Edge-based real-time diagnosis of pediatric pneumonia using lightweight CNNs and chest X-rays

Background:The aim of this meta-analysis was to evaluate the diagnostic value of lung ultrasound (LUS) in comparison to chest radiography (CXR) in children with pneumonia.Methods:Computer-based retrieval was performed on PubMed and EMBASE. Quality was evaluated according to the quality assessment of diagnostic accuracy studies-2, and Meta-Disc was adopted to perform meta-analysis. Heterogeneity was assessed using Q and I2 statistics. The pooled sensitivity, specificity, and diagnostic odds ratio (DOR) with 95% confidence intervals (CIs) as the primary outcomes were calculated for each index test.Results:Twenty two studies with a total of 2470 patients met the inclusion criteria. Our results showed that the pooled sensitivity, specificity, and DOR for children with pneumonia diagnosed by LUS were 0.95 (95% CI: 0.94 to 0.96), 0.90 (95% CI: 0.87 to 0.92), and 137.49 (95% CI: 60.21 to 313.98), respectively. The pooled sensitivity, specificity, and DOR for pediatric pneumonia diagnosed by CXR was 0.91 (95% CI: 0.90 to 0.93), 1.00 (95% CI: 0.99 to 1.00), and 369.66 (95% CI: 137.14 to 996.47), respectively. Four clinical signs, including pulmonary consolidation, positive air bronchogram, abnormal pleural line, and pleural effusion were most frequently observed using LUS in the screening of children with pneumonia.Conclusions:The available evidence suggests that LUS is a reliable, valuable, and alternative method to CXR for the diagnosis of pediatric pneumonia.

Diagnosing Pneumonia by Medical History and Physical Examination

Postoperative pneumonia is a frequent complication after cardiac surgery, and its diagnosis is difficult. Little is known about the diagnostic accuracy of lung ultrasound (LUS) in the detection of pneumonia in cardiac surgical patients. The substitution of chest radiography by colour Doppler LUS (LUS-sCPIS) in the simplified clinical pulmonary infection score (sCPIS) could improve the diagnosis of pneumonia following cardiac surgery. The aim of this study was to compare the diagnostic accuracy of LUS-sCPIS and of sCPIS alone in the detection of postoperative pneumonia after cardiac surgery. A prospective study of diagnostic accuracy. A Surgical Intensive Care Unit of a French University Hospital. Fifty-one patients with acute respiratory failure within 72 h after cardiac surgery were enrolled between January and May 2015. The two index tests, LUS-sCPIS and sCPIS, were calculated for all patients at the onset of acute respiratory failure. The reference standard for the diagnosis of pneumonia was based on the consensus of three physicians, blind to the sCPIS and LUS-sCPIS data, based on a posthoc review of all the clinical, radiological and microbiological evidence. The diagnostic accuracy of LUS-sCPIS was compared with that of sCPIS in the detection of postoperative pneumonia. Pneumonia was diagnosed in 26 out of 51 patients. The LUS-sCPIS detected the presence of pneumonia with a sensitivity of 92% (95% CI 0.85 to 0.99) and a specificity of 68% (95% CI 0.55 to 0.81). The sCPIS detected the presence of pneumonia with a sensitivity of 35% (95% CI 0.22 to 0.48) and a specificity of 84% (95% CI 0.74 to 0.94). The area under the curve (AUC) of LUS-sCPIS at 0.80 (95% CI 0.69 to 0.91) was higher than the AUC of sCPIS at 0.59 (95% CI 0.47 to 0.71; P = 0.0008). Compared with sCPIS, LUS-sCPIS improved diagnostic accuracy in the detection of postoperative pneumonia in patients with acute respiratory failure after cardiac surgery. It could be a useful bedside tool to guide pneumonia management. Clinicaltrials.gov identifier: NCT03279887.

The diagnosis of community-acquired pneumonia (CAP) is based mainly on the patient's medical history and physical examination. However, in severe cases a further evaluation including chest X-ray (CXR) may be necessary. At present, lung ultrasound (LUS) is not included in the diagnostic work-up of pneumonia. To describe the ultrasonographic appearance of CAP at presentation and during the follow-up. A total of 102 patients with clinical signs and symptoms suggesting pneumonia, who underwent a clinically driven CXR, were evaluated by LUS on the same day. LUS signs of pneumonia included subpleural lung consolidation, B-lines, pleural line abnormalities, and pleural effusion. The diagnostic gold standard was the ex-post diagnosis of pneumonia made by two independent experienced pediatricians on the basis of clinical presentation, CXR and clinical course following British Thoracic Guidelines recommendations. A final diagnosis of pneumonia was confirmed in 89/102 patients. LUS was positive for the diagnosis of pneumonia in 88/89 patients, whereas CXR was positive in 81/89. Only one patient with normal LUS examination had an abnormal CXR, whereas 8 patients with normal CXR had an abnormal LUS. LUS was able to detect pleural effusion resulting from complicated pneumonia in 16 cases, whereas CXR detected pleural effusion in 3 cases. LUS is a simple and reliable imaging tool, not inferior to CXR in identifying pleuro-pulmonary alterations in children with suspected pneumonia. During the course of the disease, LUS allows a radiation-free follow-up of these abnormalities.

