Chest X-ray Diagnosis Research Articles

Pre-trained Convolutional Neural Network Model Based Pneumonia Classification from Chest X-Ray Images

Pneumonia is a life-threatening infectious lung disease in humans. It causes the air sacs of the lungs to be filled up with fluid or pus. This ranges from mild to severe, based on the type of germ causing the infection. There are various causes for pneumonia. One of the major causes include bacteria like Streptococcus and viruses like influenza, corona virus and so on. The treatment for bacterial pneumonia varies from viral pneumonia. So, identifying the type of pneumonia is essential to provide a better treatment. According to the World Health Organization, pneumonia is responsible for one out of every three deaths in India. The early detection of pneumonia is essential to provide the treatment in the right time. The diagnosis of chest X-rays needs expert radiotherapists for evaluation. Even after the detection, it is a challenge for the doctors to classify pneumonia accurately. The system aims in classifying pneumonia from chest X-ray images thereby helps to save time and to increase accuracy in prediction. Deep Learning algorithms have been effective in analyzing medical images and has gained much attention for disease classification. This paper presents Convolutional Neural Network (CNN) architecture to extract features from the image and classify them using the extracted features. A pre-trained Xception model along with data augmentation techniques is employed to classify the chest x-ray images as bacterial or viral pneumonia with better accuracy. This helps the doctors to provide timely assistance to the patients.

Multi-Institution Encrypted Medical Imaging AI Validation Without Data Sharing

Adoption of artificial intelligence medical imaging applications is often impeded by barriers between healthcare systems and algorithm developers given that access to both private patient data and commercial model IP is important to perform pre-deployment evaluation. This work investigates a framework for secure, privacy-preserving and AI-enabled medical imaging inference using CrypTFlow2, a state-of-the-art end-to-end compiler allowing cryptographically secure 2-party Computation (2PC) protocols between the machine learning model vendor and target patient data owner. A common DenseNet-121 chest x-ray diagnosis model was evaluated on multi-institutional chest radiographic imaging datasets both with and without CrypTFlow2 on two test sets spanning seven sites across the US and India, and comprising 1,149 chest x-ray images. We measure comparative AUROC performance between secure and insecure inference in multiple pathology classification tasks, and explore model output distributional shifts and resource constraints introduced by secure model inference. Secure inference with CrypTFlow2 demonstrated no significant difference in AUROC for all diagnoses, and model outputs from secure and insecure inference methods were distributionally equivalent. The use of CrypTFlow2 may allow off-the-shelf secure 2PC between healthcare systems and AI model vendors for medical imaging, without changes in performance, and can facilitate scalable pre-deployment infrastructure for real-world secure model evaluation without exposure to patient data or model IP.

Multi-Institution Encrypted Medical Imaging Ai Validation Without Data Sharing

Adoption of artificial intelligence medical imaging applications is often impeded by barriers between healthcare systems and algorithm developers given that access to both private patient data and commercial model IP is important to perform pre-deployment evaluation. This work investigates a framework for secure, privacy-preserving and AI-enabled medical imaging inference using CrypTFlow2, a state-of-the-art end-to-end compiler allowing cryptographically secure 2-party Computation (2PC) protocols between the machine learning model vendor and target patient data owner. A common DenseNet-121 chest x-ray diagnosis model was evaluated on multi-institutional chest radiographic imaging datasets both with and without CrypTFlow2 on two test sets spanning seven sites across the US and India, and comprising 1,149 chest x-ray images. We measure comparative AUROC performance between secure and insecure inference in multiple pathology classification tasks, and explore model output distributional shifts and resource constraints introduced by secure model inference. Secure inference with CrypTFlow2 demonstrated no significant difference in AUROC for all diagnoses, and model outputs from secure and insecure inference methods were distributionally equivalent. The use of CrypTFlow2 may allow off-the-shelf secure 2PC between healthcare systems and AI model vendors for medical imaging, without changes in performance, and can facilitate scalable pre-deployment infrastructure for real-world secure model evaluation without exposure to patient data or model IP.

