prospects and Challenges.

Artificial neural networks (ANNs) ability to learn, correct errors, and transform a large amount of raw data into beneficial medical decisions for treatment and care has increased in popularity for enhanced patient safety and quality of care. Therefore, this paper reviews the critical role of ANNs in providing valuable insights for patients' healthcare decisions and efficient disease diagnosis. We study different types of ANNs in the existing literature that advance ANNs' adaptation for complex applications. Specifically, we investigate ANNs' advances for predicting viral, cancer, skin, and COVID-19 diseases. Furthermore, we propose a deep convolutional neural network (CNN) model called ConXNet, based on chest radiography images, to improve the detection accuracy of COVID-19 disease. ConXNet is trained and tested using a chest radiography image dataset obtained from Kaggle, achieving more than 97% accuracy and 98% precision, which is better than other existing state-of-the-art models, such as DeTraC, U-Net, COVID MTNet, and COVID-Net, having 93.1%, 94.10%, 84.76%, and 90% accuracy and 94%, 95%, 85%, and 92% precision, respectively. The results show that the ConXNet model performed significantly well for a relatively large dataset compared with the aforementioned models. Moreover, the ConXNet model reduces the time complexity by using dropout layers and batch normalization techniques. Finally, we highlight future research directions and challenges, such as the complexity of the algorithms, insufficient available data, privacy and security, and integration of biosensing with ANNs. These research directions require considerable attention for improving the scope of ANNs for medical diagnostic and treatment applications.

Tuberculosis contact-tracing results in childhood: a retrospective study in a tertiary-care children’s hospital in Turkey

ABSTRACT Background Smear-positive adults with tuberculosis are the main source of childhood tuberculosis. The evaluation of children exposed to tuberculosis and determination of the disease stages are the cornerstones of managing childhood tuberculosis. Aim To determine the frequency of tuberculous contact, latent tuberculosis infection and tuberculosis disease in children who were in contact with smear-positive adults. Methods This is a single-centre, retrospective study. The medical records of children exposed to tuberculosis (<18 years old) between 2014 and 2018 were investigated. After diagnosing the index cases, the children were referred to the hospital. To identify the children in contact with adults with tuberculosis, a careful medical history, demographic features and physical examination, tuberculin skin test, postero-anterior and lateral chest radiographs, and, if necessary, chest computed tomography and microbiological tests were undertaken. The children’s final diagnosis, treatment regimens and follow-up were documented. The sensitivity, specificity and positive and negative predictive values, tuberculin skin test and chest radiograph imaging were assessed and compared with computed tomography results. Results A total of 150 paediatric patients were exposed to 88 index cases. These were fathers in 29.3% of cases and mothers in 10% of cases. Of the children, 131 (87.3%) were asymptomatic, and physical examination was normal in all children, apart from one who had respiratory symptoms. The tuberculin skin test results were positive in 60 (43%) patients and chest radiograph was abnormal in 100 (66%) children. Findings were consistent with tuberculosis in 34 (40%) of the 84 patients who underwent computed tomography. Fifty (38.5%) of the remaining children were defined as having been in contact with a case of tuberculosis, 41 (31.5%) had latent tuberculous infection and 39 (30%) had tuberculosis disease. Conclusion Pulmonary tuberculosis is asymptomatic in most children but with meticulous use of computed tomography it can be detected in asymptomatic children who have had close contact with tuberculosis. Abbreviation: AFB: acid-fast bacilli; AUC: area under the curve; BCG: bacillus Calmette-Guérin; CI: confidence interval; CT: computed tomography; CXR: chest radiograph; HIV: human immunodeficiency virus; ICD-10: International Classification of Diseases 10; LTBI: latent tuberculosis infection; MDR-TB: multi-drug-resistant tuberculosis; NPV: negative predictive value; PCR: polymerase chain reaction; PPV: positive predictive value; ROC: receiver operating characteristics; SD: standard deviation; TB: tuberculosis; TST: tuberculin skin test; XDR-TB: extensively drug-resistant tuberculosis.

