Experienced Radiologists Research Articles

Analysis of Periprocedural X-ray Exposure in Transarterial Radioembolization with Glass or Resin Microspheres.

Background: Transarterial Radioembolization (TARE) is an effective treatment option for both primary and secondary liver malignancies. However, challenging anatomical conditions can lead to prolonged fluoroscopy times (FT), elevated doses of periprocedural X-radiation (DAP), and increased use of contrast agents (CAs). In this study, we examined the influence of our radiologists' experience and the choice of microspheres on X-ray exposure and CA doses in TARE. Material and Methods: Datasets comprising 161 TARE and 164 preprocedural evaluation angiographies (TARE-EVA) were analyzed. Our study focused on assessing DAP, FT, and CA concerning both microsphere types, the radiologist's experience, and whether the same radiologist performed both the TARE-EVA and the actual TARE. Results: In TARE, the use of resin microspheres resulted in significantly higher FT and CA compared to glass microspheres (14.3 ± 1.6 min vs. 10.6 ± 1.1 min and 43 ± 2.2 mL vs. 33.6 ± 2.1 mL, p < 0.05), with no notable differences in DAP (p = 0.13). Experienced radiologists demonstrated reduced FT/DAP, with a 19% decrease in DAP and 53% in FT during the evaluation angiography (p < 0.05) and a 49% reduction in DAP during the actual TARE (p < 0.05), with no statistical differences in FT. Performing TARE and TARE-EVA under the same radiologist led to a 43% reduction in DAP and a 25% decrease in FT (p < 0.05, respectively). Conclusions: To mitigate X-radiation exposure, it is advisable for radiologists to undergo thorough training, and, ideally, the same radiologist should conduct both the TARE and the TARE-EVA. While the use of glass spheres may decrease intraarterial CA, it does not significantly impact periprocedural X-ray exposure.

One-Stop Automated Diagnostic System for Carpal Tunnel Syndrome in Ultrasound Images Using Deep Learning

Ultrasound (US) examination has unique advantages in diagnosing carpal tunnel syndrome (CTS), although identification of the median nerve (MN) and diagnosis of CTS depend heavily on the expertise of examiners. In the aim of alleviating this problem, we developed a one-stop automated CTS diagnosis system (OSA-CTSD) and evaluated its effectiveness as a computer-aided diagnostic tool. We combined real-time MN delineation, accurate biometric measurements and explainable CTS diagnosis into a unified framework, called OSA-CTSD. We then collected a total of 32,301 static images from US videos of 90 normal wrists and 40 CTS wrists for evaluation using a simplified scanning protocol. The proposed model exhibited better segmentation and measurement performance than competing methods, with a Hausdorff distance (95th percentile) score of 7.21 px, average symmetric surface distance score of 2.64 px, Dice score of 85.78% and intersection over union score of 76.00%. In the reader study, it exhibited performance comparable to the average performance of experienced radiologists in classifying CTS and outperformed inexperienced radiologists in terms of classification metrics (e.g., accuracy score 3.59% higher and F1 score 5.85% higher). Diagnostic performance of the OSA-CTSD was promising, with the advantages of real-time delineation, automation and clinical interpretability. The application of such a tool not only reduces reliance on the expertise of examiners but also can help to promote future standardization of the CTS diagnostic process, benefiting both patients and radiologists.

Learnability of Thyroid Nodule Assessment on Ultrasonography: Using a Big Data Set.

The aims of the work described here were to evaluate the learnability of thyroid nodule assessment on ultrasonography (US) using a big data set of US images and to evaluate the diagnostic utilities of artificial intelligence computer-aided diagnosis (AI-CAD) used by readers with varying experience to differentiate benign and malignant thyroid nodules. Six college freshmen independently studied the "learning set" composed of images of 13,560 thyroid nodules, and their diagnostic performance was evaluated after their daily learning sessions using the "test set" composed of images of 282 thyroid nodules. The diagnostic performance of two residents and an experienced radiologist was evaluated using the same "test set." After an initial diagnosis, all readers once again evaluated the "test set" with the assistance of AI-CAD. Diagnostic performance of almost all students increased after the learning program. Although the mean areas under the receiver operating characteristic curves (AUROCs) of residents and the experienced radiologist were significantly higher than those of students, the AUROCs of five of the six students did not differ significantly compared with that of the one resident. With the assistance of AI-CAD, sensitivity significantly increased in three students, specificity in one student, accuracy in four students and AUROC in four students. Diagnostic performance of the two residents and the experienced radiologist was better with the assistance of AI-CAD. A self-learning method using a big data set of US images has potential as an ancillary tool alongside traditional training methods. With the assistance of AI-CAD, the diagnostic performance of readers with varying experience in thyroid imaging could be further improved.

