Added Value of Ultra-low-dose Computed Tomography, Dose Equivalent to Chest X-Ray Radiography, for Diagnosing Chest Pathology

Reinvestigating the performance of artificial intelligence classification algorithms on COVID-19 X-Ray and CT images

There are concerns that artificial intelligence (AI) algorithms may create underdiagnosis bias by mislabeling patient individuals with certain attributes (e.g., female and young) as healthy. Addressing this bias is crucial given the urgent need for AI diagnostics facing rapidly spreading infectious diseases like COVID-19. We find the prevalent AI diagnostic models show an underdiagnosis rate among specific patient populations, and the underdiagnosis rate is higher in some intersectional specific patient populations (for example, females aged 20-40 years). Additionally, we find training AI models on heterogeneous datasets (positive and negative samples from different datasets) may lead to poor model generalization. The model's classification performance varies significantly across test sets, with the accuracy of the better performance being over 40% higher than that of the poor performance. In conclusion, we developed an AI bias analysis pipeline to help researchers recognize and address biases that impact medical equality and ethics.

Optimizing Diagnosis and Management of Community-acquired Pneumonia in the Emergency Department

Pneumonia is split into 3 diagnostic categories: community-acquired pneumonia (CAP), health care-associated pneumonia, and ventilator-associated pneumonia. This classification scheme is driven not only by the location of infection onset but also by the predominant associated causal microorganisms. Pneumonia is diagnosed in over 1.5 million US emergency department visits annually (1.2% of all visits), and most pneumonia diagnosed by emergency physicians is CAP.

Utilizing Deep Learning and Computed Tomography to Determine Pulmonary Nodule Activity in Patients With Nontuberculous Mycobacterial-Lung Disease

PURPOSE

To develop and evaluate a deep convolutional neural network (DCNN) model for the classification of acute and chronic lung nodules from nontuberculous mycobacterial-lung disease (NTM-LD) on computed tomography (CT).

MATERIALS AND METHODS

We collected a data set of 650 nodules (316 acute and 334 chronic) from the CT scans of 110 patients with NTM-LD. The data set was divided into training, validation, and test sets in a ratio of 4:1:1. Bounding boxes were used to crop the 2D CT images down to the area of interest. A DCNN model was built using 11 convolutional layers and trained on these images. The performance of the model was evaluated on the hold-out test set and compared with that of 3 radiologists who independently reviewed the images.

RESULTS

The DCNN model achieved an area under the receiver operating characteristic curve of 0.806 for differentiating acute and chronic NTM-LD nodules, corresponding to sensitivity, specificity, and accuracy of 76%, 68%, and 72%, respectively. The performance of the model was comparable to that of the 3 radiologists, who had area under the receiver operating characteristic curve, sensitivity, specificity, and accuracy of 0.693 to 0.771, 61% to 82%, 59% to 73%, and 60% to 73%, respectively.

CONCLUSIONS

This study demonstrated the feasibility of using a DCNN model for the classification of the activity of NTM-LD nodules on chest CT. The model performance was comparable to that of radiologists. This approach can potentially and efficiently improve the diagnosis and management of NTM-LD.

Photon Counting Computed Tomography with the Radiation Dose of a Chest X-Ray: Feasibility and Diagnostic Yield

INTRODUCTION

Photon counting (PC) detectors allow a reduction of the radiation dose in CT. Chest X-ray (CXR) is known to have a low sensitivity and specificity for detection of pneumonic infiltrates. The aims were to establish an ultra-low-dose CT (ULD-CT) protocol at a PC-CT with the radiation dose comparable to the dose of a CXR and to evaluate its clinical yield in patients with suspicion of pneumonia.

METHODS

A ULD-CT protocol was established with the aim to meet the radiation dose of a CXR. In this retrospective study, all adult patients who received a ULD-CT of the chest with suspected pneumonia were included. Radiation exposure of ULD-CT and CXR was calculated. The clinical significance (new diagnosis, change of therapy, additional findings) and limitations were evaluated by a radiologist and a pulmonologist considering previous CXR and clinical data.

RESULTS

Twenty-seven patients (70% male, mean age 68 years) were included. With our ULD-CT protocol, the radiation dose of a CXR could be reached (mean radiation exposure 0.11 mSv). With ULD-CT, the diagnosis changed in 11 patients (41%), there were relevant additional findings in 4 patients (15%), an infiltrate (particularly fungal infiltrate under immunosuppression) could be ruled out with certainty in 10 patients (37%), and the therapy changed in 10 patients (37%). Two patients required an additional CT with contrast medium to rule out a pulmonary embolism or pleural empyema.

CONCLUSIONS

With ULD-CT, the radiation dose of a CXR could be reached while the clinical impact is higher with change in diagnosis in 41%.

Detection of COVID-19 using edge devices by a light-weight convolutional neural network from chest X-ray images

Deep learning is a highly significant technology in clinical treatment and diagnostics nowadays. Convolutional Neural Network (CNN) is a new idea in deep learning that is being used in the area of computer vision. The COVID-19 detection is the subject of our medical study. Researchers attempted to increase the detection accuracy but at the cost of high model complexity. In this paper, we desire to achieve better accuracy with little training space and time so that this model easily deployed in edge devices. In this paper, a new CNN design is proposed that has three stages: pre-processing, which removes the black padding on the side initially; convolution, which employs filter banks; and feature extraction, which makes use of deep convolutional layers with skip connections. In order to train the model, chest X-ray images are partitioned into three sets: learning(0.7), validation(0.1), and testing(0.2). The models are then evaluated using the test and training data. The LMNet, CoroNet, CVDNet, and Deep GRU-CNN models are the other four models used in the same experiment. The propose model achieved 99.47% & 98.91% accuracy on training and testing respectively. Additionally, it achieved 97.54%, 98.19%, 99.49%, and 97.86% scores for precision, recall, specificity, and f1-score respectively. The proposed model obtained nearly equivalent accuracy and other similar metrics when compared with other models but greatly reduced the model complexity. Moreover, it is found that proposed model is less prone to over fitting as compared to other models.

Effective processing pipeline PACE 2.0 for enhancing chest x-ray contrast and diagnostic interpretability

Preprocessing is an essential task for the correct analysis of digital medical images. In particular, X-ray imaging might contain artifacts, low contrast, diffractions or intensity inhomogeneities. Recently, we have developed a procedure named PACE that is able to improve chest X-ray (CXR) images including the enforcement of clinical evaluation of pneumonia originated by COVID-19. At the clinical benchmark state of this tool, there have been found some peculiar conditions causing a reduction of details over large bright regions (as in ground-glass opacities and in pleural effusions in bedridden patients) and resulting in oversaturated areas. Here, we have significantly improved the overall performance of the original approach including the results in those specific cases by developing PACE2.0. It combines 2D image decomposition, non-local means denoising, gamma correction, and recursive algorithms to improve image quality. The tool has been evaluated using three metrics: contrast improvement index, information entropy, and effective measure of enhancement, resulting in an average increase of 35% in CII, 7.5% in ENT, 95.6% in EME and 13% in BRISQUE against original radiographies. Additionally, the enhanced images were fed to a pre-trained DenseNet-121 model for transfer learning, resulting in an increase in classification accuracy from 80 to 94% and recall from 89 to 97%, respectively. These improvements led to a potential enhancement of the interpretability of lesion detection in CXRs. PACE2.0 has the potential to become a valuable tool for clinical decision support and could help healthcare professionals detect pneumonia more accurately.

Ultra-low-dose CT vs. chest X-ray in non-traumatic emergency department patients - a prospective randomised crossover cohort trial

Background

Ultra-low-dose CT (ULDCT) examinations of the chest at only twice the radiation dose of a chest X-ray (CXR) now offer a valuable imaging alternative to CXR. This trial prospectively compares ULDCT and CXR for the detection rate of diagnoses and their clinical relevance in a low-prevalence cohort of non-traumatic emergency department patients.

Methods

In this prospective crossover cohort trial, 294 non-traumatic emergency department patients with a clinically indicated CXR were included between May 2nd and November 26th of 2019 (www.clinicaltrials.gov: NCT03922516). All participants received both CXR and ULDCT, and were randomized into two arms with inverse reporting order. The detection rate of CXR was calculated from 'arm CXR' (n = 147; CXR first), and of ULDCT from 'arm ULDCT' (n = 147; ULDCT first). Additional information reported by the second exam in each arm was documented. From all available clinical and imaging data, expert radiologists and emergency physicians built a compound reference standard, including radiologically undetectable diagnoses, and assigned each finding to one of five clinical relevance categories for the respective patient.