1521. Risk factors associated with complications of influenza A or B infection, southern Puerto Rico, 2016-2019

BackgroundInfluenza is one of the most common viral respiratory diseases in the United States with 9–45 million cases per year. In Puerto Rico, more than 18,000 cases of influenza were reported during the 2019/20 season. The Sentinel Enhanced Dengue Surveillance System (SEDSS) conducts acute febrile illness surveillance for inpatient and outpatients at a southern Puerto Rico tertiary care hospital.MethodsWe performed a retrospective analysis of disease severity among laboratory-confirmed influenza cases enrolled in SEDSS. Cases with complications such as pneumonia or bronchitis evidenced by chest x-ray or clinical diagnosis were classified as severe.ResultsBetween January 2016 and December 2019, there were 2,835 laboratory-confirmed influenza cases. More than half (1,662, 59%) were aged 0–20 years, 51% (n=1,447)were female, and 4% (n=124)required hospital admission. Among all influenza cases, influenza A virus was most common (n = 1,963, 69%) followed by influenza B virus (n = 872, 31%). Odds of having influenza A were higher among adults (OR=1.62; 95% CI:1.38-1.92). Chronic disease history among influenza cases included: asthma (n= 507, 18%), hypertension (n = 337, 12%), hypercholesterolemia (n = 157, 6%), coronary heart disease (n = 79, 3%), immunodeficiency (n = 11, 1%), and kidney disease (n = 10, 1%). Overall, 61(2%) influenza cases developed complications, 21 developed pneumonia of which 85% had chest x-ray performed. One-third of influenza cases with pneumonia or bronchitis had preexisting conditions like asthma (33%) and hypertension (34%). Seventy five percent of Influenza A cases with asthma developed bronchitis or pneumonia. The odds of having severe Influenza were higher among patients with underlying hypertension compared to those without hypertension (OR= 3.8; 95% CI:2.0–7.1). Moreover, influenza cases with asthma had a higher odd of developing pneumonia or bronchitis as compared to those without asthma (OR=2.3; 95% CI: 1.3–4.0). While asthma is a recognized risk factor for influenza complications, isolated hypertension is not.ConclusionIdentifying people at risk for influenza complications provides a guide for clinicians and public health officials to implement preventive measures and improve clinical outcomes.Disclosures All Authors: No reported disclosures

Learning Hierarchical Attention for Weakly-Supervised Chest X-Ray Abnormality Localization and Diagnosis.

We consider the problem of abnormality localization for clinical applications. While deep learning has driven much recent progress in medical imaging, many clinical challenges are not fully addressed, limiting its broader usage. While recent methods report high diagnostic accuracies, physicians have concerns trusting these algorithm results for diagnostic decision-making purposes because of a general lack of algorithm decision reasoning and interpretability. One potential way to address this problem is to further train these models to localize abnormalities in addition to just classifying them. However, doing this accurately will require a large amount of disease localization annotations by clinical experts, a task that is prohibitively expensive to accomplish for most applications. In this work, we take a step towards addressing these issues by means of a new attention-driven weakly supervised algorithm comprising a hierarchical attention mining framework that unifies activation- and gradient-based visual attention in a holistic manner. Our key algorithmic innovations include the design of explicit ordinal attention constraints, enabling principled model training in a weakly-supervised fashion, while also facilitating the generation of visual-attention-driven model explanations by means of localization cues. On two large-scale chest X-ray datasets (NIH ChestX-ray14 and CheXpert), we demonstrate significant localization performance improvements over the current state of the art while also achieving competitive classification performance.