Interobserver reliability of classifying shoulder calcific tendinopathy on plain radiography and ultrasound

Introduction: Shoulder calcific tendinopathy is a frequent cause of shoulder pain. Diagnosis is usually based on ultrasound (US) and/or x-ray. US is considered an inherently operator-dependent imaging modality and, interobserver variability has previously been described by experts in the musculoskeletal US. The main objective of this study is to assess the interobserver agreement for shoulder calcific tendinopathy attending to the size, type, and location of calcium analyzed in plain film and ultrasound among trained musculoskeletal radiologists.Material and methods: From June 2018 to May 2019, we conducted a prospective study. Patients diagnosed with shoulder pain related to calcific tendinopathy were included. Two different experienced musculoskeletal radiologists evaluated independently the plain film and the US.Results: Forty patients, with a mean age of 54.6 years, were included. Cohen´s kappa coefficient of 0.721 and 0.761 was obtained for the type of calcium encountered in plain film and the US, respectively. The location of calcification obtained a coefficient of 0.927 and 0.760 in plain film and US, respectively. The size of the calcification presented an intraclass correlation coefficient (ICC) of 0.891 and 0.86 in plain film and US respectively. No statistically significant differences were found in either measurement.Conclusion: This study shows very good interobserver reliability of type and size measurement (plain film and US) of shoulder calcifying tendinopathy in experienced musculoskeletal radiologists.Level of Evidence: III, prospective inter-observer study

Deep learning diagnostic performance and visual insights in differentiating benign and malignant thyroid nodules on ultrasound images.

This study aims to construct and evaluate a deep learning model, utilizing ultrasound images, to accurately differentiate benign and malignant thyroid nodules. The objective includes visualizing the model's process for interpretability and comparing its diagnostic precision with a cohort of 80 radiologists. We employed ResNet as the classification backbone for thyroid nodule prediction. The model was trained using 2096 ultrasound images of 655 distinct thyroid nodules. For performance evaluation, an independent test set comprising 100 cases of thyroid nodules was curated. In addition, to demonstrate the superiority of the artificial intelligence (AI) model over radiologists, a Turing test was conducted with 80 radiologists of varying clinical experience. This was meant to assess which group of radiologists' conclusions were in closer alignment with AI predictions. Furthermore, to highlight the interpretability of the AI model, gradient-weighted class activation mapping (Grad-CAM) was employed to visualize the model's areas of focus during its prediction process. In this cohort, AI diagnostics demonstrated a sensitivity of 81.67%, a specificity of 60%, and an overall diagnostic accuracy of 73%. In comparison, the panel of radiologists on average exhibited a diagnostic accuracy of 62.9%. The AI's diagnostic process was significantly faster than that of the radiologists. The generated heat-maps highlighted the model's focus on areas characterized by calcification, solid echo and higher echo intensity, suggesting these areas might be indicative of malignant thyroid nodules. Our study supports the notion that deep learning can be a valuable diagnostic tool with comparable accuracy to experienced senior radiologists in the diagnosis of malignant thyroid nodules. The interpretability of the AI model's process suggests that it could be clinically meaningful. Further studies are necessary to improve diagnostic accuracy and support auxiliary diagnoses in primary care settings.

Diagnosis and treatment of intravenous leiomyomatosis

Intravenous leiomyomatosis is a rare type of tumor that is histologically benign but biologically invasive. It originates from the smooth muscle of the uterine or the uterine vein. It can grow through the uterus and extend into the pelvic cavity, or grow along the veins without invading the wall of the venous vessel itself. The tumors are estrogen-dependent and can metastasize through the bloodstream. Thus, in addition to continuous growth, some tumors exhibit isolated growths in the venous system and heart chambers or show disseminated growth in the lungs, although distant metastasis to other regions usually do not occur. Currently, there is limited research on this disease, the majority of which are case reports, surgical experience summaries, and differentiation from ordinary gynecological myomas in terms of pathogenesis and radiological diagnostic experience. There are two main theories on the origin of the disease: uterine smooth muscle and smooth muscle of the uterine veins. Some studies have verified the role of estrogen, progesterone receptor-related pathways, and angiogenesis in the development of the disease. The clinical symptoms of this disease are varied, depending on the affected area. In the early stages, when the tumor only affects the pelvic cavity, patients show mild symptoms resulting from pelvic organ compression. When it progresses to the inferior vena cava and heart, patients show more complex symptoms resulting from venous return obstruction, cardiac obstruction, and hemodynamics appearing. Different institutions have proposed different disease staging and classification strategies for different clinical purposes. Some are based on the affected area of the lesion; others are based on the size of the tumor. Although surgery remains the main treatment for this disease, the specific surgical approach, adjuvant drug therapy, and prognosis still need further exploration.