Findings

Detection rates for main diagnoses by CXR and ULDCT (mean effective dose: 0.22 mSv) were 9.1% (CI [5.2, 15.5]; 11/121) and 20.1% (CI [14.2, 27.7]; 27/134; P = 0.016), respectively. As an additional imaging modality, ULDCT added 9.1% (CI [5.2, 15.5]; 11/121) of main diagnoses to prior CXRs, whereas CXRs did not add a single main diagnosis (0/134; P < 0.001). Notably, ULDCT also offered higher detection rates than CXR for all other clinical relevance categories, including findings clinically irrelevant for the respective emergency department visit with 78.5% (CI [74.0, 82.5]; 278/354) vs. 16.2% (CI [12.7, 20.3]; 58/359) as a primary modality and 68.2% (CI [63.3, 72.8]; 245/359) vs. 2.5% (CI [1.3, 4.7]; 9/354) as an additional imaging modality.

Interpretation

In non-traumatic emergency department patients, ULDCT of the chest offered more than twice the detection rate for main diagnoses compared to CXR.

Funding

The Department of Biomedical Imaging and Image-guided Therapy of Medical University of Vienna received funding from Siemens Healthineers (Erlangen, Germany) to employ two research assistants for one year.

The yield of chest X-ray or ultra-low-dose chest-CT in emergency department patients suspected of pulmonary infection without respiratory symptoms or signs

OBJECTIVE

The yield of pulmonary imaging in patients with suspected infection but no respiratory symptoms or signs is probably limited, ultra-low-dose CT (ULDCT) is known to have a higher sensitivity than Chest X-ray (CXR). Our objective was to describe the yield of ULDCT and CXR in patients clinically suspected of infection, but without respiratory symptoms or signs, and to compare the diagnostic accuracy of ULDCT and CXR.

METHODS

In the OPTIMACT trial, patients suspected of non-traumatic pulmonary disease at the emergency department (ED) were randomly allocated to undergo CXR (1210 patients) or ULDCT (1208 patients). We identified 227 patients in the study group with fever, hypothermia, and/or elevated C-reactive protein (CRP) but no respiratory symptoms or signs, and estimated ULDCT and CXR sensitivity and specificity in detecting pneumonia. The final day-28 diagnosis served as the clinical reference standard.

RESULTS

In the ULDCT group, 14/116 (12%) received a final diagnosis of pneumonia, versus 8/111 (7%) in the CXR group. ULDCT sensitivity was significantly higher than that of CXR: 13/14 (93%) versus 4/8 (50%), a difference of 43% (95% CI: 6 to 80%). ULDCT specificity was 91/102 (89%) versus 97/103 (94%) for CXR, a difference of - 5% (95% CI: - 12 to 3%). PPV was 54% (13/24) for ULDCT versus 40% (4/10) for CXR, NPV 99% (91/92) versus 96% (97/101).

CONCLUSION

Pneumonia can be present in ED patients without respiratory symptoms or signs who have a fever, hypothermia, and/or elevated CRP. ULDCT's sensitivity is a significant advantage over CXR when pneumonia has to be excluded.

CLINICAL RELEVANCE STATEMENT

Pulmonary imaging in patients with suspected infection but no respiratory symptoms or signs can result in the detection of clinically significant pneumonia. The increased sensitivity of ultra-low-dose chest CT compared to CXR is of added value in vulnerable and immunocompromised patients.

KEY POINTS

• Clinical significant pneumonia does occur in patients who have a fever, low core body temperature, or elevated CRP without respiratory symptoms or signs. • Pulmonary imaging should be considered in patients with unexplained symptoms or signs of infections. • To exclude pneumonia in this patient group, ULDCT's improved sensitivity is a significant advantage over CXR.

Associated radiation exposure from medical imaging and excess lifetime risk of developing cancer in pediatric patients with pulmonary hypertension

Pediatric patients with pulmonary hypertension (PH) receive imaging studies that use ionizing radiation (radiation) such as computed tomography (CT) and cardiac catheterization to guide clinical care. Radiation exposure is associated with increased cancer risk. It is unknown how much radiation pediatric PH patients receive. The objective of this study is to quantify radiation received from imaging and compute associated lifetime cancer risks for pediatric patients with PH. Electronic health records between 2012 and 2022 were reviewed and radiation dose data were extracted. Organ doses were estimated using Monte Carlo modeling. Cancer risks for each patient were calculated from accumulated exposures using National Cancer Institute tools. Two hundred and forty-nine patients with PH comprised the study cohort; 97% of patients had pulmonary arterial hypertension, PH due to left heart disease, or PH due to chronic lung disease. Mean age at the time of the first imaging study was 2.5 years (standard deviation [SD] = 4.9 years). Patients underwent a mean of 12 studies per patient per year, SD = 32. Most (90%) exams were done in children <5 years of age. Radiation from CT and cardiac catheterization accounted for 88% of the total radiation dose received. Cumulative mean effective dose was 19 mSv per patient (SD = 30). Radiation dose exposure resulted in a mean increased estimated lifetime cancer risk of 7.6% (90% uncertainty interval 3.0%-14.2%) in females and 2.8% (1.2%-5.3%) in males. Careful consideration for the need of radiation-based imaging studies is warranted, especially in the youngest of children.

Diagnostic clues for the identification of pediatric foreign body aspirations and consideration of novel imaging techniques

OBJECTIVE

To better understand the diagnosis of foreign body aspiration by elucidating key components of its clinical presentation.

METHODS

This is a retrospective cohort study of pediatric patients with suspected foreign body aspiration. We collected information regarding demographics, history, symptoms, physical exam, imaging, and operative findings for rigid bronchoscopies. An evaluation of these findings for an association with foreign body aspiration and the overall diagnostic algorithm was performed.

RESULTS

518 pediatric patients presented with 75.2 % presenting within one day of the inciting event. Identified history findings included wheeze (OR: 5.83, p < 0.0001) and multiple encounters (OR: 5.46, p < 0.0001). Oxygen saturation was lower in patients with foreign body aspiration (97.3 %, p < 0.001). Identified physical exam findings included wheeze (OR: 7.38, p < 0.001) and asymmetric breath sounds (OR: 5.48, p < 0.0001). The sensitivity and specificity of history findings was 86.7 % and 23.1 %, physical exam was 60.8 % and 88.4 %, and chest radiographs was 45.3 % and 88.0 %. 25 CT scans were performed with a sensitivity and specificity of 100 % and 85.7 %. Combining two components of the diagnostic algorithm yielded a high sensitivity and moderate specificity; the best combination was the history and physical exam. 186 rigid bronchoscopies were performed with 65.6 % positive for foreign body aspiration.

CONCLUSION

Accurate diagnosis of foreign body aspiration requires careful history taking and examination. Low-dose CT should be included in the diagnostic algorithm. The combination of any two components of the diagnostic algorithm is the most accurate for foreign body aspiration.

A Real Time Method for Distinguishing COVID-19 Utilizing 2D-CNN and Transfer Learning

Rapid identification of COVID-19 can assist in making decisions for effective treatment and epidemic prevention. The PCR-based test is expert-dependent, is time-consuming, and has limited sensitivity. By inspecting Chest R-ray (CXR) images, COVID-19, pneumonia, and other lung infections can be detected in real time. The current, state-of-the-art literature suggests that deep learning (DL) is highly advantageous in automatic disease classification utilizing the CXR images. The goal of this study is to develop models by employing DL models for identifying COVID-19 and other lung disorders more efficiently. For this study, a dataset of 18,564 CXR images with seven disease categories was created from multiple publicly available sources. Four DL architectures including the proposed CNN model and pretrained VGG-16, VGG-19, and Inception-v3 models were applied to identify healthy and six lung diseases (fibrosis, lung opacity, viral pneumonia, bacterial pneumonia, COVID-19, and tuberculosis). Accuracy, precision, recall, f1 score, area under the curve (AUC), and testing time were used to evaluate the performance of these four models. The results demonstrated that the proposed CNN model outperformed all other DL models employed for a seven-class classification with an accuracy of 93.15% and average values for precision, recall, f1-score, and AUC of 0.9343, 0.9443, 0.9386, and 0.9939. The CNN model equally performed well when other multiclass classifications including normal and COVID-19 as the common classes were considered, yielding accuracy values of 98%, 97.49%, 97.81%, 96%, and 96.75% for two, three, four, five, and six classes, respectively. The proposed model can also identify COVID-19 with shorter training and testing times compared to other transfer learning models.

Limited Clinical Impact of Ultralow-Dose Computed Tomography in Suspected Community-Acquired Pneumonia

Patients clinically suspected of community-acquired pneumonia (CAP) were randomized between ultralow-dose chest computed tomography ([ULDCT] 261 patients) and chest radiograph ([CXR] 231 patients). We did not find evidence that performing ULDCT instead of CXR affects antibiotic treatment policy or patient outcomes. However, in a subgroup of afebrile patients, there were more patients diagnosed with CAP in the ULDCT group (ULDCT, 106 of 608 patients; CXR, 71 of 654 patients; P = .001).

