Liver fat quantification in photon counting CT in head to head comparison with clinical MRI - First experience.

PURPOSE: To compare liver fat quantification between MRI and photon-counting CT (PCCT). METHOD: A cylindrical phantom with inserts containing six concentrations of oil (0, 10, 20, 30, 50 and 100%) and oil-iodine mixtures (0, 10, 20, 30 and 50% fat +3 mg/mL iodine) was imaged with a PCCT (NAEOTOM Alpha) and a 1.5 T MRI system (MR 450w, IDEAL-IQ sequence), using clinical parameters. An IRB-approved prospective clinical evaluation included 12 obese adult patients with known fatty liver disease (seven women, mean age: 61.5 ± 13 years, mean BMI: 30.3 ± 4.7 kg/m2). Patients underwent a same-day clinical MRI and PCCT of the abdomen. Liver fat fractions were calculated for four segments (I, II, IVa and VII) using in- and opposed-phase on MRI ((Meanin - Meanopp)/2*Meanin) and iodine-fat, tissue decomposition analysis in PCCT (Syngo.Via VB60A). CT and MRI Fat fractions were compared using two-sample t-tests with equal variance. Statistical analysis was performed using RStudio (Version1.4.1717). RESULTS: Phantom results showed no significant differences between the known fat fractions (P = 0.32) or iodine (P = 0.6) in comparison to PCCT-measured concentrations, and no statistically significant difference between known and MRI-measured fat fractions (P = 0.363). In patients, the mean fat signal fraction measured on MRI and PCCT was 13.1 ± 9.9% and 12.0 ± 9.0%, respectively, with an average difference of 1.1 ± 1.9% between the modalities (P = 0.138). CONCLUSION: First experience shows promising accuracy of liver fat fraction quantification for PCCT in obese patients. This method may improve opportunistic screening for CT in the future.

Computed Tomography-Based Radiomics Analysis for Prediction of Response to Neoadjuvant Chemotherapy in Breast Cancer Patients.

PURPOSE: Previous studies have pointed out that magnetic resonance- and fluorodeoxyglucose positron emission tomography-based radiomics had a high predictive value for the response of the neoadjuvant chemotherapy (NAC) in breast cancer by respectively characterizing tumor heterogeneity of the relaxation time and the glucose metabolism. However, it is unclear whether computed tomography (CT)-based radiomics based on density heterogeneity can predict the response of NAC. This study aimed to develop and validate a CT-based radiomics nomogram to predict the response of NAC in breast cancer. METHODS: A total of 162 breast cancer patients (110 in the training cohort and 52 in the validation cohort) who underwent CT scans before receiving NAC and had pathological response results were retrospectively enrolled. Grades 4 to 5 cases were classified as response to NAC. According to the Miller-Payne grading system, grades 1 to 3 cases were classified as nonresponse to NAC. Radiomics features were extracted, and the optimal radiomics features were obtained to construct a radiomics signature. Multivariate logistic regression was used to develop the clinical prediction model and the radiomics nomogram that incorporated clinical characteristics and radiomics score. We assessed the performance of different models, including calibration and clinical usefulness. RESULTS: Eight optimal radiomics features were obtained. Human epidermal growth factor receptor 2 status and molecular subtype showed statistical differences between the response group and the nonresponse group. The radiomics nomogram had more favorable predictive efficacy than the clinical prediction model (areas under the curve, 0.82 vs 0.70 in the training cohort; 0.79 vs 0.71 in the validation cohort). The Delong test showed that there are statistical differences between the clinical prediction model and the radiomics nomogram ( z = 2.811, P = 0.005 in the training cohort). The decision curve analysis showed that the radiomics nomogram had higher overall net benefit than the clinical prediction model. CONCLUSION: The radiomics nomogram based on CT radiomics signature and clinical characteristics has favorable predictive efficacy for the response of NAC in breast cancer.

Effective deep learning approaches for predicting COVID-19 outcomes from chest computed tomography volumes.