Clinical Application of Dark-blood Imaging in Head and Neck CT Angiography: Effect on Image Quality and Plaque Visibility

The aim of this study was to explore the potential of a newly developed dark-blood imaging technique to improve image quality and plaque visibility in head and neck computed tomography (CT) angiography. Patients who underwent triphasic head and neck CT angiography scans from August 2021 to March 2023 were retrospectively enrolled (mean age 67.23±10.81 [SD] years, range 43-85 years, 64.7% male). The CT protocol consists of pre-contrast, arterial and delayed phases. Dark-blood images were postprocessed with the contrast-enhancement boost (CE-boost) technique. The quantitative assessment involved evaluating the CT value, image noise, contrast-to-noise ratio (CNR), and signal-to-noise ratio (SNR) of calcified plaque and non-calcified plaque. The plaque CNR relative to the vessel lumen (CNRplaque-lumen), vessel wall (CNRplaque-wall), and adjacent muscle (CNRplaque-muscle) was respectively calculated. Two experienced radiologists independently evaluated the CT images (5, best; 1, worst) by four characteristics including calcified plaque visibility, non-calcified plaque visibility, diagnostic confidence, and overall image quality. Inter-rater variability was also evaluated. The artery stenosis rate and plaque burden on dark-blood images were measured and compared with arterial phases. The intraclass correlation coefficient (ICC) was used for consistency analysis. The diagnostic accuracy of dark-blood images for the stenosis rate was evaluated by the area under the curve (AUC). A total of 43 patients with 54 calcified plaques and 34 non-calcified plaques were assessed in this study. When compared with pre-contrast and delayed phase, dark-blood images yielded significantly higher CNRplaque-lumen and CNRplaque-muscle of calcified (219.79±159.20 and 181.23±112.12, respectively) and non-calcified (30.30±29.11 and 6.28±4.75, respectively) plaques (all p<0.001). Calcified plaque SNR of dark-blood showed equal or slightly lower than other phases (p>0.05 or p=0.02). A major increase was observed in the non-calcified plaque SNR of dark-blood compared to the arterial phase (5.56±3.71 vs. 4.23±3.56, p=0.02), although there were no apparent differences compared to pre-contrast and delayed phases (p>0.05). In subjective analyzes, the calcified plaque visibility (4.99±0.07), non-calcified plaque visibility (4.62±0.48), overall image quality (4.81±0.34), and diagnostic confidence (4.74±0.36) in dark-blood images dominated the highest scores (p<0.001). The subjective scores of radiologists exhibited good consistency (all kappa value>0.7). The dark-blood image and the arterial phase image exhibited good consistency in identifying the stenosis rate (p<0.001). In the evaluation of plaque burden, the interobserver agreement for dark-blood images was higher compared to arterial phase images (ICC = 0.870 vs. 0.729). Compared to conventional triphasic head and neck CT angiography, the CE-boost derived dark-blood imaging demonstrated a significant improvement in image quality and visibility for both calcified and non-calcified plaque assessment.

A deep weakly semi-supervised framework for endoscopic lesion segmentation.

In the field of medical image analysis, accurate lesion segmentation is beneficial for the subsequent clinical diagnosis and treatment planning. Currently, various deep learning-based methods have been proposed to deal with the segmentation task. Albeit achieving some promising performances, the fully-supervised learning approaches require pixel-level annotations for model training, which is tedious and time-consuming for experienced radiologists to collect. In this paper, we propose a weakly semi-supervised segmentation framework, called Point Segmentation Transformer (Point SEGTR). Particularly, the framework utilizes a small amount of fully-supervised data with pixel-level segmentation masks and a large amount of weakly-supervised data with point-level annotations (i.e., annotating a point inside each object) for network training, which largely reduces the demand of pixel-level annotations significantly. To fully exploit the pixel-level and point-level annotations, we propose two regularization terms, i.e., multi-point consistency and symmetric consistency, to boost the quality of pseudo labels, which are then adopted to train a student model for inference. Extensive experiments are conducted on three endoscopy datasets with different lesion structures and several body sites (e.g., colorectal and nasopharynx). Comprehensive experimental results finely substantiate the effectiveness and the generality of our proposed method, as well as its potential to loosen the requirements of pixel-level annotations, which is valuable for clinical applications.