Review on chest pathogies detection systems using deep learning techniques

Chest radiography is the standard and most affordable way to diagnose, analyze, and examine different thoracic and chest diseases. Typically, the radiograph is examined by an expert radiologist or physician to decide about a particular anomaly, if exists. Moreover, computer-aided methods are used to assist radiologists and make the analysis process accurate, fast, and more automated. A tremendous improvement in automatic chest pathologies detection and analysis can be observed with the emergence of deep learning. The survey aims to review, technically evaluate, and synthesize the different computer-aided chest pathologies detection systems. The state-of-the-art of single and multi-pathologies detection systems, which are published in the last five years, are thoroughly discussed. The taxonomy of image acquisition, dataset preprocessing, feature extraction, and deep learning models are presented. The mathematical concepts related to feature extraction model architectures are discussed. Moreover, the different articles are compared based on their contributions, datasets, methods used, and the results achieved. The article ends with the main findings, current trends, challenges, and future recommendations.

Bio-medical imaging (X-ray, CT, ultrasound, ECG), genome sequences applications of deep neural network and machine learning in diagnosis, detection, classification, and segmentation of COVID-19: a Meta-analysis & systematic review

This review investigates how Deep Machine Learning (DML) has dealt with the Covid-19 epidemic and provides recommendations for future Covid-19 research. Despite the fact that vaccines for this epidemic have been developed, DL methods have proven to be a valuable asset in radiologists' arsenals for the automated assessment of Covid-19. This detailed review debates the techniques and applications developed for Covid-19 findings using DL systems. It also provides insights into notable datasets used to train neural networks, data partitioning, and various performance measurement metrics. The PRISMA taxonomy has been formed based on pretrained(45 systems) and hybrid/custom(17 systems) models with radiography modalities. A total of 62 systems with respect to X-ray(32), CT(19), ultrasound(7), ECG(2), and genome sequence(2) based modalities as taxonomy are selected from the studied articles. We originate by valuing the present phase of DL and conclude with significant limitations. The restrictions contain incomprehensibility, simplification measures, learning from incomplete labeled data, and data secrecy. Moreover, DML can be utilized to detect and classify Covid-19 from other COPD illnesses. The proposed literature review has found many DL-based systems to fight against Covid19. We expect this article will assist in speeding up the procedure of DL for Covid-19 researchers, including medical, radiology technicians, and data engineers.

Ultra-low-dose emergency chest computed tomography protocols in three vendors: A technical note

Background

In suspected community-acquired pneumonia (CAP), chest CT is superior to the routinely obtained radiographs (CXR), but administers higher radiation doses. However, ultra-low-dose CT (ULDCT) has shown promising results.

Purpose

To compare radiation dose and image quality using standard and ULDCT protocols designed for a multicenter study encompassing three CT scanner models from GE, Canon, and Siemens.

Material and methods

Patients with suspected CAP were referred for non-contrast standard dose chest CT (NCCT) and ULDCT. Effective radiation dose and Contrast-to-Noise Ratio (CNR) was calculated.

Results

Mean effective doses were GE (n = 10) 6.93 mSv in NCCT and 0.27 mSv in ULDCT; Canon (n = 9) 3.48 in mSv NCCT and 1.11 mSv in ULDCT; Siemens (n = 10) 2.85 mSv in NCCT and 0.45 mSv in ULDCT. CNR was reduced by 29-39% in ULDCT.

Conclusion

The proposed CT protocols yielded dose reductions of 96%, 68%, and 84% using a GE, Canon, and Siemens scanner, respectively.

Patient's effective dose and performance assessment of computed radiography systems

Computed tomography is widely used for planar imaging. Previous studies showed that CR systems involve higher patient radiation doses compared to digital systems. Therefore, assessing the patient's dose and CR system performance is necessary to ensure that patients received minimal dose with the highest possible image quality. The study was performed at three medical diagnostic centers in Sudan: Medical Corps Hospital (MCH), Advance Diagnostic Center (ADC), and Advance Medical Center (AMC). The following tools were used in this study: Tape measure, Adhesive tape, 1.5 mm copper filtration (>10 × 10 cm), TO 20 threshold contrast test object, Resolution test object (e.g., Huttner 18), MI geometry test object or lead ruler, Contact mish, Piranha (semiconductor detector), Small lead or copper block (∼5 × 5 cm), and Steel ruler, to do a different type of tests (Dark Noise, Erasure cycle efficiency, Sensitivity Index calibration, Sensitivity Index consistency, Uniformity, Scaling errors, Blurring, Limiting spatial Resolution, Threshold, and Laser beam Function. Entrance surface air kerma (ESAK (mGy) was calculated from patient exposure parameters using DosCal software for three imaging modalities. A total of 199 patients were examined (112 chest X rays, 77 lumbar spine). The mean and standard deviation (sd) for patients ESAK (mGy) were 2.56 ± 0.1 mGy and 1.6 mGy for the Anteroposterior (AP) and lateral projections for the lumbar spine, respectively. The mean and sd for the patient's chest doses were 0.1 ± 0.01 for the chest X-ray procedures. The three medical diagnostic centers' CR system performance was evaluated and found that all of the three centers have good CR system functions. All the centers satisfy all the criteria of acceptable visual tests. CR's image quality and sensitivity were evaluated, and the CR image is good because it has good contrast and resolution. All the CR system available in the medical centers and upgraded from old X-ray systems to new systems, has been found to work well. The patient's doses were comparable for the chest X-ray procedures, while patients' doses from the lumbar spine showed variation up to 2 folds due to the variation in patients' weight and X-ray machine setting. Patients dose optimization is recommended to ensure the patients received a minimal dose while obtaining the diagnostic findings.

Diagnosing Lung Abnormalities Related to Heart Failure in Chest Radiogram, Lung Ultrasound and Thoracic Computed Tomography

Heart failure (HF) is a multidisciplinary disease affecting almost 1-2% of the adult population worldwide. Symptoms most frequently reported by patients suffering from HF include dyspnoea, cough or exercise intolerance, which is equally often observed in many pulmonary diseases. The spectrum of lung changes related to HF is wide. The knowledge of different types of these abnormalities is essential to distinguish patients with HF from patients with lung diseases or both disorders and thus avoid unnecessary diagnostics or therapies. In this review, we aimed to summarise recent research concerning the spectrum of lung abnormalities related to HF in three frequently used lung imaging techniques: chest X-ray (CXR), lung ultrasound (LUS) and chest computed tomography (CT). We discussed the most prevalent abnormalities in the above-mentioned investigations in the context of consecutive pathophysiological stages identified in HF: (i) redistribution, (ii) interstitial oedema, and (iii) alveolar oedema. Finally, we compared the utility of these imaging tools in the clinical setting. In conclusion, we consider LUS the most useful and promising imaging technique due to its high sensitivity, repeatability and accessibility. However, the value of CXR and chest CT is their potential for establishing a differential diagnosis.

Ultra-Low Dose CT Chest in Acute COVID-19 Pneumonia: A Pilot Study from India

The rapid increase in the number of CT acquisitions during the COVID-19 pandemic raised concerns about increased radiation exposure to patients and the resultant radiation-induced health risks. It prompted researchers to explore newer CT techniques like ultra-low dose CT (ULDCT), which could improve patient safety. Our aim was to study the utility of ultra-low dose CT (ULDCT) chest in the evaluation of acute COVID-19 pneumonia with standard-dose CT (SDCT) chest as a reference standard. This was a prospective study approved by the institutional review board. 60 RT-PCR positive COVID-19 patients with valid indication for CT chest underwent SDCT and ULDCT. ULDCT and SDCT were compared in terms of objective (noise and signal-to-noise ratio) and subjective (noise, sharpness, artifacts and diagnostic confidence) image quality, various imaging patterns of COVID-19, CT severity score and effective radiation dose. The sensitivity, specificity, positive and negative predictive value, and diagnostic accuracy of ULDCT for detecting lung lesions were calculated by taking SDCT as a reference standard. The mean age of subjects was 47.2 ± 10.7 years, with 66.67% being men. 90% of ULDCT scans showed no/minimal noise and sharp images, while 93.33% had image quality of high diagnostic confidence. The major imaging findings detected by SDCT were GGOs (90%), consolidation (76.67%), septal thickening (60%), linear opacities (33.33%), crazy-paving pattern (33.33%), nodules (30%), pleural thickening (30%), lymphadenopathy (30%) and pleural effusion (23.33%). Sensitivity, specificity and diagnostic accuracy of ULDCT for detecting most of the imaging patterns were 100% (p < 0.001); except for GGOs (sensitivity: 92.59%, specificity: 100%, diagnostic accuracy: 93.33%), consolidation (sensitivity: 100%, specificity: 71.43%, diagnostic accuracy: 93.33%) and linear opacity (sensitivity: 90.00%, specificity: 100%, diagnostic accuracy: 96.67%). CT severity score (range: 15-25) showed 100% concordance on SDCT and ULDCT, while effective radiation dose was 4.93 ± 1.11 mSv and 0.26 ± 0.024 mSv, respectively. A dose reduction of 94.38 ± 1.7% was achieved with ULDCT. Compared to SDCT, ULDCT chest yielded images of reasonable and comparable diagnostic quality with the advantage of significantly reduced radiation dose; thus, it can be a good alternative to SDCT in the evaluation of COVID-19 pneumonia.

