OBJECTIVESThis study aimed to develop a dual-input convolutional neural network (CNN)–based deep-learning algorithm that utilizes both anteroposterior (AP) and lateral elbow radiographs for the ...automated detection of pediatric supracondylar fracture in conventional radiography, and assess its feasibility and diagnostic performance.
MATERIALS AND METHODSTo develop the deep-learning model, 1266 pairs of AP and lateral elbow radiographs examined between January 2013 and December 2017 at a single institution were split into a training set (1012 pairs, 79.9%) and a validation set (254 pairs, 20.1%). We performed external tests using 2 types of distinct datasetsone temporally and the other geographically separated from the model development. We used 258 pairs of radiographs examined in 2018 at the same institution as a temporal test set and 95 examined between January 2016 and December 2018 at another hospital as a geographic test set. Images underwent preprocessing, including cropping and histogram equalization, and were input into a dual-input neural network constructed by merging 2 ResNet models. An observer study was performed by radiologists on the geographic test set. The area under the receiver operating characteristic curve (AUC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the model and human readers were calculated and compared.
RESULTSOur trained model showed an AUC of 0.976 in the validation set, 0.985 in the temporal test set, and 0.992 in the geographic test set. In AUC comparison, the model showed comparable results to the human readers in the geographic test set; the AUCs of human readers were in the range of 0.977 to 0.997 (Pʼs > 0.05). The model had a sensitivity of 93.9%, a specificity of 92.2%, a PPV of 80.5%, and an NPV of 97.8% in the temporal test set, and a sensitivity of 100%, a specificity of 86.1%, a PPV of 69.7%, and an NPV of 100% in the geographic test set. Compared with the developed deep-learning model, all 3 human readers showed a significant difference (Pʼs < 0.05) using the McNemar test, with lower specificity and PPV in the model. On the other hand, there was no significant difference (Pʼs > 0.05) in sensitivity and NPV between all 3 human readers and the proposed model.
CONCLUSIONSThe proposed dual-input deep-learning model that interprets both AP and lateral elbow radiographs provided an accurate diagnosis of pediatric supracondylar fracture comparable to radiologists.
Cone-beam computed tomography (CBCT) systems have been designed for imaging hard tissues of the maxillofacial region. CBCT is capable of providing sub-millimetre resolution in images of high ...diagnostic quality, with short scanning times (10-70 seconds) and radiation dosages reportedly up to 15 times lower than those of conventional CT scans. Increasing availability of this technology provides the dental clinician with an imaging modality capable of providing a 3-dimensional representation of the maxillofacial skeleton with minimal distortion. This article provides an overview of currently available maxillofacial CBCT systems and reviews the specific application of various CBCT display modes to clinical dental practice.
Due to the coronavirus pandemic, all routine dental care in the UK ceased on 25 March 2020. Liverpool University Dental Hospital (LUDH) responded by commencing an emergency dental service on the same ...date. Clinicians were redeployed within the Hospital to meet the needs of the service, including staffing of the radiology department. LUDH followed Royal College recommendations by taking extraoral radiographs in preference to intraoral radiographs due to the risk of inducing an aerosol. Issues were identified with clinical diagnosis from sectional panoramic radiographs, which led to the introduction of extraoral bitewings being taken as an alternative. A quality assurance audit found that these images provided a substantially lower radiation dose and produced excellent quality images with improved diagnostic accuracy. This article aims to summarise how our radiography practices changed in response to the coronavirus and how the lessons that we have learnt provide an opportunity to modify and improve future practice, beyond the pandemic.
The primary objective was to study the degree of agreement between the chest ultrasound (CUS) studies and chest x-ray (CXR) studies in postoperative pediatric cardiac surgical patients regarding the ...diagnosis of thoracic abnormalities, and also to compare the diagnostic performance of CUS in reference to CXR for the detection of thoracic abnormalities. The secondary objective was to compare the necessity for interventions done on the basis of CUS and CXR findings in the postoperative setting.
A prospective observational study.
At a postoperative pediatric cardiac surgical intensive care unit in a tertiary-care center.