The rapid evolution of the novel coronavirus disease (COVID-19) pandemic has resulted in an urgent need for effective clinical tools to reduce transmission and manage severe illness. Numerous teams are quickly developing artificial intelligence approaches to these problems, including using deep learning to predict COVID-19 diagnosis and prognosis from chest computed tomography (CT) imaging data. In this work, we assess the value of aggregated chest CT data for COVID-19 prognosis compared to clinical metadata alone. We develop a novel patient-level algorithm to aggregate the chest CT volume into a 2D representation that can be easily integrated with clinical metadata to distinguish COVID-19 pneumonia from chest CT volumes from healthy participants and participants with other viral pneumonia. Furthermore, we present a multitask model for joint segmentation of different classes of pulmonary lesions present in COVID-19 infected lungs that can outperform individual segmentation models for each task. We directly compare this multitask segmentation approach to combining feature-agnostic volumetric CT classification feature maps with clinical metadata for predicting mortality. We show that the combination of features derived from the chest CT volumes improve the AUC performance to 0.80 from the 0.52 obtained by using patients' clinical data alone. These approaches enable the automated extraction of clinically relevant features from chest CT volumes for risk stratification of COVID-19 patients.

Application of Unenhanced Computed Tomography Texture Analysis to Differentiate Pancreatic Adenosquamous Carcinoma from Pancreatic Ductal Adenocarcinoma.

OBJECTIVE: The objective of this study was to investigate the application of unenhanced computed tomography (CT) texture analysis in differentiating pancreatic adenosquamous carcinoma (PASC) from pancreatic ductal adenocarcinoma (PDAC). METHODS: Preoperative CT images of 112 patients (31 with PASC, 81 with PDAC) were retrospectively reviewed. A total of 396 texture parameters were extracted from AnalysisKit software for further texture analysis. Texture features were selected for the differentiation of PASC and PDAC by the Mann-Whitney U test, univariate logistic regression analysis, and the minimum redundancy maximum relevance algorithm. Furthermore, receiver operating characteristic (ROC) curve analysis was performed to evaluate the diagnostic performance of the texture feature-based model by the random forest (RF) method. Finally, the robustness and reproducibility of the predictive model were assessed by the 10-times leave-group-out cross-validation (LGOCV) method. RESULTS: In the present study, 10 texture features to differentiate PASC from PDAC were eventually retained for RF model construction after feature selection. The predictive model had a good classification performance in differentiating PASC from PDAC, with the following characteristics: sensitivity, 95.7%; specificity, 92.5%; accuracy, 94.3%; positive predictive value (PPV), 94.3%; negative predictive value (NPV), 94.3%; and area under the ROC curve (AUC), 0.98. Moreover, the predictive model was proved to be robust and reproducible using the 10-times LGOCV algorithm (sensitivity, 90.0%; specificity, 71.3%; accuracy, 76.8%; PPV, 59.0%; NPV, 95.2%; and AUC, 0.80). CONCLUSION: The unenhanced CT texture analysis has great potential for differentiating PASC from PDAC.

A Retrospectively Study: Diagnosis of Pathological Types of Malignant Lung Tumors by Dual-layer Detector Spectral Computed Tomography.

Object: By retrospectively analyzing the energy spectrum of squamous cell carcinoma, adenocarcinoma, small cell lung cancer (SCLC), and pulmonary metastases that underwent dual-layer detector spectral computed tomography (DLCT) 3-phase scan of the chest, we explored the value of a multiparameter energy spectrum in the assessment of pathological types of lung tumors. Methods: Cases of squamous cell carcinoma (n = 20), adenocarcinoma (n = 24), SCLC (n = 26), and metastases (n = 14) were collected. Then the largest cross-sectional area (LCA) of the lesion, computed tomography (CT) values in the plain scan phase, arterial and venous phases (HU, HUa, and HUv), iodine concentration, and effective atomic number in the arterial and venous phases (ICa, ICv, Zeff[a], and Zeff[v]) were measured and compared among the nonsmall cell lung cancer (NSCLC), SCLC and metastases, and other 3 groups of SCLC, squamous cell carcinoma, and adenocarcinoma. Results: Only the LCA is statistically different among SCLC, NSCLC, and metastases (P < .05). And the treated subgroup analysis did not show significant differences among the groups. However, the untreated subgroup analysis showed that there was a significant difference between NSCLC and metastases in LCA, SCLC and metastases in ICa, NSCLC and SCLC in HUv, NSCLC and SCLC in Zeff(v) (P < .05). Conclusion: The energy spectrum parameters of DLCT have a certain clinical value in distinguishing NSCLC from SCLC in the Zeff(v) and distinguishing SCLC from metastases in the ICa.