A topic-based hierarchical publish/subscribe messaging middleware for COVID-19 detection in X-ray image and its metadata

Putting real-time medical data processing applications into practice comes with some challenges such as scalability and performance. Processing medical images from different collaborators is an example of such applications, in which chest X-ray data are processed to extract knowledge. It is not easy to process data and get the required information in real time using central processing techniques when data get very large in size. In this paper, real-time data are filtered and forwarded to the right processing node by using the proposed topic-based hierarchical publish/subscribe messaging middleware in the distributed scalable network of collaborating computation nodes instead of classical approaches of centralized computation. This enables processing streaming medical data in near real time and makes a warning system possible. End users have the capability of filtering/searching. The returned search results can be images (COVID-19 or non-COVID-19) and their meta-data are gender and age. Here, COVID-19 is detected using a novel capsule network-based model from chest X-ray images. This middleware allows for a smaller search space as well as shorter times for obtaining search results.

Agreement between low-dose and ultra-low-dose chest CT for the diagnosis of viral pneumonia imaging patterns during the COVID-19 pandemic

Background

Chest CT scan has an important role in the diagnosis and management of COVID-19 infection. A major concern in radiologic assessment of the patients is the radiation dose. Research has been done to evaluate low-dose chest CT in the diagnosis of pulmonary lesions with promising findings. We decided to determine diagnostic performance of ultra-low-dose chest CT in comparison to low-dose CT for viral pneumonia during the COVID-19 pandemic.

Results

167 patients underwent both low-dose and ultra-low-dose chest CT scans. Two radiologists blinded to the diagnosis independently examined ultra-low-dose chest CT scans for findings consistent with COVID-19 pneumonia. In case of any disagreement, a third senior radiologist made the final diagnosis. Agreement between two CT protocols regarding ground-glass opacity, consolidation, reticulation, and nodular infiltration were recorded. On low-dose chest CT, 44 patients had findings consistent with COVID-19 infection. Ultra-low-dose chest CT had sensitivity and specificity values of 100% and 98.4%, respectively for diagnosis of viral pneumonia. Two patients were falsely categorized to have pneumonia on ultra-low-dose CT scan. Positive predictive value and negative predictive value of ultra-low-dose CT scan were respectively 95.7% and 100%. There was good agreement between low-dose and ultra-low-dose methods (kappa = 0.97; P < 0.001). Perfect agreement between low-dose and ultra-low-dose scans was found regarding diagnosis of ground-glass opacity (kappa = 0.83, P < 0.001), consolidation (kappa = 0.88, P < 0.001), reticulation (kappa = 0.82, P < 0.001), and nodular infiltration (kappa = 0.87, P < 0.001).

Conclusion

Ultra-low-dose chest CT scan is comparable to low-dose chest CT for detection of lung infiltration during the COVID-19 outbreak while maintaining less radiation dose. It can also be used instead of low-dose chest CT scan for patient triage in circumstances where rapid-abundant PCR tests are not available.

Precise phase retrieval for propagation-based images using discrete mathematics

The ill-posed problem of phase retrieval in optics, using one or more intensity measurements, has a multitude of applications using electromagnetic or matter waves. Many phase retrieval algorithms are computed on pixel arrays using discrete Fourier transforms due to their high computational efficiency. However, the mathematics underpinning these algorithms is typically formulated using continuous mathematics, which can result in a loss of spatial resolution in the reconstructed images. Herein we investigate how phase retrieval algorithms for propagation-based phase-contrast X-ray imaging can be rederived using discrete mathematics and result in more precise retrieval for single- and multi-material objects and for spectral image decomposition. We validate this theory through experimental measurements of spatial resolution using computed tomography (CT) reconstructions of plastic phantoms and biological tissues, using detectors with a range of imaging system point spread functions (PSFs). We demonstrate that if the PSF substantially suppresses high spatial frequencies, the potential improvement from utilising the discrete derivation is limited. However, with detectors characterised by a single pixel PSF (e.g. direct, photon-counting X-ray detectors), a significant improvement in spatial resolution can be obtained, demonstrated here at up to 17%.

Chest CT scan for the screening of air anomalies at risk of pulmonary barotrauma for the initial medical assessment of fitness to dive in a military population

Introduction: The presence of intra-pulmonary air lesions such as cysts, blebs and emphysema bullae, predisposes to pulmonary barotrauma during pressure variations, especially during underwater diving activities. These rare accidents can have dramatic consequences. Chest radiography has long been the baseline examination for the detection of respiratory pathologies in occupational medicine. It has been replaced since 2018 by the thoracic CT scan for military diving fitness in France. The objective of this work was to evaluate the prevalence of the pulmonary abnormalities of the thoracic CT scan, and to relate them to the characteristics of this population and the results of the spirometry.

Methods: 330 records of military diving candidates who underwent an initial assessment between October 2018 and March 2021 were analyzed, in a single-center retrospective analysis. The following data were collected: sex, age, BMI, history of respiratory pathologies and smoking, treatments, allergies, diving practice, results of spirometry, reports of thoracic CT scans, as well as fitness decision.

Results: The study included 307 candidates, mostly male, with a median age of 25 years. 19% of the subjects had abnormal spirometry. We identified 25% of divers with CT scan abnormalities. 76% of the abnormal scans were benign nodules, 26% of which measured 6 mm or more. Abnormalities with an aerial component accounted for 13% of the abnormal scans with six emphysema bullae, three bronchial dilatations and one cystic lesion. No association was found between the presence of nodules and the general characteristics of the population, whereas in six subjects emphysema bullae were found statistically associated with active smoking or abnormal spirometry results.

Conclusion: The systematic performance of thoracic CT scan in a young population free of pulmonary pathology revealed a majority of benign nodules. Abnormalities with an aerial component are much less frequent, but their presence generally leads to a decision of unfitness. These results argue in favor of a systematic screening of aeric pleuro-pulmonary lesions during the initial assessment for professional divers.

Basis and current state of computed tomography perfusion imaging: a review

Computed tomography perfusion (CTP) is a functional imaging that allows for providing capillary-level hemodynamics information of the desired tissue in clinics. In this paper, we aim to offer insight into CTP imaging which covers the basics and current state of CTP imaging, then summarize the technical applications in the CTP imaging as well as the future technological potential. At first, we focus on the fundamentals of CTP imaging including systematically summarized CTP image acquisition and hemodynamic parameter map estimation techniques. A short assessment is presented to outline the clinical applications with CTP imaging, and then a review of radiation dose effect of the CTP imaging on the different applications is presented. We present a categorized methodology review on known and potential solvable challenges of radiation dose reduction in CTP imaging. To evaluate the quality of CTP images, we list various standardized performance metrics. Moreover, we present a review on the determination of infarct and penumbra. Finally, we reveal the popularity and future trend of CTP imaging.

Deep learning-based approach for detecting COVID-19 in chest X-rays

Today, 2019 Coronavirus (COVID-19) infections are a major health concern worldwide. Therefore, detecting COVID-19 in X-ray images is crucial for diagnosis, evaluation, and treatment. Furthermore, expressing diagnostic uncertainty in a report is a challenging duty but unavoidable task for radiologists. This study proposes a novel CNN (Convolutional Neural Network) model for automatic COVID-19 identification utilizing chest X-ray images. The proposed CNN model is designed to be a reliable diagnostic tool for two-class categorization (COVID and Normal). In addition to the proposed model, different architectures, including the pre-trained MobileNetv2 and ResNet50 models, are evaluated for this COVID-19 dataset (13,824 X-ray images) and our suggested model is compared to these existing COVID-19 detection algorithms in terms of accuracy. Experimental results show that our proposed model identifies patients with COVID-19 disease with 96.71 percent accuracy, 91.89 percent F1-score. Our proposed approach CNN’s experimental results show that it outperforms the most advanced algorithms currently available. This model can assist clinicians in making informed judgments on how to diagnose COVID-19, as well as make test kits more accessible.

