OBJECTIVES: To explore the synthesis of high-quality CT (sCT) from cone-beam CT (CBCT) using PE-CycleGAN for adaptive radiotherapy (ART) for nasopharyngeal carcinoma. METHODS: A perception-enhanced CycleGAN model "PE-CycleGAN" was proposed, introducing dual-contrast discriminator loss, multi-perceptual generator loss, and improved U-Net structure. CBCT and CT data from 80 nasopharyngeal carcinoma patients were used as the training set, with 7 cases as the test set. By quantifying the mean absolute error (MAE), peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), as well as the dose gamma pass rate and the relative dose deviations of the target area and organs at risk (OAR) between sCT and reference CT, the image quality and dose calculation accuracy of sCT were evaluated. RESULTS: The MAE of sCT generated by PE-CycleGAN compared to the reference CT was (56.89±13.84) HU, approximately 30% lower than CBCT's (81.06±15.86) HU (P<0.001). PE-CycleGAN's PSNR and SSIM were 26.69±2.41dB and 0.92±0.02 respectively, significantly higher than CBCT's 21.54±2.37dB and 0.86±0.05 (P<0.001), indicating substantial improvements in image quality and structural similarity. In gamma analysis, under the 2 mm/2% criterion, PE-CycleGAN's sCT achieved a pass rate of (90.13±3.75)%, significantly higher than CBCT's (81.65±3.92)% (P<0.001) and CycleGAN's (87.69±3.50)% (P<0.05). Under the 3 mm/3% criterion, PE-CycleGAN's sCT pass rate of (90.13±3.75)% was also significantly superior to CBCT's (86.92±3.51)% (P<0.001) and CycleGAN's (94.58±2.23)% (P<0.01). The mean relative dose deviation of the target area and OAR between sCT and planned CT was within ±3% for all regions, except for the Lens Dmax (Gy), which had a deviation of 3.38% (P=0.09). The mean relative dose deviations for PTVnx HI, PTVnd HI, PTVnd CI, PTV1 HI, PRV_SC, PRV_BS, Parotid, Larynx, Oral, Mandible, and PRV_ON were all less than ±1% (P>0.05). CONCLUSIONS PE-CycleGAN demonstrates the ability to rapidly synthesize high-quality sCT from CBCT, offering a promising approach for CBCT-guided adaptive radiotherapy in nasopharyngeal carcinoma.
Humans ; Cone-Beam Computed Tomography/methods* ; Nasopharyngeal Neoplasms/diagnostic imaging* ; Nasopharyngeal Carcinoma/radiotherapy* ; Radiotherapy Planning, Computer-Assisted/methods* ; Radiotherapy Dosage ; Signal-To-Noise Ratio ; Radiotherapy, Intensity-Modulated

OBJECTIVES: To analyze bone mass distribution and the factors affecting bone mass in a general Chinese Han cohort undergoing physical examinations at our center. METHODS: We retrospectively collected the data of bone mineral density (BMD) measurements from 30 599 healthy Han Chinese adults (age≥20 years) who underwent dual-energy X-ray absorptiometry scans at our hospital from July, 2013 to July, 2023. Basic parameters including height, body weight, and gender were recorded, and descriptive statistics and correlation analyses were performed using R software. RESULTS: In this cohort, the male individuals had a mean peak BMD of 1.00±0.12 g/cm² in the lumbar vertebrae, 0.94±0.14 g/cm² in the femoral neck, and 0.99±0.13 g/cm² in the total hip, significantly higher than the values in the female individuals [0.99±0.12 g/cm² in the lumbar vertebrae (P=0.022), 0.79±0.11 g/cm² in the femoral neck (P<0.001), and 0.88±0.11 g/cm² in the total hip (P<0.001)]. In the overall cohort, the BMD values of the lumbar spine and femur decreased with age after reaching their peak levels. There was a positive correlation between BMD value and body mass index (BMI) in both male and female individuals. The 2013-2014 period recorded the lowest BMD values in the lumbar, hip, and femoral neck, which tended to increase steadily in the following years (2015-2023). CONCLUSIONS Our data suggest that the BMD values vary among different populations, and future multi-center studies using more accurate BMD detection technology are warranted to capture the variation patterns of BMD with demographic characteristics of specific populations.
Humans ; Bone Density ; Absorptiometry, Photon ; Male ; Female ; Retrospective Studies ; Osteoporosis/diagnostic imaging* ; Adult ; Middle Aged ; Lumbar Vertebrae/diagnostic imaging* ; China/epidemiology* ; Femur Neck/diagnostic imaging* ; Aged ; Beijing/epidemiology* ; Young Adult