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.

Validation of the ECHS non trauma cranial CT rule in Australia: A prospective cohort study.

INTRODUCTION: Computed tomography (CT) is a commonly used imaging modality in Emergency Departments (EDs), however its use is questionable in many low yield settings. The Emergency CT Head score (ECHS) is a recently published clinical tool that assists in stratifying the need for CT brain (CTB) for patients presenting without a history of trauma. We sought to validate this tool in an Australian ED setting. METHODS: We prospectively evaluated 412 patients who received CTB without a history of trauma at a large Australian ED. We assessed them for the 4 main ECHS data points: focal neurological deficit on physical examination, new acute onset headache, transient neurological deficit, and a combination of new onset seizures with an altered conscious state. We examined their association with acute and chronic CTB findings. We then applied the ECHS to our data, calculating its sensitivity and its appropriateness at this single site via the calculation of a receiver operating curve (ROC). RESULTS: 10.2% of all CTB performed were positive for an acute or chronic abnormality. Only sex (male) and focal motor deficit were independent predictors of positive CTB at univariate analysis. The ECHS did not perform as anticipated in our population, with a ROC area under the curve of 0.498. An ECHS score of >0, which has been proposed as the threshold to not require imaging, had sensitivity of only 83.3% in our population. CONCLUSIONS: Further research and validation is required in order to safely implement the ECHS clinical score in the Australian ED setting.

Not so FAST-Chest ultrasound underdiagnoses traumatic pneumothorax.

BACKGROUND: Ultrasonography for trauma is a widely used tool in the initial evaluation of trauma patients with complete ultrasonography of trauma (CUST) demonstrating equivalence to computed tomography (CT) for detecting clinically significant abdominal hemorrhage. Initial reports demonstrated high sensitivity of CUST for the bedside diagnosis of pneumothorax. We hypothesized that the sensitivity of CUST would be greater than initial supine chest radiograph (CXR) for detecting pneumothorax. METHODS: A retrospective analysis of patients diagnosed with pneumothorax from 2018 through 2020 at a Level I trauma center was performed. Patients included had routine supine CXR and CUST performed prior to intervention as well as confirmatory CT imaging. All CUST were performed during the initial evaluation in the trauma bay by a registered sonographer. All imaging was evaluated by an attending radiologist. Subgroup analysis was performed after excluding occult pneumothorax. Immediate tube thoracostomy was defined as tube placement with confirmatory CXR within 8 hours of admission. RESULTS: There were 568 patients screened with a diagnosis of pneumothorax, identifying 362 patients with a confirmed pneumothorax in addition to CXR, CUST, and confirmatory CT imaging. The population was 83% male, had a mean age of 45 years, with 85% presenting due to blunt trauma. Sensitivity of CXR for detecting pneumothorax was 43%, while the sensitivity of CUST was 35%. After removal of occult pneumothorax (n = 171), CXR was 78% sensitive, while CUST was 65% sensitive (p < 0.01). In this subgroup, CUST had a false-negative rate of 36% (n = 62). Of those patients with a false-negative CUST, 50% (n = 31) underwent tube thoracostomy, with 85% requiring immediate placement. CONCLUSION: Complete ultrasonography of trauma performed on initial trauma evaluation had lower sensitivity than CXR for identification of pneumothorax including clinically significant pneumothorax requiring tube thoracostomy. Using CUST as the primary imaging modality in the initial evaluation of chest trauma should be considered with caution. LEVEL OF EVIDENCE: Diagnostic Test study, Level IV.

Comparison of lung image quality between CT Ark and Brilliance 64 CT during COVID-19.