Chest X-ray Image Analysis and Penehyclidine Hydrochloride Postconditioning Influence of Lung Ischemia Reperfusion Injury

Purpose: We analyze chest X-ray diagnosis results of patients undergoing cardiopulmonary bypass heart surgery in intensive care unit (ICU). In our previous study, we found that penehyclidine hydrochloride (PHC) preconditioning pretects lung function and reduced the apoptosis. Although preconditioning is effective, this clinical treatment is often given only when symptoms appear, when postconditioning is easier to administer. In this study, after confirming the lung condition via medical image analysis, we aimed to look the effect of high-dose penehyclidine hydrochloride postconditioning in lung after I/R in rats, and the apoptosis mechanisms involved. Methods: Chest X-ray was taken in 256 adult patients under cardiopulmonary bypass with heart syurery. Medical image analysis was preliminarily performed, the diagnostic results were analyzed. Rats were subjected to ischemia in left lung 45 min, and then 2 h reperfusion, and treated with PHC. We then observed the effects of PHC on the following: lung function, lung injury, oxidative stress, rate of apoptosis of lung cells, apoptosis-related proteins, and the p38 MAPK pathway. Results: Analysis results of chest X-ray suggest multiple pulmonary complications after cardiopulmonary bypass heart surgery. The laboratory results showed that high-dose penehyclidine hydrochloride postconditioning significantly protected lung function, reduced oxidative stress and apoptosis of the lungs caused by reperfusion, and inhibited activation of the p38 MAPK pathway. Conclusion: Analysis results of chest X-ray show multiple pulmonary complications after cardiopulmonary bypass heart surgery. This study suggests that high-dose penehyclidine hydrochloride can treat lung I/R injury.

Discriminative ensemble learning for few-shot chest x-ray diagnosis.

Few-shot learning is an almost unexplored area in the field of medical image analysis. We propose a method for few-shot diagnosis of diseases and conditions from chest x-rays using discriminative ensemble learning. Our design involves a CNN-based coarse-learner in the first step to learn the general characteristics of chest x-rays. In the second step, we introduce a saliency-based classifier to extract disease-specific salient features from the output of the coarse-learner and classify based on the salient features. We propose a novel discriminative autoencoder ensemble to design the saliency-based classifier. The classification of the diseases is performed based on the salient features. Our algorithm proceeds through meta-training and meta-testing. During the training phase of meta-training, we train the coarse-learner. However, during the training phase of meta-testing, we train only the saliency-based classifier. Thus, our method is first-of-its-kind where the training phase of meta-training and the training phase of meta-testing are architecturally disjoint, making the method modular and easily adaptable to new tasks requiring the training of only the saliency-based classifier. Experiments show as high as ∼19% improvement in terms of F1 score compared to the baseline in the diagnosis of chest x-rays from publicly available datasets.

Using Artificial Intelligence for COVID-19 Chest X-ray Diagnosis.

Coronavirus disease-19 (COVID-19), caused by a novel member of the coronavirus family, is a respiratory disease that rapidly reached pandemic proportions with high morbidity and mortality. In only a few months, it has had a dramatic impact on society and world economies. COVID-19 has presented numerous challenges to all aspects of health care, including reliable methods for diagnosis, treatment, and prevention. Initial efforts to contain the spread of the virus were hampered by the time required to develop reliable diagnostic methods. Artificial intelligence (AI) is a rapidly growing field of computer science with many applications for health care. Machine learning is a subset of AI that uses deep learning with neural network algorithms. It can recognize patterns and achieve complex computational tasks often far quicker and with increased precision than can humans. In this article, we explore the potential for the simple and widely available chest X-ray (CXR) to be used with AI to diagnose COVID-19 reliably. Microsoft CustomVision is an automated image classification and object detection system that is a part of Microsoft Azure Cognitive Services. We utilized publicly available CXR images for patients with COVID-19 pneumonia, pneumonia from other etiologies, and normal CXRs as a dataset to train Microsoft CustomVision. Our trained model overall demonstrated 92.9% sensitivity (recall) and positive predictive value (precision), with results for each label showing sensitivity and positive predictive value at 94.8% and 98.9% for COVID-19 pneumonia, 89% and 91.8% for non-COVID-19 pneumonia, 95% and 88.8% for normal lung. We then validated the program using CXRs of patients from our institution with confirmed COVID-19 diagnoses along with non-COVID-19 pneumonia and normal CXRs. Our model performed with 100% sensitivity, 95% specificity, 97% accuracy, 91% positive predictive value, and 100% negative predictive value. We have used a readily available, commercial platform to demonstrate the potential of AI to assist in the successful diagnosis of COVID-19 pneumonia on CXR images. The findings have implications for screening and triage, initial diagnosis, monitoring disease progression, and identifying patients at increased risk of morbidity and mortality. Based on the data, a website was created to demonstrate how such technologies could be shared and distributed to others to combat entities such as COVID-19 moving forward.