Evaluation of Ultra-Low-Dose Chest Computed Tomography Images in Detecting Lung Lesions Related to COVID-19: A Prospective Study

Background

The present study aimed to evaluate the effectiveness of ultra-low-dose (ULD) chest computed tomography (CT) in comparison with the routine dose (RD) CT images in detecting lung lesions related to COVID-19.

Methods

A prospective study was conducted during April-September 2020 at Shahid Faghihi Hospital affiliated with Shiraz University of Medical Sciences, Shiraz, Iran. In total, 273 volunteers with suspected COVID-19 participated in the study and successively underwent RD-CT and ULD-CT chest scans. Two expert radiologists qualitatively evaluated the images. Dose assessment was performed by determining volume CT dose index, dose length product, and size-specific dose estimate. Data analysis was performed using a ranking test and kappa coefficient (κ). P<0.05 was considered statistically significant.

Results

Lung lesions could be detected with both RD-CT and ULD-CT images in patients with suspected or confirmed COVID-19 (κ=1.0, P=0.016). The estimated effective dose for the RD-CT protocol was 22-fold higher than in the ULD-CT protocol. In the case of the ULD-CT protocol, sensitivity, specificity, accuracy, and positive predictive value for the detection of consolidation were 60%, 83%, 80%, and 20%, respectively. Comparably, in the case of RD-CT, these percentages for the detection of ground-glass opacity (GGO) were 62%, 66%, 66%, and 18%, respectively. Assuming the result of real-time polymerase chain reaction as true-positive, analysis of the receiver-operating characteristic curve for GGO detected using the ULD-CT protocol showed a maximum area under the curve of 0.78.

Conclusion

ULD-CT, with 94% dose reduction, can be an alternative to RD-CT to detect lung lesions for COVID-19 diagnosis and follow-up.An earlier preliminary report of a similar work with a lower sample size was submitted to the arXive as a preprint. The preprint is cited as: https://arxiv.org/abs/2005.03347.

Ultra-low-dose CT versus chest X-ray for patients suspected of pulmonary disease at the emergency department: a multicentre randomised clinical trial

BACKGROUND

Chest CT displays chest pathology better than chest X-ray (CXR). We evaluated the effects on health outcomes of replacing CXR by ultra-low-dose chest-CT (ULDCT) in the diagnostic work-up of patients suspected of non-traumatic pulmonary disease at the emergency department.

METHODS

Pragmatic, multicentre, non-inferiority randomised clinical trial in patients suspected of non-traumatic pulmonary disease at the emergency department. Between 31 January 2017 and 31 May 2018, every month, participating centres were randomly allocated to using ULDCT or CXR. Primary outcome was functional health at 28 days, measured by the Short Form (SF)-12 physical component summary scale score (PCS score), non-inferiority margin was set at 1 point. Secondary outcomes included hospital admission, hospital length of stay (LOS) and patients in follow-up because of incidental findings.

RESULTS

2418 consecutive patients (ULDCT: 1208 and CXR: 1210) were included. Mean SF-12 PCS score at 28 days was 37.0 for ULDCT and 35.9 for CXR (difference 1.1; 95% lower CI: 0.003). After ULDCT, 638/1208 (52.7%) patients were admitted (median LOS of 4.8 days; IQR 2.1-8.8) compared with 659/1210 (54.5%) patients after CXR (median LOS 4.6 days; IQR 2.1-8.8). More ULDCT patients were in follow-up because of incidental findings: 26 (2.2%) versus 4 (0.3%).

CONCLUSIONS

Short-term functional health was comparable between ULDCT and CXR, as were hospital admissions and LOS, but more incidental findings were found in the ULDCT group. Our trial does not support routine use of ULDCT in the work-up of patients suspected of non-traumatic pulmonary disease at the emergency department.

TRIAL REGISTRATION NUMBER

NTR6163.

Diagnostic validation of a deep learning nodule detection algorithm in low-dose chest CT: determination of optimized dose thresholds in a virtual screening scenario

OBJECTIVES

This study was conducted to evaluate the effect of dose reduction on the performance of a deep learning (DL)-based computer-aided diagnosis (CAD) system regarding pulmonary nodule detection in a virtual screening scenario.

METHODS

Sixty-eight anthropomorphic chest phantoms were equipped with 329 nodules (150 ground glass, 179 solid) with four sizes (5 mm, 8 mm, 10 mm, 12 mm) and scanned with nine tube voltage/current combinations. The examinations were analyzed by a commercially available DL-based CAD system. The results were compared by a comparison of proportions. Logistic regression was performed to evaluate the impact of tube voltage, tube current, nodule size, nodule density, and nodule location.

RESULTS

The combination with the lowest effective dose (E) and unimpaired detection rate was 80 kV/50 mAs (sensitivity: 97.9%, mean false-positive rate (FPR): 1.9, mean CTDIvol: 1.2 ± 0.4 mGy, mean E: 0.66 mSv). Logistic regression revealed that tube voltage and current had the greatest impact on the detection rate, while nodule size and density had no significant influence.

CONCLUSIONS

The optimal tube voltage/current combination proposed in this study (80 kV/50 mAs) is comparable to the proposed combinations in similar studies, which mostly dealt with conventional CAD software. Modification of tube voltage and tube current has a significant impact on the performance of DL-based CAD software in pulmonary nodule detection regardless of their size and composition.

KEY POINTS

• Modification of tube voltage and tube current has a significant impact on the performance of deep learning-based CAD software. • Nodule size and composition have no significant impact on the software's performance. • The optimal tube voltage/current combination for the examined software is 80 kV/50 mAs.

Four Types of Multiclass Frameworks for Pneumonia Classification and Its Validation in X-ray Scans Using Seven Types of Deep Learning Artificial Intelligence Models

Background and Motivation: The novel coronavirus causing COVID-19 is exceptionally contagious, highly mutative, decimating human health and life, as well as the global economy, by consistent evolution of new pernicious variants and outbreaks. The reverse transcriptase polymerase chain reaction currently used for diagnosis has major limitations. Furthermore, the multiclass lung classification X-ray systems having viral, bacterial, and tubercular classes—including COVID-19—are not reliable. Thus, there is a need for a robust, fast, cost-effective, and easily available diagnostic method. Method: Artificial intelligence (AI) has been shown to revolutionize all walks of life, particularly medical imaging. This study proposes a deep learning AI-based automatic multiclass detection and classification of pneumonia from chest X-ray images that are readily available and highly cost-effective. The study has designed and applied seven highly efficient pre-trained convolutional neural networks—namely, VGG16, VGG19, DenseNet201, Xception, InceptionV3, NasnetMobile, and ResNet152—for classification of up to five classes of pneumonia. Results: The database consisted of 18,603 scans with two, three, and five classes. The best results were using DenseNet201, VGG16, and VGG16, respectively having accuracies of 99.84%, 96.7%, 92.67%; sensitivity of 99.84%, 96.63%, 92.70%; specificity of 99.84, 96.63%, 92.41%; and AUC of 1.0, 0.97, 0.92 (p < 0.0001 for all), respectively. Our system outperformed existing methods by 1.2% for the five-class model. The online system takes <1 s while demonstrating reliability and stability. Conclusions: Deep learning AI is a powerful paradigm for multiclass pneumonia classification.

Personalized Chest Computed Tomography: Minimum Diagnostic Radiation Dose Levels for the Detection of Fibrosis, Nodules, and Pneumonia

OBJECTIVES

The purpose of this study was to evaluate the minimum diagnostic radiation dose level for the detection of high-resolution (HR) lung structures, pulmonary nodules (PNs), and infectious diseases (IDs).

MATERIALS AND METHODS

A preclinical chest computed tomography (CT) trial was performed with a human cadaver without known lung disease with incremental radiation dose using tin filter-based spectral shaping protocols. A subset of protocols for full diagnostic evaluation of HR, PN, and ID structures was translated to clinical routine. Also, a minimum diagnostic radiation dose protocol was defined (MIN). These protocols were prospectively applied over 5 months in the clinical routine under consideration of the individual clinical indication. We compared radiation dose parameters, objective and subjective image quality (IQ).