Assessment of inspiration and technical quality in anteroposterior thoracic radiographs using machine learning

IntroductionChest radiographs are the most performed radiographic procedure, but suboptimal technical factors can impact clinical interpretation. A deep learning model was developed to assess technical and inspiratory adequacy of anteroposterior chest radiographs. MethodsAdult anteroposterior chest radiographs (n = 2375) were assessed for technical adequacy, and if otherwise technically adequate, for adequacy of inspiration. Images were labelled by an experienced radiologist with one of three ground truth labels: inadequate technique (n = 605, 25.5 %), adequate inspiration (n = 900, 37.9 %), and inadequate inspiration (n = 870, 36.6 %). A convolutional neural network was then iteratively trained to predict these labels and evaluated using recall, precision, F1 and micro-F1, and Gradient-weighted Class Activation Mapping analysis on a hold-out test set. Impact of kyphosis on model accuracy was assessed. ResultsThe model performed best for radiographs with adequate technique, and worst for images with inadequate technique. Recall was highest (89 %) for radiographs with both adequate technique and inspiration, with recall of 81 % for images with adequate technique and inadequate inspiration, and 60 % for images with inadequate technique, although precision was highest (85 %) for this category. Per-class F1 was 80 %, 81 % and 70 % for adequate inspiration, inadequate inspiration, and inadequate technique respectively. Weighted F1 and Micro F1 scores were 78 %. Presence or absence of kyphosis had no significant impact on model accuracy in images with adequate technique. ConclusionThis study explores the promising performance of a machine learning algorithm for assessment of inspiratory adequacy and overall technical adequacy for anteroposterior chest radiograph acquisition. Implications for practiceWith further refinement, machine learning can contribute to education and quality improvement in radiology departments.

Ultrasound imaging of onychopapilloma: Evaluation of 43 cases

IntroductionOnychopapilloma is a benign tumor presenting as an isolated benign warty distal matrix and bed lesion in association with longitudinal erythronychia for which no underlying cause can be identified. Diagnosis might be challenging due to other, more severe, causes of erythronychia. ObjectivesThis study aimed to investigate the clinical and sonographic features of onychopapilloma using high-frequency ultrasound. MethodsBetween February 2022 and May 2023, a cross-sectional study was conducted, involving 39 patients displaying clinical features suggestive of onychopapilloma. These patients were selected at random and underwent examination using high-frequency ultrasound. High-frequency ultrasound assessments of the nail unit were carried out using 33 MHz and 24 MHz high-frequency linear transducers. An experienced radiologist performed the ultrasound scans while participants assumed an orthostatic seated position. Subsequently, the collected ultrasound images were independently evaluated by two radiologists specializing in nail injuries, ensuring a blinded assessment. This evaluation aimed to provide comprehensive insights into the imaging characteristics of onychopapilloma. ResultsHigh-frequency ultrasound revealed oval or linear hyperechoic masses located in the middle of the nail plate, without vascularity or posterior acoustic shadowing. Nail plate irregularity, thickening, and nail bed indentation were observed in some cases. These findings demonstrate that high-frequency ultrasound can aid in the diagnosis and surgical planning of onychopapilloma, providing detailed visualization of the lesion and surrounding structures. ConclusionsThis imaging method could represent a potential tool to differentiate onychopapilloma from other nail conditions, especially malignant neoplasms. However, further research is needed to validate these findings and assess long-term outcomes.

Correlation Between Contrast-Detail Analysis and Clinical Image Quality Assessment of Intrapulmonary Lesions in Dual-Energy Subtraction Chest Radiography Using the Two-Shot Method: A Phantom Study

Dual-energy subtraction (DES) imaging constitutes a technique that has demonstrated efficacy in enhancing the detectability of pulmonary nodules on chest radiographs. However, a simple and quantitative methodology for evaluating the clinical image quality of DES images is currently lacking. The objective of this study was to investigate the applicability of contrast-detail (C-D) phantom analysis to the visual clinical image quality evaluation of chest DES images. We employed a custom-made phantom incorporating the C-D phantom and a multipurpose anthropomorphic adult chest phantom. Two phantom sizes were utilized to simulate standard- and large-bodied adult patients for each phantom. The custom-made phantom images were scored automatically using dedicated software, yielding an inverse image quality figure (IQFinv) value. The multipurpose anthropomorphic adult chest phantom was employed in a visual grading analysis (VGA) study that was conducted by two experienced radiologists and five radiological technologists. Each nodule placed in the chest phantom image was rated on a 4-point Likert scale. A statistically significant correlation was observed between the VGA scores of the seven observers and the obtained IQFinv values. The findings of this study suggest that DES image analysis of the C-D phantom possesses the potential to be utilized for the evaluation of clinical DES image quality based on chest lesion detectability.