OBJECTIVES: We propose a novel deep learning segmentation algorithm (DSRF) based on multi-scale supervision and residual feedback strategy for precise segmentation of the optic chiasm and optic nerves in CT images of nasopharyngeal carcinoma (NPC) patients. METHODS: We collected 212 NPC CT images and their ground truth labels from SegRap2023, StructSeg2019 and HaN-Seg2023 datasets. Based on a hybrid pooling strategy, we designed a decoder (HPS) to reduce small organ feature loss during pooling in convolutional neural networks. This decoder uses adaptive and average pooling to refine high-level semantic features, which are integrated with primary semantic features to enable network learning of finer feature details. We employed multi-scale deep supervision layers to learn rich multi-scale and multi-level semantic features under deep supervision, thereby enhancing boundary identification of the optic chiasm and optic nerves. A residual feedback module that enables multiple iterations of the network was designed for contrast enhancement of the optic chiasm and optic nerves in CT images by utilizing information from fuzzy boundaries and easily confused regions to iteratively refine segmentation results under supervision. The entire segmentation framework was optimized with the loss from each iteration to enhance segmentation accuracy and boundary clarity. Ablation experiments and comparative experiments were conducted to evaluate the effectiveness of each component and the performance of the proposed model. RESULTS: The DSRF algorithm could effectively enhance feature representation of small organs to achieve accurate segmentation of the optic chiasm and optic nerves with an average DSC of 0.837 and an ASSD of 0.351. Ablation experiments further verified the contributions of each component in the DSRF method. CONCLUSIONS The proposed deep learning segmentation algorithm can effectively enhance feature representation to achieve accurate segmentation of the optic chiasm and optic nerves in CT images of NPC.
Humans ; Tomography, X-Ray Computed/methods* ; Optic Chiasm/diagnostic imaging* ; Optic Nerve/diagnostic imaging* ; Algorithms ; Nasopharyngeal Carcinoma ; Deep Learning ; Nasopharyngeal Neoplasms/diagnostic imaging* ; Neural Networks, Computer ; Image Processing, Computer-Assisted/methods*

OBJECTIVES: To develop a risk prediction model for left ventricular adverse remodeling (LVAR) based on cardiac magnetic resonance (CMR) parameters in patients undergoing percutaneous coronary intervention (PCI) for acute ST-segment elevation myocardial infarction (STEMI). METHODS: A total of 329 acute STEMI patients undergoing primary PCI at 8 medical centers from January, 2018 to December, 2021 were prospectively enrolled. The parameters of CMR, performed at 7±2 days and 6 months post-PCI, were analyzed using CVI42 software. LVAR was defined as an increase >20% in left ventricular end-diastolic volume or >15% in left ventricular end-systolic volume at 6 months compared to baseline. The patients were randomized into training (n=230) and validation (n=99) sets in a 7∶3 ratio. In the training set, potential predictors were selected using LASSO regression, followed by univariate and multivariate logistic regression to construct a nomogram. Model performance was evaluated using receiver-operating characteristic (ROC) curves, area under the curve (AUC), calibration curves, and decision curve analysis. RESULTS: LVAR occurred in 100 patients (30.40%), who had a higher incidence of major adverse cardiovascular events than those without LVAR (58.00% vs 16.16%, P<0.001). Left ventricular global longitudinal strain (LVGLS; OR=0.76, 95% CI: 0.61-0.95, P=0.015) and left atrial active strain (LAAS; OR=0.78, 95% CI: 0.67-0.92, P=0.003) were protective factors for LVAR, while infarct size (IS; OR=1.05, 95% CI: 1.01-1.10, P=0.017) and microvascular obstruction (MVO; OR=1.26, 95% CI: 1.01-1.59, P=0.048) were risk factors for LVAR. The nomogram had an AUC of 0.90 (95% CI: 0.86-0.94) in the training set and an AUC of 0.88 (95% CI: 0.81-0.94) in the validation set. CONCLUSIONS LVGLS, LAAS, IS, and MVO are independent predictors of LVAR in STEMI patients following PCI. The constructed nomogram has a strong predictive ability to provide assistance for management and early intervention of LVAR.
Humans ; Percutaneous Coronary Intervention ; Prospective Studies ; ST Elevation Myocardial Infarction/diagnostic imaging* ; Ventricular Remodeling ; Magnetic Resonance Imaging ; Male ; Female ; Middle Aged ; Risk Factors ; Aged ; Risk Assessment