RESULTS

The HR protocol was performed in 38 patients (43%), PN in 21 patients (24%), ID in 20 patients (23%), and MIN in 9 patients (10%). Radiation dose differed significantly among HR, PN, and ID (5.4, 1.2, and 0.6 mGy, respectively; P < 0.001). Differences between ID and MIN (0.2 mGy) were not significant (P = 0.262). Dose-normalized contrast-to-noise ratio was comparable among all groups (P = 0.087). Overall IQ was perfect for the HR protocol (median, 5.0) and decreased for PN (4.5), ID-CT (4.3), and MIN-CT (2.5). The delineation of disease-specific findings was high in all dedicated protocols (HR, 5.0; PN, 5.0; ID, 4.5). The MIN protocol had borderline IQ for PN and ID lesions but was insufficient for HR structures. The dose reductions were 78% (PN), 89% (ID), and 97% (MIN) compared with the HR protocols.

CONCLUSIONS

Personalized chest CT tailored to the clinical indications leads to substantial dose reduction without reducing interpretability. More than 50% of patients can benefit from such individual adaptation in a clinical routine setting. Personalized radiation dose adjustments with validated diagnostic IQ are especially preferable for evaluating ID and PN lesions.

Local binary pattern and deep learning feature extraction fusion for COVID-19 detection on computed tomography images

The deadly coronavirus virus (COVID-19) was confirmed as a pandemic by the World Health Organization (WHO) in December 2019. It is important to identify suspected patients as early as possible in order to control the spread of the virus, improve the efficacy of medical treatment, and, as a result, lower the mortality rate. The adopted method of detecting COVID-19 is the reverse-transcription polymerase chain reaction (RT-PCR), the process is affected by a scarcity of RT-PCR kits as well as its complexities. Medical imaging using machine learning and deep learning has proved to be one of the most efficient methods of detecting respiratory diseases, but to train machine learning features needs to be extracted manually, and in deep learning, efficiency is affected by deep learning architecture and low data. In this study, handcrafted local binary pattern (LBP) and automatic seven deep learning models extracted features were used to train support vector machines (SVM) and K-nearest neighbour (KNN) classifiers, to improve the performance of the classifier, a concatenated LBP and deep learning feature was proposed to train the KNN and SVM, based on the performance criteria, the models VGG-19 + LBP achieved the highest accuracy of 99.4%. The SVM and KNN classifiers trained on the hybrid feature outperform the state of the art model. This shows that the proposed feature can improve the performance of the classifiers in detecting COVID-19.

Automated detection of COVID-19 cases from chest X-ray images using deep neural network and XGBoost

Introduction

In late 2019 and after the COVID-19 pandemic in the world, many researchers and scholars tried to provide methods for detecting COVID-19 cases. Accordingly, this study focused on identifying patients with COVID-19 from chest X-ray images.

Methods

In this paper, a method for diagnosing coronavirus disease from X-ray images was developed. In this method, DenseNet169 Deep Neural Network (DNN) was used to extract the features of X-ray images taken from the patients’ chests. The extracted features were then given as input to the Extreme Gradient Boosting (XGBoost) algorithm to perform the classification task.

Results

Evaluation of the proposed approach and its comparison with the methods presented in recent years revealed that this method was more accurate and faster than the existing ones and had an acceptable performance for detecting COVID-19 cases from X-ray images. The experiments showed 98.23% and 89.70% accuracy, 99.78% and 100% specificity, 92.08% and 95.20% sensitivity in two and three-class problems, respectively.

Conclusion

This study aimed to detect people with COVID-19, focusing on non-clinical approaches. The developed method could be employed as an initial detection tool to assist the radiologists in more accurate and faster diagnosing the disease.

Implication for practice

The proposed method's simple implementation, along with its acceptable accuracy, allows it to be used in COVID-19 diagnosis. Moreover, the gradient-based class activation mapping (Grad-CAM) can be used to represent the deep neural network's decision area on a heatmap. Radiologists might use this heatmap to evaluate the chest area more accurately.

Diagnostic accuracy of ultra-low-dose chest computed tomography in an emergency department

BACKGROUND

This study examined whether ultra-low-dose chest computed tomography (ULD-CT) could improve detection of acute chest conditions.

PURPOSE

To determine (i) whether diagnostic accuracy of ULD-CT is superior to supine chest X-ray (sCXR) for acute chest conditions and (ii) the feasibility of ULD-CT in an emergency department.

MATERIAL AND METHODS

From 1 February to 31 July 2019, 91 non-traumatic patients from the Emergency Department were prospectively enrolled in the study if they received an sCXR. An ULD-CT and a non-contrast chest CT (NCCT) scan were then performed. Three radiologists assessed the sCXR and ULD-CT examinations for cardiogenic pulmonary edema, pneumonia, pneumothorax, and pleural effusion. Resources and effort were compared for sCXR and ULD-CT to evaluate feasibility. Diagnostic accuracy was calculated for sCXR and ULD-CT using NCCT as the reference standard.

RESULTS

The mean effective dose of ULD-CT was 0.05±0.01 mSv. For pleural effusion and cardiogenic pulmonary edema, no difference in diagnostic accuracy between ULD-CT and sCXR was observed. For pneumonia and pneumothorax, sensitivities were 100% (95% confidence interval [CI] 69-100) and 50% (95% CI 7-93) for ULD-CT and 60% (95% CI 26-88) and 0% (95% CI 0-0) for sCXR, respectively. Median examination time was 10 min for ULD-CT vs. 5 min for sCXR (P<0.001). For ULD-CT 1-2 more staff members were needed compared to sCXR (P<0.001). ULD-CT was rated more challenging to perform than sCXR (P<0.001).

CONCLUSION

ULD-CT seems equal or better in detecting acute chest conditions compared to sCXR. However, ULD-CT examinations demand more effort and resources.

COVID-19 Pneumonia Classification Based on NeuroWavelet Capsule Network

Since it was first reported, coronavirus disease 2019, also known as COVID-19, has spread expeditiously around the globe. COVID-19 must be diagnosed as soon as possible in order to control the disease and provide proper care to patients. The chest X-ray (CXR) has been identified as a useful diagnostic tool, but the disease outbreak has put a lot of pressure on radiologists to read the scans, which could give rise to fatigue-related misdiagnosis. Automatic classification algorithms that are reliable can be extremely beneficial; however, they typically depend upon a large amount of COVID-19 data for training, which are troublesome to obtain in the nick of time. Therefore, we propose a novel method for the classification of COVID-19. Concretely, a novel neurowavelet capsule network is proposed for COVID-19 classification. To be more precise, first, we introduce a multi-resolution analysis of a discrete wavelet transform to filter noisy and inconsistent information from the CXR data in order to improve the feature extraction robustness of the network. Secondly, the discrete wavelet transform of the multi-resolution analysis also performs a sub-sampling operation in order to minimize the loss of spatial details, thereby enhancing the overall classification performance. We examined the proposed model on a public-sourced dataset of pneumonia-related illnesses, including COVID-19 confirmed cases and healthy CXR images. The proposed method achieves an accuracy of 99.6%, sensitivity of 99.2%, specificity of 99.1% and precision of 99.7%. Our approach achieves an up-to-date performance that is useful for COVID-19 screening according to the experimental results. This latest paradigm will contribute significantly in the battle against COVID-19 and other diseases.

Enabling CT-Scans for covid detection using transfer learning-based neural networks

Today, we are coping with the pandemic, and the novel virus is covertly evolving day by day. Therefore, a precautionary system to deal with the issue is required as early as possible. The last few years were very challenging for doctors, vaccine makers, hospitals, and medical authorities to deal with the massive crowd to provide results for all patients and newcomers in the past months. Thus, these issues should be handled with a robust system that can accord with many people and deliver the results in a fraction of time without visiting public places and help reduce crowd gathering. So, to deal with these issues, we developed an AI model using transfer learning that can aid doctors and other people to get to know whether they were suffering from covid or not. In this paper, we have used VGG-19 (CNN-based) model with open-sourced COVID-CT (CTSI) dataset. The dataset consists of 349 images of COVID-19 of 216 patients and 463 images of NON-COVID-19. We have achieved an accuracy of 95%, precision of 96%, recall of 94%, and F1-Score of 96% from the experiments.Communicated by Ramaswamy H. Sarma.