The diagnosis of pediatric community-acquired pneumonia (CAP) is based on clinical criteria. Even though chest x-ray (CXR) is only recommended in severe cases, it is often requested from physicians in mild cases, thus increasing radiation exposure. Lung ultrasound (LUS) is not included in the diagnostic workup. The objective of this study was to evaluate the diagnostic performance of LUS against CXR. Children who presented to the emergency department with clinical signs suggesting CAP and had already been evaluated with a CXR were included in the study. Availability of a pediatric sonographer expert in LUS was also considered a criterion for participation. Chest x-ray and LUS were considered positive for CAP in cases of alveolar or interstitial pattern of disease. The diagnostic criterion standard was the ex post diagnosis of pneumonia, made by an independent senior expert pediatrician, after evaluation of the complete medical chart. Sixty-nine children were enrolled in the study, with 66 of 69 positive for CAP. Receiver operating characteristic curve analysis results for CXR were 95.5% sensitivity and 100% specificity, whereas for LUS, sensitivity was reported 92.42% and specificity 100%. Comparison of the 2 receiver operating characteristic curves revealed no difference in the diagnostic value of the 2 methods for the diagnosis of pneumonia (P = 0.658). However, LUS classified more cases as alveolar disease compared with CXR. Lung ultrasound plays a significant role in the detection of CAP, not inferior to CXR. The aim of this study was to encourage the use of ultrasound as a first-line examination for CAP in children.

Pneumonia is the most common infectious cause of mortality in children worldwide. Chest x-ray (CXR) has been used as a supplementary mode in the diagnosis of pneumonia in children, but its frequent use might expose children to unnecessary ionizing radiation. In this review, we present up-to-date data of an alternative mode of imaging other than CXR in the diagnosis of pneumonia in children. We found that lung ultrasound is a safe and accurate mode of imaging that can be used by a health care provider in the cases of suspected pneumonia. It is more sensitive than CXR in the diagnosis of pneumonia and obviates the need for irradiation.

Background: Pneumonia is a common and serious infec-tious disease that can cause high mortality. Lung ultrasonog-raphy is being increasingly utilized in emergency and critical settings. The role of Lung Ultrasound (LUS) in the diagnosis and follow-up of pneumonia is becoming more and more important.Aim of the Work: To compare the diagnostic accuracy of LUS against a referent Chest X-Ray (CXR), chest contrast-enhanced Computerized Tomography (CT) scan and/or clinical criteria for diagnosis and follow-up of pneumonia in critically ill adult patients.Patients and Methods: We enrolled 32 (11M, 21F) multi-morbid patients aged 61.31±12.13 years from March 2016 to October 2016. Each participant underwent CXR and bedside LUS within 6 hours from Intensive Care Unit (ICU) admission. LUS was performed by skilled clinicians, blinded to CXR results and clinical history. The final diagnosis (pneumonia vs. no-pneumonia) was established by another clinician re-viewing clinical and laboratory data independent of LUS results and possibly prescribing chest contrast-enhanced CT. Diagnostic parameters of CXR and LUS were compared.Results: 28 patients (87.5%) out of 32 patients with positive LUS had a final diagnosis of pneumonia. LUS was falsely positive in two cases (6.2%) and false negative in two patients (6.2%). The sensitivity and the specificity of LUS were 87.5% (95% CI 78.9-92.7%) and 89.3% (95% CI78.3- 91.9%) respectively.Conclusion: The study supports that LUS when conducted by highly-skilled sonographers, performs well for the diagnosis of pneumonia. Intensivist and Emergency Medicine physicians should be encouraged to learn LUS since it appears to be an established diagnostic tool in the hands of experienced phy-sicians.

Chest radiographs are the best method for diagnosing pneumonia but are often not available in developing countries. Therefore, in 1990 the World Health Organization developed guidelines for diagnosis of non-severe...