Understanding transfer learning for chest radiograph clinical report generation with modified transformer architectures

The image captioning task is increasingly prevalent in artificial intelligence applications for medicine. One important application is clinical report generation from chest radiographs. The clinical writing of unstructured reports is time consuming and error-prone. An automated system would improve standardization, error reduction, time consumption, and medical accessibility. In this paper we demonstrate the importance of domain specific pre-training and propose a modified transformer architecture for the medical image captioning task. To accomplish this, we train a series of modified transformers to generate clinical reports from chest radiograph image input. These modified transformers include: a meshed-memory augmented transformer architecture with visual extractor using ImageNet pre-trained weights, a meshed-memory augmented transformer architecture with visual extractor using CheXpert pre-trained weights, and a meshed-memory augmented transformer whose encoder is passed the concatenated embeddings using both ImageNet pre-trained weights and CheXpert pre-trained weights. We use BLEU(1-4), ROUGE-L, CIDEr, and the clinical CheXbert F1 scores to validate our models and demonstrate competitive scores with state of the art models. We provide evidence that ImageNet pre-training is ill-suited for the medical image captioning task, especially for less frequent conditions (e.g.: enlarged cardiomediastinum, lung lesion, pneumothorax). Furthermore, we demonstrate that the double feature model improves performance for specific medical conditions (edema, consolidation, pneumothorax, support devices) and overall CheXbert F1 score, and should be further developed in future work. Such a double feature model, including both ImageNet pre-training as well as domain specific pre-training, could be used in a wide range of image captioning models in medicine.

Atypical Radiology Imaging of Squamous Cell Lung Carcinoma: A Case Report and Narrative Review

Introduction: Multiple well-circumscribed round nodules, also known as “Cannonball” appearance in the lungs, are typically associated with specific pulmonary metastases. However, limited reports denote the finding, especially if correlated with primary squamous cell lung carcinoma in the younger population. This is the first case report in Indonesia of a 19-year-old male with primary squamous cell carcinoma in the lung and chest radiograph image mimicking a “Cannonball” appearance Case Presentation: A 19-year-old male was presented with a history of progressing hemoptysis for two weeks with no other abnormal respiratory symptoms but a reduction of vocal fremitus. The subject is an occasional smoker and had a history of lung tuberculosis in 2009 and a suprarenal or adrenal mass. X-ray imaging revealed multiple bilateral pulmonary nodules, mimicking the “Cannonball” appearance. A fine needle aspiration (FNA) biopsy of a nodule was done and interpreted to be primary squamous cell carcinoma. Symptomatic treatment was given and the subject was discharged on the fifth day. Conclusions: A “Cannonball” appearance or its mimic is not always typical to certain carcinoma based on what this case report finds. In the case of a tumor, radiograph findings should also be correlated with other supporting tests such as biopsy and clinical presentation.

An Efficient CNN Model for Automatic Diagnosis of Cardiomegaly from Chest Radiographic Images

An Efficient CNN Model for Automatic Diagnosis of Cardiomegaly from Chest Radiographic Images

A Deep Learning Model for the Diagnosis and Discrimination of Gram-Positive and Gram-Negative Bacterial Pneumonia for Children Using Chest Radiography Images and Clinical Information.

This study aimed to develop a deep learning model based on chest radiography (CXR) images and clinical data to accurately classify gram-positive and gram-negative bacterial pneumonia in children to guide the use of antibiotics. We retrospectively collected CXR images along with clinical information for gram-positive (n=447) and gram-negative (n=395) bacterial pneumonia in children from January 1, 2016, to June 30, 2021. Four types of machine learning models based on clinical data and six types of deep learning algorithm models based on image data were constructed, and multi-modal decision fusion was performed. In the machine learning models, CatBoost, which only used clinical data, had the best performance; its area under the receiver operating characteristic curve (AUC) was significantly higher than that of the other models (P<0.05). The incorporation of clinical information improved the performance of deep learning models that relied solely on image-based classification. Consequently, AUC and F1 increased by 5.6% and 10.2% on average, respectively. The best quality was achieved with ResNet101 (model accuracy: 0.75, recall rate: 0.84, AUC: 0.803, F1: 0.782). Our study established a pediatric bacterial pneumonia model that utilizes CXR and clinical data to accurately classify cases of gram-negative and gram-positive bacterial pneumonia. The results confirmed that the addition of image data to the convolutional neural network model significantly improved its performance. While the CatBoost-based classifier had greater advantages owing to a smaller dataset, the quality of the Resnet101 model trained using multi-modal data was comparable to that of the CatBoost model, even with a limited number of samples.