Deep Learning Reconstruction Shows Better Lung Nodule Detection for Ultra-Low-Dose Chest CT

Background Ultra-low-dose (ULD) CT could facilitate the clinical implementation of large-scale lung cancer screening while minimizing the radiation dose. However, traditional image reconstruction methods are associated with image noise in low-dose acquisitions. Purpose To compare the image quality and lung nodule detectability of deep learning image reconstruction (DLIR) and adaptive statistical iterative reconstruction-V (ASIR-V) in ULD CT. Materials and Methods Patients who underwent noncontrast ULD CT (performed at 0.07 or 0.14 mSv, similar to a single chest radiograph) and contrast-enhanced chest CT (CECT) from April to June 2020 were included in this prospective study. ULD CT images were reconstructed with filtered back projection (FBP), ASIR-V, and DLIR. Three-dimensional segmentation of lung tissue was performed to evaluate image noise. Radiologists detected and measured nodules with use of a deep learning-based nodule assessment system and recognized malignancy-related imaging features. Bland-Altman analysis and repeated-measures analysis of variance were used to evaluate the differences between ULD CT images and CECT images. Results A total of 203 participants (mean age ± standard deviation, 61 years ± 12; 129 men) with 1066 nodules were included, with 100 scans at 0.07 mSv and 103 scans at 0.14 mSv. The mean lung tissue noise ± standard deviation was 46 HU ± 4 for CECT and 59 HU ± 4, 56 HU ± 4, 53 HU ± 4, 54 HU ± 4, and 51 HU ± 4 in FBP, ASIR-V level 40%, ASIR-V level 80% (ASIR-V-80%), medium-strength DLIR, and high-strength DLIR (DLIR-H), respectively, of ULD CT scans (P < .001). The nodule detection rates of FBP reconstruction, ASIR-V-80%, and DLIR-H were 62.5% (666 of 1066 nodules), 73.3% (781 of 1066 nodules), and 75.8% (808 of 1066 nodules), respectively (P < .001). Bland-Altman analysis showed the percentage difference in long diameter from that of CECT was 9.3% (95% CI of the mean: 8.0, 10.6), 9.2% (95% CI of the mean: 8.0, 10.4), and 6.2% (95% CI of the mean: 5.0, 7.4) in FBP reconstruction, ASIR-V-80%, and DLIR-H, respectively (P < .001). Conclusion Compared with adaptive statistical iterative reconstruction-V, deep learning image reconstruction reduced image noise, increased nodule detection rate, and improved measurement accuracy on ultra-low-dose chest CT images. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Lee in this issue.

Qualitative and Quantitative Assessment of Emphysema Using Dark-Field Chest Radiography

Background Dark-field chest radiography allows for assessment of lung alveolar structure by exploiting wave optical properties of x-rays. Purpose To evaluate the qualitative and quantitative features of dark-field chest radiography in participants with pulmonary emphysema as compared with those in healthy control subjects. Materials and Methods In this prospective study conducted from October 2018 to October 2020, participants aged at least 18 years who underwent clinically indicated chest CT were screened for participation. Inclusion criteria were an ability to consent to the procedure and stand upright without help. Exclusion criteria were pregnancy, serious medical conditions, and any lung condition besides emphysema that was visible on CT images. Participants were examined with a clinical dark-field chest radiography prototype that simultaneously acquired both attenuation-based radiographs and dark-field chest radiographs. Dark-field coefficients were tested for correlation with each participant's CT-based emphysema index using the Spearman correlation test. Dark-field coefficients of adjacent groups in the semiquantitative Fleischner Society emphysema grading system were compared using a Wilcoxon Mann-Whitney U test. The capability of the dark-field coefficient to enable detection of emphysema was evaluated with receiver operating characteristics curve analysis. Results A total of 83 participants (mean age, 65 years ± 12 [standard deviation]; 52 men) were studied. When compared with images from healthy participants, dark-field chest radiographs in participants with emphysema had a lower and inhomogeneous dark-field signal intensity. The locations of focal signal intensity loss on dark-field images corresponded well with emphysematous areas found on CT images. The dark-field coefficient was negatively correlated with the quantitative CT-based emphysema index (r = -0.54, P < .001). Participants with Fleischner Society grades of mild, moderate, confluent, or advanced destructive emphysema exhibited a lower dark-field coefficient than those without emphysema (eg, 1.3 m^-1 ± 0.6 for participants with confluent or advanced destructive emphysema vs 2.6 m^-1 ± 0.4 for participants without emphysema; P < .001). The area under the receiver operating characteristic curve for detection of mild emphysema was 0.79. Conclusion Pulmonary emphysema leads to reduced signal intensity on dark-field chest radiographs, showing the technique has potential as a diagnostic tool in the assessment of lung diseases. © RSNA, 2022 See also the editorial by Hatabu and Madore in this issue.

Automatic method for classifying COVID-19 patients based on chest X-ray images, using deep features and PSO-optimized XGBoost

The COVID-19 pandemic, which originated in December 2019 in the city of Wuhan, China, continues to have a devastating effect on the health and well-being of the global population. Currently, approximately 8.8 million people have already been infected and more than 465,740 people have died worldwide. An important step in combating COVID-19 is the screening of infected patients using chest X-ray (CXR) images. However, this task is extremely time-consuming and prone to variability among specialists owing to its heterogeneity. Therefore, the present study aims to assist specialists in identifying COVID-19 patients from their chest radiographs, using automated computational techniques. The proposed method has four main steps: (1) the acquisition of the dataset, from two public databases; (2) the standardization of images through preprocessing; (3) the extraction of features using a deep features-based approach implemented through the networks VGG19, Inception-v3, and ResNet50; (4) the classifying of images into COVID-19 groups, using eXtreme Gradient Boosting (XGBoost) optimized by particle swarm optimization (PSO). In the best-case scenario, the proposed method achieved an accuracy of 98.71%, a precision of 98.89%, a recall of 99.63%, and an F1-score of 99.25%. In our study, we demonstrated that the problem of classifying CXR images of patients under COVID-19 and non-COVID-19 conditions can be solved efficiently by combining a deep features-based approach with a robust classifier (XGBoost) optimized by an evolutionary algorithm (PSO). The proposed method offers considerable advantages for clinicians seeking to tackle the current COVID-19 pandemic.

COMPARISON OF PULMONARY NODULE DETECTION, READING TIMES AND PATIENT DOSES OF ULTRA-LOW DOSE CT, STANDARD DOSE CT AND DIGITAL RADIOGRAPHY

The purpose of the present work was to evaluate performance in pulmonary nodule detection, reading times and patient doses for ultra-low dose computed tomography (ULD-CT), standard dose chest CT (SD-CT), and digital radiography (DR). Pulmonary nodules were simulated in an anthropomorphic lung phantom. Thirty cases, 18 with lesions (45 total lesions of 3-12 mm) and 12 without lesions were acquired for each imaging modality. Three radiologists interpreted the cases in a free-response study. Performance was assessed using the JAFROC figure-of-merit (FOM). Performance was not significantly different between ULD-CT and SD-CT (FOMs: 0.787 vs 0.814; ΔFOM: 0.03), but both CT techniques were superior to DR (FOM: 0.541; ΔFOM: 0.31 and 0.28). Overall, the CT modalities took longer time to interpret than DR. ULD chest CT may serve as an alternative to both SD-CT and conventional radiography, considerably reducing dose in the first case and improving diagnostic accuracy in the second.

Evaluation and comparison of performance of low-dose 128-slice CT scanner with different mAs values: A phantom study

OBJECTIVE:

Radiation dose in computed tomography (CT) has been the concern of physicists ever since the introduction of CT scan. The objective of this study was to evaluate the performance of low-dose 128-slice CT scanner with different mAs values.

MATERIALS AND METHODS:

Quantitative study was carried out at different values of mAs. Philips brilliance CT phantom with Philips ingenuity 128-slice low-dose CT scanner was chosen for this study. CT number linearity, CT number accuracy, slice thickness accuracy, high-contrast resolution, and low-contrast resolution were calculated and estimated computed tomography dose index volume (CTDI_vol) for all the mAs values were recorded. Noise was calculated for all mAs values for comparison.

RESULTS:

Data analysis shows that image quality was acceptable for all protocols. High-contrast resolution for all protocols was 20 line pairs per centimeter. Low-contrast resolution for 50 mAs images was 4 mm and 3 mm for other mAs protocols. Images acquired using 100 mAs revealed ring artifacts. CTDI_vol using 50 mAs was 33% of the CTDI_vol using 150 mAs. The dose–length product at 100 mAs was reduced to 66% of the dose–length product at 150 mAs, and the same at 50 mAs was reduced to 33%.

CONCLUSION:

It is evident here that mAs has direct impact on the radiation dose to patient. With iDose4, mAs can be reduced to 50 mAs in multislice low-dose CT scan to reduce the radiation dose with minimal effect on image quality for slice thickness 4 mm. However, noise would dominate at tube current lower than 50 mAs for 120 kVp.