BackgroundIn developing countries, pneumonia is one of the leading causes of death in children under five years of age and hence timely and accurate diagnosis is critical. In North America, pneumonia is also a common source of childhood morbidity and occasionally mortality. Clinicians traditionally have used the chest radiograph as the gold standard in the diagnosis of pneumonia, but they are becoming increasingly aware that it is not ideal. Numerous studies have shown that chest radiography findings lack precision in defining the etiology of childhood pneumonia. There is no single test that reliably distinguishes bacterial from non-bacterial causes. These factors have resulted in clinicians historically using a combination of physical signs and chest radiographs as a ‘gold standard’, though this combination of tests has been shown to be imperfect for diagnosis and assigning treatment. The objectives of this systematic review are to: 1) identify and categorize studies that have used single or multiple tests as a gold standard for assessing accuracy of other tests, and 2) given the ‘gold standard’ used, determine the accuracy of these other tests for diagnosing childhood bacterial pneumonia.Methods and FindingsSearch strategies were developed using a combination of subject headings and keywords adapted for 18 electronic bibliographic databases from inception to May 2008. Published studies were included if they: 1) included children one month to 18 years of age, 2) provided sufficient data regarding diagnostic accuracy to construct a 2×2 table, and 3) assessed the accuracy of one or more index tests as compared with other test(s) used as a ‘gold standard’. The literature search revealed 5,989 references of which 256 were screened for inclusion, resulting in 25 studies that satisfied all inclusion criteria. The studies examined a range of bacterium types and assessed the accuracy of several combinations of diagnostic tests. Eleven different gold standards were studied in the 25 included studies. Criterion validity was calculated for fourteen different index tests using eleven different gold standards. The most common gold standard utilized was blood culture tests used in six studies. Fourteen different tests were measured as index tests. PCT was the most common measured in five studies each with a different gold standard.ConclusionsWe have found that studies assessing the diagnostic accuracy of clinical, radiological, and laboratory tests for bacterial childhood pneumonia have used a heterogeneous group of gold standards, and found, at least in part because of this, that index tests have widely different accuracies. These findings highlight the need for identifying a widely accepted gold standard for diagnosis of bacterial pneumonia in children.

BackgroundGuidelines currently do not recommend the routine use of chest x-ray (CXR) in bronchiolitis. However, CXR is still performed in a high percentage of cases, mainly to diagnose or rule out pneumonia. The inappropriate use of CXR results in children exposure to ionizing radiations and increased medical costs. Lung Ultrasound (LUS) has become an emerging diagnostic tool for diagnosing pneumonia in the last decades. The purpose of this study was to assess the diagnostic accuracy and reliability of LUS for the detection of pneumonia in hospitalized children with bronchiolitis and to evaluate the agreement between LUS and CXR in diagnosing pneumonia in these patients.MethodsWe enrolled children admitted to our hospital in 2016–2017 with a diagnosis of bronchiolitis and undergone CXR because of clinical suspicion of concomitant pneumonia. LUS was performed in each child by a pediatrician blinded to the patient’s clinical, laboratory and CXR findings. An exploratory analysis was done in the first 30 patients to evaluate the inter-observer agreement between a pediatrician and a radiologist who independently performed LUS. The diagnosis of pneumonia was established by an expert clinician based on the recommendations of the British Thoracic Society guidelines.ResultsEighty seven children with bronchiolitis were investigated. A final diagnosis of concomitant pneumonia was made in 25 patients. Sensitivity and specificity of LUS for the diagnosis of pneumonia were 100% and 83.9% respectively, with an area under-the-curve of 0.92, while CXR showed a sensitivity of 96% and specificity of 87.1%. When only consolidation > 1 cm was considered consistent with pneumonia, the specificity of LUS increased to 98.4% and the sensitivity decreased to 80.0%, with an area under-the-curve of 0.89. Cohen’s kappa between pediatrician and radiologist sonologists in the first 30 patients showed an almost perfect agreement in diagnosing pneumonia by LUS (K 0.93).ConclusionsThis study shows the good accuracy of LUS in diagnosing pneumonia in children with clinical bronchiolitis. When including only consolidation size > 1 cm, specificity of LUS was higher than CXR, avoiding the need to perform CXR in these patients. Added benefit of LUS included high inter-observer agreement.Trial registrationIdentifier: NCT03280732. Registered 12 September 2017 (retrospectively registered).

Usefulness of Lung Ultrasound in the Diagnosis of Community-acquired Pneumonia in Children

Objective: The objective of this study was to determine the diagnostic accuracy of chest sonography (CS) in diagnosis of pneumonia among children.Materials and Methods: This present cross-sectional analytical diagnostic study was conducted in the department of pediatrics of Western Odisha over a period of 2 years after institutional ethics committee approval. Totally 1475 children with clinical diagnosis of pneumonia were enrolled by a simple convenient sampling technique after getting written informed consent from their parents and/or guardian. The children with congenital anomalies, immunocompromised conditions, critically ill, and CS done after 24 h of chest radiography (CR) were excluded from the study. CS and CR were performed on 1475 children, and CR was taken as the gold standard for diagnosis of pneumonia in this study. All the relevant statistics were done by appropriate statistical software.Results: Out of 1475 children, 911 (68.8%) were male and 563 (38.2%) were female. The mean age of the study population was 3.94 ± 3.42 years. The sensitivity of CS was 99.6% (98.8%, 99.9%), specificity: 89.3 (86.8%–91.4%), positive predictive value: 90.5% (88.3%–92.4%), negative predictive value: 99.5% (98.7%–99.9%), and diagnostic accuracy of 83.62%.Conclusion: CS can be used as a good screening tool for early detection of pneumonia among children.