Multi-Channel Transfer Learning of Chest X-ray Images for Screening of COVID-19

The 2019 novel coronavirus (COVID-19) has spread rapidly all over the world. The standard test for screening COVID-19 patients is the polymerase chain reaction test. As this method is time consuming, as an alternative, chest X-rays may be considered for quick screening. However, specialization is required to read COVID-19 chest X-ray images as they vary in features. To address this, we present a multi-channel pre-trained ResNet architecture to facilitate the diagnosis of COVID-19 chest X-ray. Three ResNet-based models were retrained to classify X-rays in a one-against-all basis from (a) normal or diseased, (b) pneumonia or non-pneumonia, and (c) COVID-19 or non-COVID19 individuals. Finally, these three models were ensembled and fine-tuned using X-rays from 1579 normal, 4245 pneumonia, and 184 COVID-19 individuals to classify normal, pneumonia, and COVID-19 cases in a one-against-one framework. Our results show that the ensemble model is more accurate than the single model as it extracts more relevant semantic features for each class. The method provides a precision of 94% and a recall of 100%. It could potentially help clinicians in screening patients for COVID-19, thus facilitating immediate triaging and treatment for better outcomes.

ECMO İle Tedavi Edilen ARDS Olgusu

Acute Respiratory Distress Syndrome (ARDS) is characterized with diffuse bilaterat radiologic infiltration, and a hypoxemic respiratory failure. A 48-year-old housewife hospitalized in a chest diseases hospital with the diagnosis of pneumonia was intubated for increased respiratory distress. The patient, who developed cardiac arrest three times underwent CRP for nearly 15 minutes. When she was admitted to intensive care unit of a tertiary healthcare institute, severe acidosis was detected in her first arterial blood gas sample (ph 6.9). Bilateral rales were heard in respiratory system auscultation and “cotton wool spots” were observed on her PA chest X-ray and initial diagnosis of ARDS was made. The patient was consulted with the department of infectious diseases, and zyvoxid 2x1 (avelox 1x1 meronem 2x1) were added to her antibiotherapy. The patient’s admission parameters were as follows: WBC 45000/mm3 CRP: 11 mg/dL sedimentation: 84 mg/dL. Although the patient was followed up on a mechanical ventilator with a protective ventilation strategy for 2 days, ECMO treatment was decided because PaO2 / FiO2 <100 and blood gas values did not improve. ECMO treatment was applied for 5 days. Then, the patient was extubated and transferred to the service without complications.