Lung segmentation and automatic detection of COVID-19 using radiomic features from chest CT images

This paper aims to develop an automatic method to segment pulmonary parenchyma in chest CT images and analyze texture features from the segmented pulmonary parenchyma regions to assist radiologists in COVID-19 diagnosis. A new segmentation method, which integrates a three-dimensional (3D) V-Net with a shape deformation module implemented using a spatial transform network (STN), was proposed to segment pulmonary parenchyma in chest CT images. The 3D V-Net was adopted to perform an end-to-end lung extraction while the deformation module was utilized to refine the V-Net output according to the prior shape knowledge. The proposed segmentation method was validated against the manual annotation generated by experienced operators. The radiomic features measured from our segmentation results were further analyzed by sophisticated statistical models with high interpretability to discover significant independent features and detect COVID-19 infection. Experimental results demonstrated that compared with the manual annotation, the proposed segmentation method achieved a Dice similarity coefficient of 0.9796, a sensitivity of 0.9840, a specificity of 0.9954, and a mean surface distance error of 0.0318 mm. Furthermore, our COVID-19 classification model achieved an area under curve (AUC) of 0.9470, a sensitivity of 0.9670, and a specificity of 0.9270 when discriminating lung infection with COVID-19 from community-acquired pneumonia and healthy controls using statistically significant radiomic features. The significant features measured from our segmentation results agreed well with those from the manual annotation. Our approach has great promise for clinical use in facilitating automatic diagnosis of COVID-19 infection on chest CT images.

A segmentation tool for pulmonary nodules in lung cancer screening: Testing and clinical usage

PURPOSE

With the future goal of defining a large dataset based on low-dose CT with labelled pulmonary lesions for lung cancer screening (LCS) research, the aim of this work is to propose and evaluate into a clinical context a tool for semi-automatic segmentation able to facilitate the process of labels collection from a LCS study (COSMOS, Continuous Observation of SMOking Subjects).

METHODS

Considering a preliminary set of manual annotations, a segmentation model based on a 2D-Unet was trained from scratch. Contour quality of the final 2D-Unet was assessed on an internal test set of manual annotations and on a subset of the public available LIDC dataset used as external test set. The tool for semi-automatic segmentation was then designed integrating the tested model into a Graphical User Interface. According to the opinion of two clinical users, the percentage of lesions properly contoured through the tool was quantified (Acceptance Rate, AR). The variability between segmentations derived by the two readers was estimated as mean percentage of difference (MPD) between the two sets of volumes and comparing the likelihood of malignancy derived from Volume Doubling Time (VDT).

RESULTS

Performance in test sets were found similar (DICE ~ 0.75(0.15)). Accordingly, a good mean AR (80.1%) resulted from the two readers. Variability in terms of MPD was equal to 23.6% while 2.7% was the VDTs percentage of disagreement.

CONCLUSIONS

A semi-automatic segmentation tool was developed and its applicability evaluated into a clinical context demonstrating the efficacy of the tool in facilitating the collection of labelled data.

Pneumonia Detection in Chest X-Ray Dose-Equivalent CT: Impact of Dose Reduction on Detectability by Artificial Intelligence

RATIONALE AND OBJECTIVES

There has been a significant increase of immunocompromised patients in recent years due to new treatment modalities for previously fatal diseases. This comes at the cost of an elevated risk for infectious diseases, most notably pathogens affecting the respiratory tract. Because early diagnosis and treatment of pneumonia can help reducing morbidity and mortality, we assessed the performance of a deep neural network in the detection of pulmonary infection in chest X-ray dose-equivalent computed tomography (CT).

MATERIALS AND METHODS

The 100 patients included in this retrospective study were referred to our department for suspicion of pulmonary infection and/or follow-up of known pulmonary nodules. Every patient was scanned with a standard dose (1.43 ± 0.54 mSv) and a 20 times dose-reduced (0.07 ± 0.03 mSv) CT protocol. We trained a deep neural network to perform binary classification (pulmonary consolidation present or not) and assessed diagnostic performance on both standard dose and reduced dose CT images.

RESULTS

The areas under the curve of the deep learning algorithm for the standard dose CT was 0.923 (confidence interval [CI] 95%: 0.905-0.941) and significantly higher than the areas under the curve (0.881, CI 95%: 0.859-0.903) of the reduced dose CT (p = 0.001). Sensitivity and specificity of the standard dose CT was 82.9% and 93.8%, and of the reduced dose CT 71.0% and 93.3%.

CONCLUSION

Pneumonia detection with X-ray dose-equivalent CT using artificial intelligence is feasible and may contribute to a more robust and reproducible diagnostic performance. Dose reduction lowered the performance of the deep neural network, which calls for optimization and adaption of CT protocols when using AI algorithms at reduced doses.

The utility of a dual-phase, dual-energy CT protocol in patients presenting with overt gastrointestinal bleeding

Background

Due to their easy accessibility, CT scans have been increasingly used for investigation of gastrointestinal (GI) bleeding.

Purpose

To estimate the performance of a dual-phase, dual-energy (DE) GI bleed CT protocol in patients with overt GI bleeding in clinical practice and examine the added value of portal phase and DE images.

Materials and Methods

Consecutive patients with GI bleeding underwent a two-phase DE GI bleed CT protocol. Two gastroenterologists established the reference standard. Performance was estimated using clinical CT reports. Three GI radiologists rated confidence in GI bleeding in a subset of 62 examinations, evaluating first mixed kV arterial images, then after examining additional portal venous phase images, and finally after additional DE images (virtual non-contrast and virtual monoenergetic 50 keV images).

Results

52 of 176 patients (29.5%) had GI bleeding by the reference standard. The overall sensitivity, specificity, and positive and negative predictive values of the CT GI bleed protocol for detecting GI bleeding were 65.4%, 89.5%, 72.3%, and 86.0%, respectively. In patients with GI bleeding, diagnostic confidence of readers increased after adding portal phase images to arterial phase images (p = 0.002), without additional benefit from dual energy images. In patients without GI bleeding, confidence in luminal extravasation appropriately decreased after adding portal phase, and subsequently DE images (p = 0.006, p = 0.018).

Conclusion

A two-phase DE GI bleed CT protocol had high specificity and negative predictive value in clinical practice. Portal venous phase images improved diagnostic confidence in comparison to arterial phase images alone. Dual-energy images further improved radiologist confidence in the absence of bleeding.

The use of deep learning towards dose optimization in low-dose computed tomography: A scoping review

INTRODUCTION

Low-dose computed tomography tends to produce lower image quality than normal dose computed tomography (CT) although it can help to reduce radiation hazards of CT scanning. Research has shown that Artificial Intelligence (AI) technologies, especially deep learning can help enhance the image quality of low-dose CT by denoising images. This scoping review aims to create an overview on how AI technologies, especially deep learning, can be used in dose optimisation for low-dose CT.

METHODS

Literature searches of ProQuest, PubMed, Cinahl, ScienceDirect, EbscoHost Ebook Collection and Ovid were carried out to find research articles published between the years 2015 and 2020. In addition, manual search was conducted in SweMed+, SwePub, NORA, Taylor & Francis Online and Medic.

RESULTS

Following a systematic search process, the review comprised of 16 articles. Articles were organised according to the effects of the deep learning networks, e.g. image noise reduction, image restoration. Deep learning can be used in multiple ways to facilitate dose optimisation in low-dose CT. Most articles discuss image noise reduction in low-dose CT.

CONCLUSION

Deep learning can be used in the optimisation of patients' radiation dose. Nevertheless, the image quality is normally lower in low-dose CT (LDCT) than in regular-dose CT scans because of smaller radiation doses. With the help of deep learning, the image quality can be improved to equate the regular-dose computed tomography image quality.

IMPLICATIONS TO PRACTICE

Lower dose may decrease patients' radiation risk but may affect the image quality of CT scans. Artificial intelligence technologies can be used to improve image quality in low-dose CT scans. Radiologists and radiographers should have proper education and knowledge about the techniques used.

Diagnostic yield, safety, and advantages of ultra-low dose chest CT compared to chest radiography in early stage suspected SARS-CoV-2 pneumonia: A retrospective observational study

ABSTRACT

To determine the role of ultra-low dose chest computed tomography (uld CT) compared to chest radiographs in patients with laboratory-confirmed early stage SARS-CoV-2 pneumonia.Chest radiographs and uld CT of 12 consecutive suspected SARS-CoV-2 patients performed up to 48 hours from hospital admission were reviewed by 2 radiologists. Dosimetry and descriptive statistics of both modalities were analyzed.On uld CT, parenchymal abnormalities compatible with SARS-CoV-2 pneumonia were detected in 10/12 (83%) patients whereas on chest X-ray in, respectively, 8/12 (66%) and 5/12 (41%) patients for reader 1 and 2. The average increment of diagnostic performance of uld CT compared to chest X-ray was 29%. The average effective dose was, respectively, of 0.219 and 0.073 mSv.Uld CT detects substantially more lung injuries in symptomatic patients with suspected early stage SARS-CoV-2 pneumonia compared to chest radiographs, with a significantly better inter-reader agreement, at the cost of a slightly higher equivalent radiation dose.