AIM: This study is to compare the lung image quality between shelter hospital CT (CT Ark) and ordinary CT scans (Brilliance 64) scans. METHODS: The patients who received scans with CT Ark or Brilliance 64 CT were enrolled. Their lung images were divided into two groups according to the scanner. The objective evaluation methods of signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were used. The subjective evaluation methods including the evaluation of the fine structure under the lung window and the evaluation of the general structure under the mediastinum window were compared. Kappa method was used to assess the reliability of the subjective evaluation. The subjective evaluation results were analyzed using the Wilcoxon rank sum test. SNR and CNR were tested using independent sample t tests. RESULTS: There was no statistical difference in somatotype of enrolled subjects. The Kappa value between the two observers was between 0.68 and 0.81, indicating good consistency. For subjective evaluation results, the rank sum test P value of fine structure evaluation and general structure evaluation by the two observers was ≥ 0.05. For objective evaluation results, SNR and CNR between the two CT scanners were significantly different (P<0.05). Notably, the absolute values ​​of SNR and CNR of the CT Ark were larger than Brilliance 64 CT scanner. CONCLUSION: CT Ark is fully capable of scanning the lungs of the COVID-19 patients during the epidemic in the shelter hospital.

Adrenal wash-out CT: moderate diagnostic value in distinguishing benign from malignant adrenal masses.

OBJECTIVE: Reliable results of wash-out CT in the diagnostic workup of adrenal incidentalomas are scarce. Thus, we evaluated the diagnostic accuracy of delayed wash-out CT and determined thresholds to accurately differentiate adrenal masses. DESIGN: Retrospective, single-center cohort study including 216 patients with 252 adrenal lesions who underwent delayed wash-out CT. Definitive diagnoses based on histopathology (n = 92) or comprehensive follow-up. METHODS: Size, average attenuation values of the adrenal lesions in all CT scan phases, and absolute and relative percentage wash-out (APW/RPW) were determined by an expert radiologist blinded for clinical data. Adrenal lesions with unenhanced attenuation values >10 Hounsfield units (HU) built a subgroup (n = 142). Diagnostic accuracy was calculated. RESULTS: The study group consisted of 171 adenomas, 32 other benign tumors, 11 pheochromocytomas, 9 adrenocortical carcinomas, and 29 other malignant tumors. All (potentially) malignant and 46% of benign lesions showed unenhanced attenuation values >10 HU. In this most relevant subgroup, the established thresholds of 60% for APW and 40% for RPW misclassified 35.9 and 35.2% of the masses, respectively. When we applied optimized cutoffs (APW >83%; RPW >58%) and excluded pheochromocytomas, we missed only one malignant tumor by APW and none by RPW. However, only 11 and 15% of the benign tumors were correctly identified. CONCLUSIONS: Wash-out CT with the established thresholds for APW and RPW is insufficient to reliably diagnose adrenal masses. Using the proposed cutoff of 58% for RPW, malignant tumors will be correctly identified, but the added value is limited, namely 15% of patients with benign tumors can be prevented from additional imaging or even unnecessary surgery.

MRI versus CT for the detection of pulmonary nodules: A meta-analysis.

BACKGROUND: Computed tomography (CT) is the current gold standard for the detection of pulmonary nodules but has high radiation burden. In contrast, many radiologists tried to use magnetic resonance imaging (MRI) to replace CT because MRI has no radiation burden associated. Due to the lack of high-level evidence of comparison of the diagnostic accuracy of MRI versus CT for detecting pulmonary nodules, it is unknown whether CT can be replaced successfully by MRI. Therefore, the aim of this study was to compare the diagnostic accuracy of MRI versus CT for detecting pulmonary nodules. METHODS: Electronic databases PubMed, EmBase, and Cochrane Library were systematically searched from their inception to September 2017 to identify studies in which CT/MRI was used to diagnose pulmonary nodules. According to true positive, true negative, false negative, and false positive extracted from the included studies, we calculate the pooled sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), and area under the curve (AUC) using Stata version 14.0 software (STATA Corp, TX). RESULTS: A total of 8 studies involving a total of 653 individuals were included. The pooled sensitivity, specificity, PLR, NLR, and AUC were 0.91 (95% confidence interval [CI]: 0.80-0.96), 0.76 (95%CI: 0.58-0.87), 3.72 (95%CI: 2.05-6.76), 0.12 (95%CI: 0.06-0.27), and 0.91 (95%CI: 0.88-0.93) for MRI respectively, while the pooled sensitivity, specificity, PLR, NLR, and AUC for CT were 1.00 (95%CI: 0.95-1.00), 0.99 (95%CI: 0.78-1.00), 79.35 (95%CI: 3.68-1711.06), 0.00 (95%CI: 0.00-0.06), and 1.00 (95%CI: 0.99-1.00), respectively. Further, we compared the diagnostic accuracy of CT versus MRI and found that compared with MRI, CT shows statistically higher sensitivity (odds ratio [OR] for MRI vs CT: 0.91; 95%CI: 0.85-0.98; P value .010), specificity (OR: 0.82; 95%CI: 0.69-0.97; P value .019), PLR (OR: 0.29; 95%CI: 0.10-0.83; P value 0.02), AUC (OR: 0.91; 95%CI: 0.89-0.94; P value < .001), and lower NLR (OR: 8.72; 95%CI: 1.57-48.56; P value .013). CONCLUSION: Our study suggested both CT and MRI have a high diagnostic accuracy in diagnosing pulmonary nodules, while CT was superior to MRI in sensitivity, specificity, PLR, NLR, and AUC, indicating that in terms of the currently available evidence, MRI could not replace CT in diagnosing pulmonary nodules.