Learning to Recognize Chest-Xray Images Faster and More Efficiently Based on Multi-Kernel Depthwise Convolution

The development of convolutional neural networks has promoted the progress of computer-aided diagnostic systems. Details in medical image, such as the texture and tissue structure, are crucial features for diagnosis. Therefore, large input images combined with deep convolution neural networks are adopted to boost the performance in recent research of chest X-ray diagnosis. Meanwhile, due to the variable sizes of thoracic diseases, many researchers have worked to introduce additional module to capture multi-scale feature of images in CNN. However, these efforts hardly consider the computational costs of large inputs and introduced additional modules. This paper aims to automatically diagnose diseases on chest X-rays images quickly and effectively. We propose the multi-kernel depthwise convolution(MD-Conv) which contains depthwise convolution kernels with different filter sizes in one depthwise convolution layer. MD-Conv has high calculation efficiency and few parameters. Because its ability to learn multi-scale feature based on the multi-size kernels, it is appropriate for medical images diagnosis tasks in which abnormalities varied in sizes. In addition, larger depthwise convolution kernels are adopted in MD-Conv to obtain a larger receptive field efficiently, which can ensure sufficient receptive field for high resolution inputs. MD-Conv can be easily applied in modern lightweight networks to replace the normal depthwise convolution layer. We conduct experiments on the Chest X-ray 14 Dataset, which is the largest available chest x-ray dataset, and obtain competitive results. We also evaluate the MD-Conv on the new released dataset for pediatric pneumonia diagnosis. We obtain a better performance of 98.3% AUC than original paper (96.8%) for recognize pneumonia versus normal. Meanwhile we compare the FLOPs and Params of different models to show their efficiency for chest X-rays recognition.

Assessment of serum sodium levels in patients of lower respiratory tract infections (LRTI) in children at tertiary care center

Introduction: Lower respiratory tract infection is an infection beneath the larynx which includes: Pneumonia, Wheeze associated Lower respiratory tract infection, Bronchiolitis and Empyema. The aim of this study was to find out the association among hyponatremia and LRTI in tertiary care center. Materials and Methods: The sample size was calculated to be a minimum of 50 subjects. Based on clinical signs and symptoms (as defined by WHO) and infiltrates present on chest X-ray diagnosis of LRTI such as Pneumonia, Bronchiolitis was made. Then patients were subjected to routine investigations such as: Complete blood count, Serum electrolyte (serum sodium) at the time of admission, on day 2 and day 3 by Ion selective electrodes. Results: Mild hyponatremia was found among 11 patients (16.92%), moderate among 9 (13.85%) and severe among 2 (3.08%) patients. Hyponatremia was found to be more common among 1 to 5 years age group as compared to ≥ 2months to 12 months and > 5 to 12 years age groups. Conclusion: Hyponatremia is a significantly common association among hospitalized children with lower respiratory tract infections and it is mainly due to syndrome of inappropriate antidiuretic hormone secretion (SIADH).

Effectiveness of Lung Ultrasound in Comparison with Chest X-Ray in Diagnosis of Lung Consolidation.

BACKGROUND:Lung ultrasound (US) is an available and inexpensive tool for the diagnosis of community-acquired pneumonia (CAP); it which has no hazards of radiation and can be easily used.AIM:To evaluate the efficacy of lung ultrasound in the diagnosis and follow-up of CAP.PATIENTS AND METHODS:100 patients aged from 40 to 63 years with a mean age of 52.3 ± 10 years admitted to the Critical Care Department, Cairo University with pictures of CAP. Lung US was performed for all patients initially, then a plain chest X-ray (CXR) was performed. Another lung ultrasound was performed on the 10th day after admission.RESULTS:Initial chest X-ray was correlated with the initial chest ultrasound examination in CAP diagnosis (R-value = 0.629, P < 0.001). Cohen’s κ was run to determine if there is an agreement between the findings of the initial chest X-ray findings and those of the initial chest ultrasound in CAP diagnosis. A moderate agreement was found where κ = .567 (95% CI, 0.422 to 0.712) and P < 0.001. Upon initial examination, the CXR diagnosed CAP in 48.0% of patients, while lung US diagnosed the disease in 70% of patients. Moreover, lung US was more sensitive than CXR (P-value < 0.001). Compared to the accuracy of computed tomography (CT) chest (100%) which is the gold standard for CAP diagnosis, the accuracy of lung US was 95.0%, while the accuracy of CXR was 81.0%.CONCLUSION:This study proved the effectiveness of lung ultrasound in CAP diagnosis.