Lung ultrasound (LUS) in the emergency department (ED) has shown a significant role in the diagnostic workup of pulmonary edema, pneumothorax and pleural effusions. The aim of this study is to assess the reliability of LUS for the diagnosis of acute pneumonia compared to chest X-ray (CXR) study. The study was conducted from September 2013 to March 2015. 107 patients were admitted to the ED with a clinical appearance of pneumonia. All the patients underwent a CXR study, read by a radiologist, and an LUS, performed by a trained ED physician on duty. Among the 105 patients, 68 were given a final diagnosis of pneumonia. We found a sensitivity of 0.985 and a specificity of 0.649 for LUS, and a sensitivity of 0.735 and specificity of 0.595 for CXR. The positive predictive value for LUS was 0.838 against 0.7 for CXR. The negative predictive value of LUS was 0.960 versus 0.550 for CXR. This study has shown sensitivity, positive predictive value and negative predictive value of LUS compared to the CXR study for the diagnosis of acute pneumonia. These results suggest the use of bedside thoracic US first-line diagnostic tool in patients with suspected pneumonia.

Current WHO algorithm has retained the signs and symptoms used in the older version for classifying severity of childhood pneumonia. To study the role of clinical features (including that of current WHO criteria), and oxygen saturation (SpO2) in the diagnosis of childhood pneumonia. Multicenter prospective cohort study. Children, 2 to 59 months of age, suffering from acute respiratory infection (ARI). Sensitivity, specificity, and likelihood ratios were calculated for clinical features, and SpO2. Of a total 7026 children with ARI enrolled, 13.4% had pneumonia (37% of them had severe pneumonia), according to WHO criteria. Based on any abnormality on chest x ray (CXR), 46% had pneumonia. The sensitivity and specificity of the existing WHO criteria for diagnosis of pneumonia was 56.5% and 66.2%, respectively, when compared against abnormalities in CXR. Cough and fever, each had sensitivity of >80%. Audible wheeze and breathing difficulty, each had a specificity of >80%. Sensitivity and specificity of tachypnoea were 58.7% and 63.3%, respectively. None of the clinical features alone had a sensitivity and specificity of >80%. Addition of SpO2 of <92% to chest indrawing alone or WHO criteria increased the likelihood of diagnosis of pneumonia. Current WHO criteria based on rapid respiratory rate and/or chest indrawing has modest sensitivity and specificity, considering CXR abnormalities as gold standard for diagnosis of pneumonia. Addition of SpO2 of <92% to chest indrawing alone or WHO criteria increases the probability of pneumonia diagnosis, and is important in the management of a child with pneumonia.

<b>Background:</b> Current WHO algorithm has retained the signs and symptoms used in the older version for classifying severity of childhood pneumonia. <b>Objective:</b> To study the role of clinical features (including that of current WHO criteria), and oxygen saturation (SpO2) in the diagnosis of childhood pneumonia. <b>Methods:</b> This multicenter prospective cohort study was conducted at six sites in India. Sensitivity, specificity, and likelihood ratios were calculated for clinical features, and SpO2. <b>Results:</b> Of a total 7026 children with ARI enrolled, 13.4% had pneumonia (37% of them had severe pneumonia), according to WHO criteria (Figure 1). Based on any abnormality on chest x ray (CXR), 46% had pneumonia. The sensitivity and specificity of the existing WHO criteria for diagnosis of pneumonia was 56.5% and 66.2%, respectively, when compared against abnormalities in CXR. Cough and fever, each had sensitivity of &gt;80%. Audible wheeze and breathing difficulty, each had a specificity of &gt;80%. Sensitivity and specificity of tachypnoea were 58.7% and 63.3%, respectively. None of the clinical features alone had a sensitivity and specificity of &gt;80%. Addition of SpO2 of &lt;92% to chest indrawing alone or WHO criteria increased the likelihood of diagnosis of pneumonia. <b>Conclusions:</b> Addition of SpO2 of &lt;92% to chest indrawing alone or WHO criteria increases the probability of pneumonia diagnosis, and is important in the management of a child with pneumonia.