Advantages of CT nano-contrast agent in tumor diagnosis: A retrospective study.

ABSTRACT: The purpose of this retrospective study was to explore the advantages of computed tomography (CT) nano-contrast agent in tumor diagnosis.A total of 100 patients with malignant tumor who were diagnosed in Shaanxi Province Public Hospital between January 2018 and January 2019 were included in this retrospective study. They were randomly divided into observation and control groups with 50 patients in each group. The patients in the observation group used new type of nano-contrast agent for examination, and the patients in the control group used traditional iohexol contrast agent for examination. The detection rate, misdiagnosis rate, and incidence of adverse reactions were observed. In addition, single photon emission computed tomography or CT scan was performed on patients to observe the radioactive concentration.The detection rate was 100% in the observation group and 84% in the control group, and the difference between the 2 groups was statistically significant (χ2 = 8.763, P = .001). The incidence of adverse reactions was 2% in the observation group and 30% in the control group, and the difference between the 2 groups was significantly different (χ2 = 12.683, P = .000). The radioactive concentration in the observation group was markedly higher than that in the control group (t = 19.692, P = .001).The use of CT nano-contrast agent in tumor diagnosis had higher detection rate of tumor and radioactive concentration, and it had lower misdiagnosis rate and adverse reaction rate than traditional iohexol contrast agent.

Executive summary screening for Lung Cancer: Chest guideline and expert panel report

Low-dose chest CT screening for lung cancer has become a standard of care in the United States, in large part because of the results of the National Lung Screening Trial (NLST). Additional evidence supporting the net benefit of low-dose chest CT screening for lung cancer, and increased experience in minimizing the potential harms, has accumulated since the prior iteration of these guidelines. Here, we update the evidence base for the benefit, harms, and implementation of low-dose chest CT screening. We use the updated evidence base to provide recommendations where the evidence allows, and statements based on experience and expert consensus where it does not. Approved panelists reviewed previously developed key questions using the Population, Intervention, Comparator, Outcome format to address the benefit and harms of low-dose CT screening, and key areas of program implementation. A systematic literature review was conducted using MEDLINE via PubMed, Embase, and the Cochrane Library on a quarterly basis since the time of the previous guideline publication. Reference lists from relevant retrievals were searched, and additional papers were added. Retrieved references were reviewed for relevance by two panel members. The quality of the evidence was assessed for each critical or important outcome of interest using the Grading of Recommendations, Assessment, Development and Evaluation approach. Meta-analyses were performed where appropriate. Important clinical questions were addressed based on the evidence developed from the systematic literature review. Graded recommendations and ungraded statements were drafted, voted on, and revised until consensus was reached. The systematic literature review identified 75 additional studies that informed the response to the 12 key questions that were developed. Additional clinical questions were addressed resulting in seven graded recommendations and nine ungraded consensus statements. Evidence suggests that low-dose CT screening for lung cancer can result in a favorable balance of benefit and harms. The selection of screen-eligible individuals, the quality of imaging and image interpretation, the management of screen-detected findings, and the effectiveness of smoking cessation interventions can impact this balance.