BACKGROUND AND OBJECTIVES

The Philippines faces challenges in the screening of tuberculosis (TB), one of them being the shortage in the health workforce who are skilled and allowed to screen TB. Deep learning neural networks (DLNNs) have shown potential in the TB screening process utilizing chest radiographs (CXRs). However, local studies on AIbased TB screening are limited. This study evaluated qXR3.0 technology's diagnostic performance for TB screening in Filipino adults aged 15 and older. Specifically, we evaluated the specificity and sensitivity of qXR3.0 compared to radiologists' impressions and determined whether it meets the World Health Organization (WHO) standards.

A prospective cohort design was used to perform a study on comparing screening and diagnostic accuracies of qXR3.0 and two radiologist gradings in accordance with the Standards for Reporting Diagnostic Accuracy (STARD). Subjects from two clinics in Metro Manila which had qXR 3.0 seeking consultation at the time of study were invited to participate to have CXRs and sputum collected. Radiologists' and qXR3.0 readings and impressions were compared with respect to the reference standard Xpert MTB/RiF assay. Diagnostic accuracy measures were calculated.

With 82 participants, qXR3.0 demonstrated 100% sensitivity and 72.7% specificity with respect to the reference standard. There was a strong agreement between qXR3.0 and radiologists' readings as exhibited by the 0.7895 (between qXR 3.0 and CXRs read by at least one radiologist), 0.9362 (qXR 3.0 and CXRs read by both radiologists), and 0.9403 (qXR 3.0 and CXRs read as not suggestive of TB by at least one radiologist) concordance indices.

qXR3.0 demonstrated high sensitivity to identify presence of TB among patients, and meets the WHO standard of at least 70% specificity for detecting true TB infection. This shows an immense potential for the tool to supplement the shortage of radiologists for TB screening in the country. Future research directions may consider larger sample sizes to confirm these findings and explore the economic value of mainstream adoption of qXR 3.0 for TB screening.

INTRODUCTION: Medical errors commonly occur in medical imaging departments. These errors are frequently influenced by patient safety culture. This study aimed to develop a suitable patient safety culture assessment tool for medical imaging departments. METHODS: Staff members of a teaching hospital medical imaging department were invited to complete the generic short version of the Safety Attitude Questionnaire (SAQ). Internal consistency and reliability were evaluated using Cronbach's α. Confirmatory factor analysis (CFA) was conducted to examine model fit. A cut-off of 60% was used to define the percentage positive responses (PPR). PPR values were compared between occupational groups. RESULTS: A total of 300 complete responses were received and the response rate was 75.4%. In reliability analysis, the Cronbach's α for the original 32-item SAQ was 0.941. Six subscales did not demonstrate good fit with CFA. A modified five-subscale, 22-item model (SAQ-MI) showed better fit (goodness-to-fit index ≥0.9, comparative fit index ≥ 0.9, Tucker-Lewis index ≥0.9 and root mean square error of approximation ≤0.08). The Cronbach's α for the 22 items was 0.921. The final five subscales were safety and teamwork climate, job satisfaction, stress recognition, perception of management and working condition, with PPR of 62%, 68%, 57%, 61% and 60%, respectively. Statistically significant differences in PPR were observed between radiographers, doctors and others occupational groups. CONCLUSION The modified five-factor, 22-item SAQ-MI is a suitable tool for the evaluation of patient safety culture in a medical imaging department. Differences in patient safety culture exist between occupation groups, which will inform future intervention studies.
Humans ; Surveys and Questionnaires ; Patient Safety ; Attitude of Health Personnel ; Diagnostic Imaging ; Reproducibility of Results ; Male ; Female ; Adult ; Job Satisfaction ; Factor Analysis, Statistical ; Middle Aged ; Hospitals, Teaching ; Safety Management ; Organizational Culture ; Medical Errors/prevention & control*