Interobserver variability in the determination of bone mineral density in Hounsfield units from differently configured fields of measurement in the cancellous bone of vertebral bodies from elderly body donors

BackgroundDue to the absence of suitable diagnostic procedures, osteoporosis (OP) is frequently detected late or not at all. Many elderly persons undergo computed tomographies (CT). The routine determination of Hounsfield units (HU) in bone as a part of these examinations could close a gap here. MethodsSpines were extracted from 22 body donors, fixed in a PVC water phantom, and subjected to a high-resolution CT investigation. Cancellous bone was examined and its bone mineral density measured in HU from cervical vertebra 3 to lumbar vertebra 5 (484 vertebral bodies). On sagittal sections, a circular and a rectangular region of interest (ROI) were defined in mid-vertebral cancellous bone, positioned manually, and the measurements were performed by three experienced radiologists. Bone mineral density (BMD), measured in mg/cm3, was used to determine the presence of OP. ResultsAll of the spines were osteoporotic. In the presence of a BMD below 60 mg/cm3 and HU values below 63.36 in lumbar vertebrae, there were significantly more vertebral body fractures in the thoracic and thoracolumbar spine. No difference was observed between the manually positioned circular and rectangular regions of interest (ROI) on the sagittal CT section (p > 0.05). Similar HU counts were obtained by the individual examiners (p > 0.05). The following formula was used to determine QCT values on a non-contrasted CT of the spine: QCT = 0.6 × HU + 13.7. ConclusionsMeasurement of the density of cancellous bone in HU can be used to determine BMD for estimating demineralization. Quantitative BMD values in mg/cm3, which can be calculated from the HU data, concur well with QCT values.

Impact of relaxing incisions on maxillofacial growth following Sommerlad–Furlow modified technique in patients with isolated cleft palate: a preliminary comparative study

ObjectiveTo estimate the impact of relaxing incisions on maxillofacial growth following Sommerlad-Furlow modified technique in patients with isolated cleft palate.Study designA Retrospective Cohort Study.MethodsA total of 90 participants, 60 patients with non-syndromic isolated soft and hard cleft palate underwent primary palatoplasty (30 patients received the Sommerlad-Furlow modified technique without relaxing incision (S.F−RI group), and 30 received Sommerlad-Furlow modified technique with relaxing (S.F+RI group) with no significant difference found between them regarding the cleft type, cleft width, and age at repair. While the other 30 were healthy noncleft participants with skeletal class I pattern as a Control group. The control group (C group) was matched with the patient groups in number, age, and sex. All participants had lateral cephalometric radiographs at least 5 years old age. The lateral cephalometric radiographs were taken with the same equipment by the same experienced radiologist while the participants were in centric occlusion and a standardized upright position, with the transporionic axis and Frankfort horizontal plane parallel to the surface of the floor. A well-trained assessor (S. Elayah) used DOLPHIN Imaging Software to trace twice to eliminate measurement errors. All the study variables were measured using stable landmarks, including 12 linear and 10 angular variants.ResultsThe mean age at collection of cephalograms was 6.03 ± 0.80 in the S.F+RI group, 5.96 ± 0.76 in the S.F−RI group, and 5.91 ± 0.87 in the C group.Regarding cranial base, the results showed no statistically significant differences between the three groups in S–N and S–N-Ba. While the S.F+R.I group had a significantly shortest S-Ba than the S.F−R.I & C groups (P = 0.01 & P < 0.01), but there was no statistically significant difference between S.F−R.I & C groups (P = 0.71).Regarding the skeletal maxilla, there was no significant difference between the S.F+R.I and S.F−R.I groups in all linear measurements (N-ANS and S-PM) except Co-A, the S.F+R.I group had significantly shorter Co-A than the S.F−R.I & C groups (P = < 0.01). While the angular measurement, S.F+R.I group had significantly less SNA angle than the S.F−R.I & C groups (P = < 0.01).Regarding mandibular bone, there were no statistically significant differences in all linear and angular mandibular measurements between the S.F+R.I and S.F−R.I.groups.Regarding intermaxillary relation, the S.F+R.I group had significant differences in Co-Gn—Co-A and ANB compared to the S.F−R.I & C groups (P = < 0.01). While there was no statistically significant difference in PP-MP between the three groups.ConclusionAs a preliminary report, the Sommerlad-Furlow modified technique without relaxing incisions was found to have a good maxillary positioning in the face and a satisfactory intermaxillary relationship compared to the Sommerlad-Furlow modified technique with relaxing incisions.