The diagnostic value of chest X-ray in coronavirus disease 2019: A comparative study of X-ray and CT.

As the coronavirus disease 2019 (COVID-19) epidemic spreads around the world, the demand for imaging examinations increases accordingly. The value of conventional chest radiography (CCR) remains unclear. In this study, we aimed to investigate the diagnostic value of CCR in the detection of COVID-19 through a comparative analysis of CCR and CT. This study included 49 patients with 52 CT images and chest radiographs of pathogen-confirmed COVID-19 cases and COVID-19-suspected cases that were found to be negative (non-COVID-19). The performance of CCR in detecting COVID-19 was compared to CT imaging. The major signatures that allowed for differentiation between COVID-19 and non-COVID-19 cases were also evaluated. Approximately 75% (39/52) of images had positive findings on the chest x-ray examinations, while 80.7% (42/52) had positive chest CT scans. The COVID-19 group accounted for 88.4% (23/26) of positive chest X-ray examinations and 96.1% (25/26) of positive chest CT scans. The sensitivity, specificity, and accuracy of CCR for abnormal shadows were 88%, 80%, and 87%, respectively, for all patients. For the COVID-19 group, the accuracy of CCR was 92%. The primary signature on CCR was flocculent shadows in both groups. The shadows were primarily in the bi-pulmonary, which was significantly different from non-COVID-19 patients (p = 0.008). The major CT finding of COVID-19 patients was ground-glass opacities in both lungs, while in non-COVID-19 patients, consolidations combined with ground-glass opacities were more common in one lung than both lungs (p = 0.0001). CCR showed excellent performance in detecting abnormal shadows in patients with confirmed COVID-19. However, it has limited value in differentiating COVID-19 patients from non-COVID-19 patients. Through the typical epidemiological history, laboratory examinations, and clinical symptoms, combined with the distributive characteristics of shadows, CCR may be useful to identify patients with possible COVID-19. This will allow for the rapid identification and quarantine of patients.

A real-world demonstration of machine learning generalizability in the detection of intracranial hemorrhage on head computerized tomography.

Machine learning (ML) holds great promise in transforming healthcare. While published studies have shown the utility of ML models in interpreting medical imaging examinations, these are often evaluated under laboratory settings. The importance of real world evaluation is best illustrated by case studies that have documented successes and failures in the translation of these models into clinical environments. A key prerequisite for the clinical adoption of these technologies is demonstrating generalizable ML model performance under real world circumstances. The purpose of this study was to demonstrate that ML model generalizability is achievable in medical imaging with the detection of intracranial hemorrhage (ICH) on non-contrast computed tomography (CT) scans serving as the use case. An ML model was trained using 21,784 scans from the RSNA Intracranial Hemorrhage CT dataset while generalizability was evaluated using an external validation dataset obtained from our busy trauma and neurosurgical center. This real world external validation dataset consisted of every unenhanced head CT scan (n = 5965) performed in our emergency department in 2019 without exclusion. The model demonstrated an AUC of 98.4%, sensitivity of 98.8%, and specificity of 98.0%, on the test dataset. On external validation, the model demonstrated an AUC of 95.4%, sensitivity of 91.3%, and specificity of 94.1%. Evaluating the ML model using a real world external validation dataset that is temporally and geographically distinct from the training dataset indicates that ML generalizability is achievable in medical imaging applications.

Effect of focused protocol on reducing radiation dose for children who require computed tomography for suspected appendicitis.