INTRODUCTION: Foreign body (FB) ingestion is a common paediatric emergency. While guidelines exist for urgent intervention, less is known of the natural progress of FBs passing through the gastrointestinal tract (GIT). We reviewed these FB transit times in an outpatient cohort. METHODS: A retrospective review was performed on all children (≤18 years) treated for radiopaque FB ingestion at two major tertiary paediatric centres from 2015 to 2016. Demographic data, FB types, outcomes and hospital visits (emergency department [ED] and outpatient) were recorded. All cases discharged from the ED with outpatient follow-up were included. We excluded those who were not given follow-up appointments and those admitted to inpatient wards. We categorised the outcomes into confirmed passage (ascertained via abdominal X-ray or reported direct stool visualisation by patients/caregivers) and assumed passage (if patients did not attend follow-up appointments). RESULTS: Of the 2,122 ED visits for FB ingestion, 350 patients who were given outpatient follow-up appointments were reviewed (median age 4.35 years [range: 0.5-14.7], 196 [56%] male). The largest proportion (16%) was aged 1-2 years. Coins were the most common ingested FB, followed by toys. High-risk FB (magnets or batteries) formed 9% of cases ( n =33). The 50 th centile for FB retention was 8, 4 and 7 days for coins, batteries and other radiopaque FBs, respectively; all confirmed passages occurred at 37, 7 and 23 days, respectively. Overall, 197 (68%) patients defaulted on their last given follow-up. CONCLUSION This study provides insight into the transit times of FB ingested by children, which helps medical professionals to decide on the optimal time for follow-up visits and provide appropriate counsel to caregivers.
Adolescent ; Child ; Child, Preschool ; Female ; Humans ; Infant ; Male ; Eating ; Emergency Service, Hospital ; Foreign Bodies/diagnostic imaging* ; Gastrointestinal Tract/diagnostic imaging* ; Gastrointestinal Transit ; Retrospective Studies ; Singapore ; Tertiary Care Centers

INTRODUCTION: This study was conducted to evaluate the efficacy of postoperative computed tomography (CT) measurements of aortic lumen volumes in predicting aortic-related complications following acute type A aortic dissection (ATAAD) repair. METHODS: We conducted a single-institution retrospective aortic volumetric analysis of patients after ascending aorta replacement performed during 2001-2015. The volumetric measurements of total lumen (total-L), true lumen (TL), false lumen (FL), as well as the TL:FL ratio from the first and second postoperative computer angiograms were obtained. A generalised structural equation model was created to analyse the predictive utility of TL:FL ratio. RESULTS: One hundred and twenty-five patients underwent surgical intervention, of whom 97 patients were eventually discharged and analysed for postoperative complications. A total of 19 patients were included in the final analysis. Patients with late postoperative aortic complications had a significantly higher FL volume and total-L volume on the first (FL volume P = 0.041, total-L volume P = 0.05) and second (FL volume P = 0.01, total-L volume P = 0.007) postoperative scans. The odds of having aortic complications were raised by 1% with a 1 cm 3 increase in total-L volume and by 2% with a 1 cm 3 increase in FL volume. The TL:FL ratio was significantly lower in patients who developed complications. CONCLUSION Postoperative CT volumetric measurements in patients who developed complications are characterised by a significant increase in the FL volume and total-L volume from the first postoperative scans. Patients with disproportionately expanded FL presenting with TL:FL ratios less than 1 were associated with aortic complications. Hence, the TL:FL ratio may be a reliable and useful parameter to monitor postoperative disease progression and to evaluate the risk of late complications in ATAAD patients.
Humans ; Male ; Female ; Retrospective Studies ; Singapore ; Aortic Dissection/diagnostic imaging* ; Middle Aged ; Postoperative Complications/diagnostic imaging* ; Aged ; Tomography, X-Ray Computed ; Aortic Aneurysm/diagnostic imaging* ; Aorta/surgery* ; Adult ; Treatment Outcome ; Computed Tomography Angiography