Lung ultrasound in the diagnosis of neonatal respiratory failure prior to patient transport.

Lung ultrasound (LUS) at the point-of-care is a new method that is increasingly used in neonatology. The aim of this study was to determine the utility of the addition of LUS prior to the interhospital transport of neonates with respiratory failure. LUS was performed on 50 newborns with respiratory failure prior to transport to a tertiary neonatal intensive care unit. We analyzed the performance of LUS for diagnosing the cause of respiratory failure, the concordance between LUS, chest X-ray (CXR) and final clinical diagnosis, and the impact of LUS on clinical decision making before transport. LUS sensitivity for the diagnosis of respiratory distress syndrome was 91.3% (95%CI: 70.5-98.5%), and specificity was 92.6% (95%CI: 74.2-98.7%), whereas sensitivity and specificity of CXR were 69.6% (95%CI: 47.0-85.9%) and 81.5% (95%CI: 61.2-92.9%), respectively. For the recognition of pneumothorax (PTX) LUS had a sensitivity of 83.3% (95%CI: 36.5-99.1%) and a specificity of 100% (95%CI: 89.9-100%). For CXR, sensitivity was 16.7% (95%CI: 0.01-63.5%) and specificity was 97.7% (95%CI: 86.4-99.9%). The agreement between LUS and CXR in diagnosing the cause of respiratory failure was substantial (κ of 0.57 [95%CI: 0.40-0.74]) and the agreement between LUS and the final clinical diagnosis was very good (κ of 0.86 [95%CI: 0.74-0.98]). In 42% of the patients, a LUS examination prior to transport indicated the need for endotracheal tube repositioning or PTX decompression. LUS may be a reliable imaging technique for differentiating the causes of respiratory failure before neonatal transport. Use of LUS may optimize the care of infants during transport.

Echocardiography in patients with chest x-ray diagnosis of cardiomegaly at an academic hospital in Zambia

&#x0D; Background: Cardiomegaly is the most diagnosed condition on Chest X-ray (CXR) at the study site. However, little is found on the association between cardiomegaly and echocardiography (ECHO) findings at this site. The main objective of this study was to establish whether there is an association between cardiomegaly diagnosed on a CXR and ECHO findings of the same patient.&#x0D; Methods and Materials: This retrospective study involved the retrieving of data from files of 124 patients who had cardiomegaly diagnosed on CXR and had undergone an ECHO examination. The study sought to establish whether there is an association between cardiomegaly diagnosed on a CXR and ECHO findings of the same patients.&#x0D; Results: There was a positive association between cardiomegaly diagnosed on CXR and ECHO findings and the association was significant at 5%. In addition, there was also a strong association between age and cardiomegaly and the association was also significant at 5%. The Chi-square results at 5% level of significance showed that there is not enough evidence to suggest an association between gender and cardiomegaly. Therefore, both women and men have an equal chance of acquiring cardiomegaly. Left Ventricular Diastolic Dysfunction was the most prevalent condition, having been seen in more than 50% of patients. Most of the conditions identified in patients were very severe. In addition, 70% of these patients had a raised blood pressure with 43% having severely raised blood pressure.The test results also showed that 66% of patients in the sample had increased blood flow on Doppler measured at certain points of the heart with 6% of the patients having severely increased blood flow. Patients with increased blood flow showed an increase in blood pressure with those with severe increased blood flow showing a strong positive association with severe high blood pressure.&#x0D; Conclusion: These findings show that there is a strong association between cardiomegaly diagnosed on CXR to ECHO findings of the same patient. Therefore, echocardiography should be done on all patients diagnosed with cardiomegaly on CXR. In addition, priority should be given to elderly patients (65 years and above) as most of these had both severe cardiomegaly and one or more cardiovascular diseases.&#x0D;