BACKGROUND: Despite efforts to incorporate ultrasound into the evaluation of children for appendicitis, computed tomography (CT) is often used to aid in its diagnosis. CT scans, however, expose children to a considerable amount of radiation. In 2017, our institution began using a height-based Focused CT protocol for children with suspected appendicitis in need of CT. OBJECTIVE: To compare the radiation dose received by children with suspected appendicitis who underwent a Standard CT of the abdomen and pelvis (CTAP) with that of a Focused CT. METHODS: We conducted a retrospective study of children <18 years who underwent a CT scan for suspected appendicitis (2014-2020). We included all patients whose indication for CT was "appendicitis" or "right lower quadrant pain" and excluded those whose CT scan record lacked a radiation dose report. The effective radiation dose delivered was calculated using the dose-length product from the dose report. We compared the effective dose of those who received a Standard CTAP to those who received a Focused CT. To account for differences in radiation dose over time and by CT scanner, analyses were adjusted for CT dose index volume (CTDIvol) and size-specific dose estimate (SSDE) using quantile regression. RESULTS: A total of 474 patients who underwent CT were included. Prior to CT, 362(76%) had received an ultrasound. In total, 309(65%) patients underwent a Standard CTAP and 165(35%) underwent a Focused CT. The appendix was identified in 259(84%) Standard CTAPs compared to 151(92%) Focused CTs (p = 0.02). Compared to the Standard CTAP, children who received a Focused CT were exposed to a significantly lower effective dose (relative difference: CTDI-adjusted -13%[95% CI:-21,-5]; SSDE-adjusted -14%[95% CI:-24,-3]). CONCLUSIONS: Our height-based Focused CT protocol reduces radiation for children undergoing CT evaluation for suspected appendicitis without sacrificing diagnostic accuracy. Further study is needed to validate these findings at other institutions.

Acceptance and image quality of high-resolution peripheral quantitative computed tomography of the metacarpophalangeal joints in rheumatoid arthritis.

OBJECTIVE: High-resolution peripheral quantitative computed tomography (HR-pQCT) requires longer immobilization time than conventional radiography, which challenges patient acceptance and image quality. Therefore, the aim was to investigate the acceptance of HR-pQCT in patients with rheumatoid arthritis (RA), and secondly the effect of an inflatable hand immobilization device on motion artefacts of the metacarpophalangeal (MCP) joints. METHODS: Fifty patients with established RA and a median (interquartile range) age of 64.3 (55.0-71.2) years had their MCP joints scanned by HR-pQCT with the hand positioned with and without an inflatable immobilization device followed by a full radiographic examination and a questionnaire on the imaging experience. The comparability of the erosion measures was investigated with and without the immobilization device using Bland-Altman plot and intrareader repeatability by intraclass correlation coefficient. The motion artefacts were graded for each acquisition, and intrareader repeatability was investigated by Cohen's kappa coefficient. RESULTS: Forty percent of the patients preferred HR-pQCT imaging, only 6% preferred conventional X-ray. Seventy-four percent reported it was not difficult to keep their fingers steady during the scan. Sixty percent of the patients reported the immobilization device helped keep their fingers steady. However, as motion artefacts were sparse, no clinically relevant difference was observed concerning the effect of the immobilization device on readability. The intrareader repeatability and comparability for the erosion measures were excellent. CONCLUSION: The high patient acceptance adds to the feasibility of HR-pQCT imaging of MCP joints in RA. The inflatable immobilization device did not reduce motion-induced image degradation.

Variation in pediatric cervical spine imaging across trauma centers-A cause for concern?

BACKGROUND: Traumatic pediatric cervical spine injury can be challenging to diagnose, and the clinical algorithms meant to aid physicians differ from adult trauma protocols. Despite the existence of standardized guidelines, imaging decisions may vary according to physician education, subjective assessment, and experience with pediatric trauma patients. Our study investigates the rates of pediatric posttraumatic cervical spine imaging across trauma centers, hypothesizing that more specialized centers will have lower rates of advanced cervical spine imaging. METHODS: The 2015 to 2016 Trauma Quality Improvement Program database was reviewed for patients younger than 18 years- to assess rates of cervical spine imaging on presentation across different trauma centers. Propensity stratification logistic regression was performed controlling for patient- and center-specific variables. p Values less than 0.05 were considered significant. RESULTS: Of 110,769 pediatric trauma patients, 35.2% were female, and the average age was 9.6 years. Overall, 3.6% had cervical spine computed tomography (CT) and less than 1% had cervical spine MRI or X-ray. Compared with all others, Level I trauma centers were significantly less likely to use cervical spine CT for the initial evaluation of younger (≤14 years) but not older trauma patients (adjusted odds ratio [AOR], 0.89; 95% confidence interval [CI], 0.80-0.99; AOR, 0.97; 95% CI, 0.87-1.09); Level I centers had higher odds of cervical spine MRI use, but only for patients 14 years or younger (AOR, 1.63; 95% CI, 1.09-2.44). Pediatric-designated trauma centers had significantly lower odds of cervical spine CT (≤14 years: AOR, 0.70; 95% CI, 0.63-0.78; >14 years: AOR, 0.67; 95% CI, 0.67-0.75) and higher odds of cervical spine X-ray (≤14 years: AOR, 4.75; 95% CI, 3.55-6.36; >14 years: AOR, 4.50; 95% CI, 2.72-7.45) for all ages, but higher odds of cervical spine MRI for younger patients only (≤14 years: AOR, 2.10; 95% CI, 1.38-3.21). CONCLUSION: Level I and pediatric designations were associated with lower rates of cervical spine CT. Pediatric centers were also more likely to use cervical spine X-ray. This variability of imaging use further supports the need to disseminate and educate providers on pediatric-specific cervical spine evaluation guidelines. LEVEL OF EVIDENCE: Prognostic and epidemiological, level III.