Clinical analysis of retropharyngeal lymph node metastasis in patients with hypopharyngeal squamous cell carcinoma

Objective: To analyze the incidence and the related risk factors of retropharyngeal lymph node metastasis in patients with hypopharyngeal squamous cell carcinoma, evaluate the accuracy of preoperative enhanced CT in judging retropharyngeal lymph node metastasis, and investigate the impact of retropharyngeal lymph node metastasis on the prognosis. Methods: Retrospective analyses were made on 398 patients with hypopharyngeal squamous cell carcinoma who underwent surgery as the primary therapy and accepted retropharyngeal lymph node exploration and clearance during surgery in Shandong Provincial ENT Hospital from January 2014 to December 2019. Multivariate logistic regression analysis was used to clarify the related risk factors of retropharyngeal lymph node metastasis. Multivariate Cox regression analysis was used to investigate the impact of retropharyngeal lymph node metastasis on prognosis. The retropharyngeal lymph nodes of 218 cases with available preoperative enhanced CT images were evaluated by two experienced radiologists and compared with postoperative pathological results. Results: Retropharyngeal lymph node metastasis were confirmed in 54 of 398 (13.6%) cases according to postoperative pathology. The sensitivity and specificity of preoperative enhanced CT in the diagnosis of retropharyngeal lymph node metastasis were 34.6% and 91.1%, respectively, and the overall accuracy was 84.4%. Multivariate logistic regression analysis showed that the site of the primary lesion and pathological N stage were independent risk factors for retropharyngeal lymph node metastasis in hypopharyngeal squamous cell carcinoma. Patients with primary lesion located in the posterior wall of hypopharynx (OR=4.83, 95% CI: 1.27-18.40), N2 stage (OR=6.30, 95% CI: 2.25-17.67), and N3 stage (OR=26.89, 95% CI: 5.76-125.58) were prone to retropharyngeal lymph node metastasis. The 5-year overall survival rate of the 398 patients was 50.4%, and the 5-year disease-free survival rate was 48.3%. Multivariate Cox regression analysis showed that T stage, N stage, retropharyngeal lymph node metastasis, and radiotherapy were independent influencing factors for overall survival (T stage: HR=1.28, 95% CI: 1.06-1.54; N stage: HR=1.26, 95% CI: 1.14-1.40; retropharyngeal lymph node metastasis: HR=2.13, 95% CI: 1.47-3.08; radiotherapy: HR=0.54, 95% CI: 0.38-0.76) and disease-free survival of patients with hypopharyngeal squamous cell carcinoma (T stage: HR=1.26, 95% CI: 1.06-1.51; N stage: HR=1.25, 95% CI: 1.13-1.37; retropharyngeal lymph node metastasis: HR=2.24, 95% CI: 1.56-3.21; radiotherapy: HR=0.55, 95% CI: 0.40-0.77). Conclusions: Metastasis of retropharyngeal lymph nodes in hypopharyngeal squamous cell carcinoma is not rare. Enhanced CT is of low accuracy and limited value in diagnosing retropharyngeal lymph node metastasis. Primary lesions located in the posterior wall of the hypopharyngx, N2 stage, and N3 stage are independent high-risk factors for retropharyngeal lymph node metastasis. The prognosis of hypopharyngeal cancer patients with retropharyngeal lymph node metastasis is worse, and active surgical exploration and clearance can effectively reduce the mortality caused by retropharyngeal lymph node metastasis.

Artificial Intelligence Applied to Chest X-ray: A Reliable Tool to Assess the Differential Diagnosis of Lung Pneumonia in the Emergency Department.

Considering the large number of patients with pulmonary symptoms admitted to the emergency department daily, it is essential to diagnose them correctly. It is necessary to quickly solve the differential diagnosis between COVID-19 and typical bacterial pneumonia to address them with the best management possible. In this setting, an artificial intelligence (AI) system can help radiologists detect pneumonia more quickly. We aimed to test the diagnostic performance of an AI system in detecting COVID-19 pneumonia and typical bacterial pneumonia in patients who underwent a chest X-ray (CXR) and were admitted to the emergency department. The final dataset was composed of three sub-datasets: the first included all patients positive for COVID-19 pneumonia (n = 1140, namely "COVID-19+"), the second one included all patients with typical bacterial pneumonia (n = 500, "pneumonia+"), and the third one was composed of healthy subjects (n = 1000). Two radiologists were blinded to demographic, clinical, and laboratory data. The developed AI system was used to evaluate all CXRs randomly and was asked to classify them into three classes. Cohen's κ was used for interrater reliability analysis. The AI system's diagnostic accuracy was evaluated using a confusion matrix, and 95%CIs were reported as appropriate. The interrater reliability analysis between the most experienced radiologist and the AI system reported an almost perfect agreement for COVID-19+ (κ = 0.822) and pneumonia+ (κ = 0.913). We found 96% sensitivity (95% CIs = 94.9-96.9) and 79.8% specificity (76.4-82.9) for the radiologist and 94.7% sensitivity (93.4-95.8) and 80.2% specificity (76.9-83.2) for the AI system in the detection of COVID-19+. Moreover, we found 97.9% sensitivity (98-99.3) and 88% specificity (83.5-91.7) for the radiologist and 97.5% sensitivity (96.5-98.3) and 83.9% specificity (79-87.9) for the AI system in the detection of pneumonia+ patients. Finally, the AI system reached an accuracy of 93.8%, with a misclassification rate of 6.2% and weighted-F1 of 93.8% in detecting COVID+, pneumonia+, and healthy subjects. The AI system demonstrated excellent diagnostic performance in identifying COVID-19 and typical bacterial pneumonia in CXRs acquired in the emergency setting.