One hundred sixty patients between the age of 2 months to 18 years undergoing elective cardiac surgery for various congenital heart diseases.
After obtaining permission from the institutional ethics committee, 160 pediatric cardiac surgical patients were studied prospectively in the postoperative period. On the day of surgery (postoperative day POD 0), bedside CXR was done in the immediate postoperative period. After bedside CXR, CUS examination was performed and then interpreted by the principal investigator. The CXR was interpreted by the surgical team. Provisional diagnosis was made by the principal investigator and surgical team. Any intervention required was decided based on CXR or CUS findings or both. The procedure was repeated in the morning of POD 1.
The degree of agreement between CUS studies and CXR studies in detecting abnormalities was evaluated by Cohen's kappa (k) statistics. The diagnostic performance of CUS was compared with that of CXR using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy. Overall, kappa analysis (k) showed substantial agreement between the findings of the CUS and CXR studies (k = 0.749). The diagnostic performance of CUS, as compared with CXR, was found to have a sensitivity of 96.9%, specificity of 84.75%, PPV of 73.4%, NPV of 98.43%, and diagnostic accuracy of 88.44%. In 94 abnormal findings, the interventions were done based on CUS or CXR findings or both. Overall, there was a substantial agreement (k = 0.787) between CUS and CXR regarding the necessity for interventions.
The degree of agreement between CUS and CXR studies was substantial for atelectasis, interstitial edema, and diaphragmatic weakness. The degree of agreement between CUS and CXR studies was almost perfect for pneumothorax and fair for pleural effusion. More CUS studies detected intrathoracic pathologies than CXR studies. The CUS also detected abnormalities earlier than CXR and was found to be useful for the early institution of intervention therapy in patients with interstitial edema and atelectasis. It would be reasonable to conclude that CUS may be considered in some instances as an alternative to CXR.
Purpose:
The authors previously proposed an image‐based technique Y. Lin et al. Med. Phys. 39, 7019–7031 (2012) to assess the perceptual quality of clinical chest radiographs. In this study, an ...observer study was designed and conducted to validate the output of the program against rankings by expert radiologists and to establish the ranges of the output values that reflect the acceptable image appearance so the program output can be used for image quality optimization and tracking.
Methods:
Using an IRB‐approved protocol, 2500 clinical chest radiographs (PA/AP) were collected from our clinical operation. The images were processed through our perceptual quality assessment program to measure their appearance in terms of ten metrics of perceptual image quality: lung gray level, lung detail, lung noise, rib–lung contrast, rib sharpness, mediastinum detail, mediastinum noise, mediastinum alignment, subdiaphragm–lung contrast, and subdiaphragm area. From the results, for each targeted appearance attribute/metric, 18 images were selected such that the images presented a relatively constant appearance with respect to all metrics except the targeted one. The images were then incorporated into a graphical user interface, which displayed them into three panels of six in a random order. Using a DICOM calibrated diagnostic display workstation and under low ambient lighting conditions, each of five participating attending chest radiologists was tasked to spatially order the images based only on the targeted appearance attribute regardless of the other qualities. Once ordered, the observer also indicated the range of image appearances that he/she considered clinically acceptable. The observer data were analyzed in terms of the correlations between the observer and algorithmic rankings and interobserver variability. An observer‐averaged acceptable image appearance was also statistically derived for each quality attribute based on the collected individual acceptable ranges.
Results:
The observer study indicated that, for each image quality attribute, the averaged observer ranking strongly correlated with the algorithmic ranking (linear correlation coefficient R > 0.92), with highest correlation (R = 1) for lung gray level and the lowest (R = 0.92) for mediastinum noise. There was a strong concordance between the observers in terms of their rankings (i.e., Kendall's tau agreement > 0.84). The observers also generally indicated similar tolerance and preference levels in terms of acceptable ranges, as 85% of the values were close to the overall tolerance or preference levels and the differences were smaller than 0.15.
Conclusions:
The observer study indicates that the previously proposed technique provides a robust reflection of the perceptual image quality in clinical images. The results established the range of algorithmic outputs for each metric that can be used to quantitatively assess and qualify the appearance quality of clinical chest radiographs.