Acute Stroke Imaging Research Roadmap IV: Imaging Selection and Outcomes in Acute Stroke Clinical Trials and Practice.

BACKGROUND AND PURPOSE: The Stroke Treatment Academic Industry Roundtable (STAIR) sponsored an imaging session and workshop during the Stroke Treatment Academic Industry Roundtable XI via webinar on October 1 to 2, 2020, to develop consensus recommendations, particularly regarding optimal imaging at primary stroke centers. METHODS: This forum brought together stroke neurologists, neuroradiologists, neuroimaging research scientists, members of the National Institute of Neurological Disorders and Stroke, industry representatives, and members of the US Food and Drug Administration to discuss imaging priorities in the light of developments in reperfusion therapies, particularly in an extended time window, and reinvigorated interest in brain cytoprotection trials. RESULTS: The imaging session summarized and compared the imaging components of recent acute stroke trials and debated the optimal imaging strategy at primary stroke centers. The imaging workshop developed consensus recommendations for optimizing the acquisition, analysis, and interpretation of computed tomography and magnetic resonance acute stroke imaging, and also recommendations on imaging strategies for primary stroke centers. CONCLUSIONS: Recent positive acute stroke clinical trials have extended the treatment window for reperfusion therapies using imaging selection. Achieving rapid and high-quality stroke imaging is therefore critical at both primary and comprehensive stroke centers. Recommendations for enhancing stroke imaging research are provided.

A transfer learning approach to facilitate ComBat-based harmonization of multicentre radiomic features in new datasets.

PURPOSE: To facilitate the demonstration of the prognostic value of radiomics, multicenter radiomics studies are needed. Pooling radiomic features of such data in a statistical analysis is however challenging, as they are sensitive to the variability in scanner models, acquisition protocols and reconstruction settings, which is often unavoidable in a multicentre retrospective analysis. A statistical harmonization strategy called ComBat was utilized in radiomics studies to deal with the "center-effect". The goal of the present work was to integrate a transfer learning (TL) technique within ComBat-and recently developed alternate versions of ComBat with improved flexibility (M-ComBat) and robustness (B-ComBat)-to allow the use of a previously determined harmonization transform to the radiomic feature values of new patients from an already known center. MATERIAL AND METHODS: The proposed TL approach were incorporated in the four versions of ComBat (standard, B, M, and B-M ComBat). The proposed approach was evaluated using a dataset of 189 locally advanced cervical cancer patients from 3 centers, with magnetic resonance imaging (MRI) and positron emission tomography (PET) images, with the clinical endpoint of predicting local failure. The impact performance of the TL approach was evaluated by comparing the harmonization achieved using only parts of the data to the reference (harmonization achieved using all the available data). It was performed through three different machine learning pipelines. RESULTS: The proposed TL technique was successful in harmonizing features of new patients from a known center in all versions of ComBat, leading to predictive models reaching similar performance as the ones developed using the features harmonized with all the data available. CONCLUSION: The proposed TL approach enables applying a previously determined ComBat transform to new, previously unseen data.