INTRODUCTION: Radiology plays an integral role in fracture detection in the emergency department (ED). After hours, when there are fewer reporting radiologists, most radiographs are interpreted by ED physicians. A minority of these interpretations may miss diagnoses, which later require the callback of patients for further management. Artificial intelligence (AI) has been viewed as a potential solution to augment the shortage of radiologists after hours. We explored the efficacy of an AI solution in the detection of appendicular and pelvic fractures for adult radiographs performed after hours at a general hospital ED in Singapore, and estimated the potential monetary and non-monetary benefits. METHODS: One hundred and fifty anonymised abnormal radiographs were retrospectively collected and fed through an AI fracture detection solution. The radiographs were re-read by two radiologist reviewers and their consensus was established as the reference standard. Cases were stratified based on the concordance between the AI solution and the reviewers' findings. Discordant cases were further analysed based on the nature of the discrepancy into overcall and undercall subgroups. Statistical analysis was performed to evaluate the accuracy, sensitivity and inter-rater reliability of the AI solution. RESULTS: Ninety-two examinations were included in the final study radiograph set. The AI solution had a sensitivity of 98.9%, an accuracy of 85.9% and an almost perfect agreement with the reference standard. CONCLUSION An AI fracture detection solution has similar sensitivity to human radiologists in the detection of fractures on ED appendicular and pelvic radiographs. Its implementation offers significant potential measurable cost, manpower and time savings.
Humans ; Singapore ; Emergency Service, Hospital ; Fractures, Bone/diagnostic imaging* ; Artificial Intelligence ; Retrospective Studies ; Adult ; Male ; Female ; Cost Savings ; Middle Aged ; Pelvic Bones/diagnostic imaging* ; Reproducibility of Results ; Aged ; Sensitivity and Specificity ; Radiography

INTRODUCTION: In this study, we aimed to compare the performance of a convolutional neural network (CNN)-based deep learning model that was trained on a dataset of normal and abnormal paediatric elbow radiographs with that of paediatric emergency department (ED) physicians on a binomial classification task. METHODS: A total of 1,314 paediatric elbow lateral radiographs (patient mean age 8.2 years) were retrospectively retrieved and classified based on annotation as normal or abnormal (with pathology). They were then randomly partitioned to a development set (993 images); first and second tuning (validation) sets (109 and 100 images, respectively); and a test set (112 images). An artificial intelligence (AI) model was trained on the development set using the EfficientNet B1 network architecture. Its performance on the test set was compared to that of five physicians (inter-rater agreement: fair). Performance of the AI model and the physician group was tested using McNemar test. RESULTS: The accuracy of the AI model on the test set was 80.4% (95% confidence interval [CI] 71.8%-87.3%), and the area under the receiver operating characteristic curve (AUROC) was 0.872 (95% CI 0.831-0.947). The performance of the AI model vs. the physician group on the test set was: sensitivity 79.0% (95% CI: 68.4%-89.5%) vs. 64.9% (95% CI: 52.5%-77.3%; P = 0.088); and specificity 81.8% (95% CI: 71.6%-92.0%) vs. 87.3% (95% CI: 78.5%-96.1%; P = 0.439). CONCLUSION The AI model showed good AUROC values and higher sensitivity, with the P-value at nominal significance when compared to the clinician group.
Humans ; Deep Learning ; Child ; Retrospective Studies ; Male ; Female ; Radiography/methods* ; ROC Curve ; Elbow/diagnostic imaging* ; Neural Networks, Computer ; Child, Preschool ; Elbow Joint/diagnostic imaging* ; Emergency Service, Hospital ; Adolescent ; Infant ; Artificial Intelligence