Final-year medical students’ perspective: a survey on the use of computed tomography in sepsis

ObjectivesTo determine the perspective of final-year medical students on the use of computed tomography (CT) in patients with sepsis.MethodsA total of 207 questionnaires were distributed to final-year medical students at a large university medical center, and 113 returned questionnaires met the criteria for inclusion in the analysis. Questions referred to sepsis guidelines, CT indications, and the use of contrast agents. Control variables included a level of practical experience as a final-year student (trimester of student’s practical year) and previous radiological experience. Statistical hypothesis tests such as the Mann-Whitney U test and chi-square test were performed.ResultsThe majority of participating students, 85% (n = 91/107), considered a Systemic Organ Failure Assessment (SOFA) score ≥ 2 as a diagnostic criterion for sepsis. The presence of ≥ 2 positive systemic inflammatory response syndrome (SIRS) criteria was considered relevant for diagnosing sepsis by 34% (n = 34/100). Ninety-nine percent (n = 64/65) of the participants who fully agreed with a SOFA score ≥ 2 being relevant for diagnosing sepsis would also use it as an indication for a CT scan. Seventy-six percent (n = 78/103) of the students rated a known severe allergic reaction to contrast agents as an absolute contraindication for its administration. Ninety-five percent (n = 78/82) considered radiation exposure as problematic in CT examinations, especially in repeat CTs.ConclusionMost final-year medical students were familiar with the sepsis criteria. Still, some referred to outdated diagnostic criteria. Participants saw the ability to plan further patient management based on CT as a major benefit. Most participants were aware of radiation as a risk of CT.Critical relevance statementMore detailed knowledge of CT in septic patients should be implemented in the medical curriculum. Retraining of medical students could help increase student confidence potentially improving patient care.Key points1. Whereas the majority of final-year medical students were familiar with sepsis criteria, some referred to outdated diagnostic criteria.2. Participants saw the ability to plan further patient management based on CT as a major benefit.3. Most participants were aware of radiation as a risk of CT.Graphical

Perceptions of work-life balance and gender issues in private radiology in Australia.

Women comprise almost 30% of practising clinical radiologists in Australia. Despite three-quarters of clinical radiologists working in private practice, there is limited research regarding the experience of female radiologists in private practice and issues surrounding work-life balance. The primary aim of this study was to identify gender issues within private radiology and investigate work-life balance, focusing on issues relevant for female practitioners. Clinical radiologists working in private practice at five major Australian private radiology service providers were sent an email on behalf of the researchers, containing study information and a link to an online questionnaire. Descriptive statistics, chi-square analysis and Fisher's exact test were used. Binary logistic regression was used to determine odds ratios and 95% confidence intervals. Responses to short-answer questions were analysed manually and grouped into themes. There were consistent gender differences in perceptions of gender and its implication upon working and career progression in radiology private practice. There was near unanimous agreement that private radiologists should be able to work flexible/reduced working hours for a proportionate decrease in pay. Two themes with distinct gender differences related to pay equity and transparency. Key themes relating to work-life balance for private radiologists included workload, setting boundaries and flexibility of work hours in private radiology. This preliminary survey highlights areas related to work-life balance and gender issues relevant for female practitioners in private practice radiology in Australia that can be targeted for further investigation and improvement, namely pay equity and transparency, and part-time leadership opportunities.

MRI-based automated multitask deep learning system to evaluate supraspinatus tendon injuries.