Echocardiography in Patients with Chest X-ray Diagnosis of Cardiomegaly at an Academic Hospital in Zambia

Background: Cardiomegaly is the most diagnosed condition on Chest X-ray (CXR) at the study site. However, little is found on the association between cardiomegaly and echocardiography (ECHO) findings at this site. The main objective of this study was to establish whether there is an association between cardiomegaly diagnosed on a CXR and ECHO findings of the same patient.&#x0D; Methods and Materials: This retrospective study involved the retrieving of data from files of 124 patients who had cardiomegaly diagnosed on CXR and had undergone an ECHO examination. The study sought to establish whether there is an association between cardiomegaly diagnosed on a CXR and ECHO findings of the same patients.&#x0D; Results: There was a positive association between cardiomegaly diagnosed on CXR and ECHO findings and the association was significant at 5%. In addition, there was also a strong association between age and cardiomegaly and the association was also significant at 5%. The Chi-square results at 5% level of significance showed that there is not enough evidence to suggest an association between gender and cardiomegaly. Therefore, both women and men have an equal chance of acquiring cardiomegaly. Left Ventricular Diastolic Dysfunction was the most prevalent condition, having been seen in more than 50% of patients. Most of the conditions identified in patients were very severe. In addition, 70% of these patients had a raised blood pressure with 43% having severely raised blood pressure.The test results also showed that 66% of patients in the sample had increased blood flow on Doppler measured at certain points of the heart with 6% of the patients having severely increased blood flow. Patients with increased blood flow showed an increase in blood pressure with those with severe increased blood flow showing a strong positive association with severe high blood pressure.&#x0D; Conclusion: These findings show that there is a strong association between cardiomegaly diagnosed on CXR to ECHO findings of the same patient. Therefore, echocardiography should be done on all patients diagnosed with cardiomegaly on CXR. In addition, priority should be given to elderly patients (65 years and above) as most of these had both severe cardiomegaly and one or more cardiovascular diseases.

Value of the Chest X-Rays in Diagnosis of Children with Heart Disease in the Age Group 1-4 Years

AbstractBackground: Chest X-ray continues to have an important role in diagnosis of heart diseases.Aim of Study: We aim by this the study to evaluate the use of chest X-ray in the Pediatric Cardiology Unit in Assiut University Children Hospital for children (1-4 years old) with suspected heart disease.Patients and Methods: This was a six-month, case series clinical study undertaken in the inpatient Pediatric Cardiology Unit, Assiut University Hospital, Assiut City, Egypt; between 1 of April 2016 to 30 of September 2016.Results: Chest X-ray is an important tool in evaluating heart disease in children. Noninvasive imaging such as echocar-diography and cardiac MRI provide valuable and detailed assessment of the cardiovascular system; however, the cost incurred from these diagnostic procedures is significant making their routine use difficult.Conclusion: Auditing the guidelines of interpretation of the chest radiograph for children aged one year to four years at the Pediatric Cardiology Unit of Assiut University Hospital for Children, it could be concluded: Chest roentgenogram is accurate in detecting cardiac site and situs.

Improving TB case notification in northern Uganda: evidence of a quality improvement-guided active case finding intervention