To establish an automated, multitask, MRI-based deep learning system for the detailed evaluation of supraspinatus tendon (SST) injuries. According to arthroscopy findings, 3087 patients were divided into normal, degenerative, and tear groups (groups 0-2). Group 2 was further divided into bursal-side, articular-side, intratendinous, and full-thickness tear groups (groups 2.1-2.4), and external validation was performed with 573 patients. Visual geometry group network 16 (VGG16) was used for preliminary image screening. Then, the rotator cuff multitask learning (RC-MTL) model performed multitask classification (classifiers 1-4). A multistage decision model produced the final output. Model performance was evaluated by receiver operating characteristic (ROC) curve analysis and calculation of related parameters. McNemar's test was used to compare the differences in the diagnostic effects between radiologists and the model. The intraclass correlation coefficient (ICC) was used to assess the radiologists' reliability. p < 0.05 indicated statistical significance. In the in-group dataset, the area under the ROC curve (AUC) of VGG16 was 0.92, and the average AUCs of RC-MTL classifiers 1-4 were 0.99, 0.98, 0.97, and 0.97, respectively. The average AUC of the automated multitask deep learning system for groups 0-2.4 was 0.98 and 0.97 in the in-group and out-group datasets, respectively. The ICCs of the radiologists were 0.97-0.99. The automated multitask deep learning system outperformed the radiologists in classifying groups 0-2.4 in both the in-group and out-group datasets (p < 0.001). The MRI-based automated multitask deep learning system performed well in diagnosing SST injuries and is comparable to experienced radiologists. Our study established an automated multitask deep learning system to evaluate supraspinatus tendon (SST) injuries and further determine the location of SST tears. The model can potentially improve radiologists' diagnostic efficiency, reduce diagnostic variability, and accurately assess SST injuries. • A detailed classification of supraspinatus tendon tears can help clinical decision-making. • Deep learning enables the detailed classification of supraspinatus tendon injuries. • The proposed automated multitask deep learning system is comparable to radiologists.

Ultrasonographic assessment of magnetic growing rods overestimates the lengthening of the thoracic spine compared to radiographs in early-onset scoliotic patients.

Magnetic growing rods (MGRs) are one of the most common procedures to treat early-onset scoliosis (EOS). Radiographic examinations (X-ray) or ultrasonographic (US) assessments are used to evaluate the lengthening of the rods. X-ray exposes patients to radiation, while the US has not been validated and may be affected by the radiologist's ability to assess elongation. The research question of the present study is to compare the difference between US and X-ray growth assessments in EOS patients treated with MGRs. The study enrolled 65 patients consecutively from July 2011 to July 2022. Noninvasive lengthening was performed every fourmonths, and X-ray follow-up was performed at different intervals. An experienced radiologist assessed the mean US rod elongation per session. The mean elongation/session of T2-T12 and T2-S1 was calculated, and the results were compared using an independent t-test. The mean age at operation was 8.8 ± 2years, and the mean follow-up was four± twoyears. The average rod elongation assessed by the US was 3.1 ± 0.1mm. The average rod elongation evaluated by X-ray was 1.2 ± 2.9mm (T2-T12) and 1.8 ± 1.9mm (T2-S1). The difference between the values measured by US and X-ray was statistically significant in the T2-T12 group (p < 0.05) and not significant in the T2-S1 group (p = 0.34). This is the most extensive single-center study comparing US and X-ray data for MGRs in EOS patients. US overestimates thoracic spine elongation compared to X-ray. US elongation analysis could be appropriate in long thoracolumbar curves.

Evaluation of imaging findings in gastrointestinal anisakiasis in emergency CT and ultrasound

BackgroundTo assess the frequency of appearance of various signs (isolated and grouped) in emergency imaging tests in patients with anisakiasis, according to the location of gastrointestinal tract involvement.MethodsRetrospective review by two experienced radiologists of emergency ultrasounds and CTs performed on patients admitted in the Emergency Department of our hospital with later confirmed anisakiasis (2010–2021), assessing the presence of signs suggesting anisakiasis. Calculation of the frequency of appearance according to the gastric or intestinal location, as well as the most common grouped signs.ResultsOut of 231 total patients with anisakiasis, imaging studies were performed in 144: abdominopelvic ultrasound in 43 cases and CT in 111 (both techniques in 31). In cases with gastric occurrence (34), in CT the wall stratification (100%), wall thickening (97%), fat stranding (91%) and ascitic fluid (82%) were predominant. In the intestinal cases (105), in CT (95) the wall thickening (100%), fat stranding (92%) and mesenteric vessel engorgement (83%) were usual; in ultrasound (40), ascitic fluid and wall thickening (70% in both cases) were frequently observed. The frequency of grouped appearance of the mentioned signs was 82% in gastric cases, 80% in intestinal cases and 50% in ultrasounds. Multisegment involvement in CT reached 28% (gastric + intestinal) and 11% (only intestinal) of cases.ConclusionsThe most frequent CT findings in patients with gastric anisakiasis are wall stratification and thickening, fat stranding and ascitic fluid. In the intestinal cases, wall thickening, fat oedema and vessel engorgement are the most often observed findings.Critical relevance statementThe presence of different radiological signs makes it advisable to include anisakiasis in the differential diagnosis of acute abdomen. Intestinal and multifocal involvement rates are greater than previously reported.Key points• In gastric anisakiasis, CT frequently shows wall stratification and thickening, fat stranding and ascitic fluid.• In intestinal anisakiasis, CT often presents wall thickening, fat stranding and vessel engorgement.• In intestinal anisakiasis, ultrasounds most frequently show ascitic fluid and wall thickening.Graphical