BackgroundStrategies to identify and treat undiagnosed prevalent cases that have not sought diagnostic services on their own, are necessary to treat TB in patients earlier and interrupt transmission. Late presentation for medical services of symptomatic patients require special efforts to detect early and notify TB in high risk populations. An intervention that combined quality improvement with facility-led active case finding (QI-ACF) was implemented in 10 districts of Northern Uganda with the highest TB burden to improve case notification among populations at highest risk of TB.MethodsUsing QI-ACF intervention approach in 48 facilities, we; 1) targeted key vulnerable populations, 2) engaged district and facility teams in TB systems strengthening, 3) conducted systematic screening and diagnosis in vulnerable groups (people living with HIV, fishing communities, and prisoners), and 4) trained health workers on national x-ray diagnosis guidelines for smear-negative patients. Facility-led QI-ACF meant that health care providers identified the target population, mobilized and massively screened suspects, and addressed gaps in documentation. Chest X-ray diagnosis was promoted for smear-negative TB among those suspects whose sputum examination was negative. The effect of the intervention on case notification was then assessed separately over the post intervention period.ResultsOver all TB case notification in the intervention districts increased from 171 to 223 per 100,000 population between the baseline months of October–December 2016 and end line month of April–June 2017. TB patient contacts had the majority of TB positive cases identified during active case finding (40, 6.1%). Fishing communities had the highest TB positivity rate at 6.8%. Prisoners accounted for the lowest number of TB positive cases at 34 (2.3%).ConclusionTargeting should be applied at all levels of TB intervention to improve yield: targeting districts and facilities with the lowest rates of case notification and targeting index patient contacts, HIV clients, and fishing communities. Screening tools are useful to guide health workers to identify presumptive cases. Efforts to improve availability of x-ray for TB diagnosis contributed to almost half of the new cases identified. Having all HIV patients who were eligible for viral load provide sputum for TB screening proved easy to implement.

Lung Ultrasound for the Differential Diagnosis of Respiratory Distress in Neonates

Background: Respiratory distress (RD) is the most common neonatal illness. Lung ultrasound (LUS) is a technique previously tested in neonatal studies on RD, but literature regarding its routine clinical applicability is still lacking. Objective: To assess the concordance between LUS performed by neonatologists with different training levels and chest X-ray (CXR) for the diagnosis of RD in newborns during the first 24 h of life. Methods: We enrolled newborns with RD during the first 24 h of life. Patients underwent LUS and CXR. LUS and CXR diagnosis were compared to evaluate concordance. Twenty percent of patients received two LUS (one from an experienced and one from a novice sonographer) to calculate the interobserver agreement. The difference in time needed to reach a diagnosis with LUS and CXR, and from novice and expert operators, was measured. Results: We studied 124 patients; 134 diagnoses were reported. The concordance between LUS and CXR diagnosis was 91% (95% CI 86–96%) with a κ statistic of 0.88 (95% CI 0.81–0.94). The median time to diagnosis was shorter for LUS (9.5 min, IQR 5–15) than for CXR (50 min, IQR 33–64) (p < 0.0001). In 25/124 patients, LUS was performed by both novice and experienced sonographers with complete concordance. The median time to diagnosis was shorter for expert (9 min, IQR 5–15) than novice operators (15 min, IQR 10–20) (p < 0.0002). Conclusion: LUS and CXR have a high concordance in the differential diagnosis of neonatal RD in the first 24 h of life. LUS has a shorter operation time than CXR.

An Ensemble SSL Algorithm for Efficient Chest X-Ray Image Classification

A critical component in the computer-aided medical diagnosis of digital chest X-rays is the automatic detection of lung abnormalities, since the effective identification at an initial stage constitutes a significant and crucial factor in patient’s treatment. The vigorous advances in computer and digital technologies have ultimately led to the development of large repositories of labeled and unlabeled images. Due to the effort and expense involved in labeling data, training datasets are of a limited size, while in contrast, electronic medical record systems contain a significant number of unlabeled images. Semi-supervised learning algorithms have become a hot topic of research as an alternative to traditional classification methods, exploiting the explicit classification information of labeled data with the knowledge hidden in the unlabeled data for building powerful and effective classifiers. In the present work, we evaluate the performance of an ensemble semi-supervised learning algorithm for the classification of chest X-rays of tuberculosis. The efficacy of the presented algorithm is demonstrated by several experiments and confirmed by the statistical nonparametric tests, illustrating that reliable and robust prediction models could be developed utilizing a few labeled and many unlabeled data.