Non-contrast-enhanced Computed Tomography Research Articles

RFLSE: Joint radiomics feature‐enhanced level‐set segmentation for low‐contrast SPECT/CT tumour images

AbstractDoctors typically use non‐contrast‐enhanced computed tomography (NCECT) in the treatment of kidney cancer to map kidney and tumour structural information to functional imaging single‐photon emission computed tomography, which is then used to assess patient kidney function and predict postoperative recovery. However, the assessment of kidney function and formulation of surgical plans is constrained by the low contrast of tumours in NCECT, which hinders the acquisition of accurate tumour boundaries. Therefore, this study designed a radiomics feature‐enhanced level‐set evolution (RFLSE) to precisely segment small‐sample low‐contrast kidney tumours. Integration of high‐dimensional radiomics features into the level‐set energy function enhances the edge detection capability of low‐contrast kidney tumours. The use of sensitive radiomics features to control the regional term parameters achieves adaptive adjustment of the curve evolution amplitude, improving the level‐set segmentation process. The experimental data used low‐contrast, limited‐sample tumours provided by hospitals, as well as the public datasets BUSI18 and KiTS19. Comparative results with advanced energy functionals and deep learning models demonstrate the precision and robustness of RFLSE segmentation. Additionally, the application value of RFLSE in assisting doctors with accurately marking tumours and generating high‐quality pseudo‐labels for deep learning datasets is demonstrated.

The role of CT brain findings in the early diagnosis of infantile encephalitic beriberi.

Thiamine deficiency disease may occur in infants from thiamine-deficient mothers in developing countries, as well as in infants fed solely with soy-based formula. Thiamine deficiency in infants may present with acute neurological manifestations of infantile encephalitic beriberi. To review the role of noncontrast CT brain findings in infantile encephalitic beriberi in early diagnosis. A retrospective review of noncontrast CT scans of the brain in 21 infants with acute-onset infantile encephalitic beriberi was carried out. On noncontrast-enhanced CT brain, hypodense lesions were seen symmetrically in the putamen in all the babies; symmetric hypodensities were seen in the caudate nuclei in 14/21 (67%), in dorsomedial thalami/hypothalamic/subthalamic area in 4/21 (19%), and in the globi pallidi in 2/21 (9.5%) of the infants. Recognition of symmetrical hypodense lesions in the basal ganglia and medial thalami/hypothalamic/subthalamic area on noncontrast CT scan of the brain are important early features to recognize in encephalitic beriberi in at-risk infants. IEBB is a cause of hypodense bilateral basal ganglia and may be identified by this finding in the appropriate clinical settings.

The ratio of the maximum density values: a new method for predicting hemorrhagic transformation in acute ischemic stroke patients undergoing mechanical thrombectomy.

It is challenging yet critical to differentiate between hemorrhagic transformation (HT) and contrast extravasation on non-contrast-enhanced computed tomography (NCCT) scans following mechanical thrombectomy (MT) in patients with acute ischemic stroke. We propose a new method called the ratio of maximum density values (RMDV) to minimize the confusion of contrast extravasation and to evaluate the diagnostic significance of RMDV in predicting HT on immediate post-interventional NCCT scans. We conducted a retrospective analysis of the prospective patients' database who received MT for acute ischemic stroke caused by occlusion of the intracranial large artery and showed postinterventional cerebral hyperdensities (PCHDs) on NCCT scans immediately after MT. Based on the subsequent NCCT scans, we divided patients with PCHDs into the HT and the non-HT groups. The clinical characters and radiological details were collected and compared to the two groups. We assessed the ability of RMDV >1 to predict HT by analyzing the receiver operating characteristic curve. One hundred and three patients showed PCHDs; 58 (56.31%) were classified as HT, while 45 (43.69%) were classified as non-HT. The only notable distinction between the two groups was the proportion of RMDV >1 in the HT group. The correlation between HT and RMDV >1 with an area under the curve of 0.826 (95% confidence interval, 0.739 to 0.894). The sensitivity, specificity, positive, and negative predictive values of RMDV >1 on NCCT for predicting HT were 89.66, 75.56, 82.54, and 85.00%, respectively. The utilization of RMDV >1 on immediate NCCT scans after MT can predict early HT with good sensitivity and specificity.

A deep learning approach to remove contrast from contrast-enhanced CT for proton dose calculation.

Non-Contrast Enhanced CT (NCECT) is normally required for proton dose calculation while Contrast Enhanced CT (CECT) is often scanned for tumor and organ delineation. Possible tissue motion between these two CTs raises dosimetry uncertainties, especially for moving tumors in the thorax and abdomen. Here we report a deep-learning approach to generate NCECT directly from CECT. This method could be useful to avoid the NCECT scan, reduce CT simulation time and imaging dose, and decrease the uncertainties caused by tissue motion between otherwise two different CT scans. A deep network was developed to convert CECT to NCECT. The network receives a 3D image from CECT images as input and generates a corresponding contrast-removed NCECT image patch. Abdominal CECT and NCECT image pairs of 20 patients were deformably registered and 8000 image patch pairs extracted from the registered image pairs were utilized to train and test the model. CTs of clinical proton patients and their treatment plans were employed to evaluate the dosimetric impact of using the generated NCECT for proton dose calculation. Our approach achieved a Cosine Similarity score of 0.988 and an MSE value of 0.002. A quantitative comparison of clinical proton dose plans computed on the CECT and the generated NCECT for five proton patients revealed significant dose differences at the distal of beam paths. V100% of PTV and GTV changed by 3.5% and 5.5%, respectively. The mean HU difference for all five patients between the generated and the scanned NCECTs was ∼4.72, whereas the difference between CECT and the scanned NCECT was ∼64.52, indicating a ∼93% reduction in mean HU difference. A deep learning approach was developed to generate NCECTs from CECTs. This approach could be useful for the proton dose calculation to reduce uncertainties caused by tissue motion between CECT and NCECT.

Comparison of preoperative CT- and MRI-based multiparametric radiomics in the prediction of lymph node metastasis in rectal cancer

ObjectiveTo compare computed tomography (CT)- and magnetic resonance imaging (MRI)-based multiparametric radiomics models and validate a multi-modality, multiparametric clinical-radiomics nomogram for individual preoperative prediction of lymph node metastasis (LNM) in rectal cancer (RC) patients.Methods234 rectal adenocarcinoma patients from our retrospective study cohort were randomly selected as the training (n = 164) and testing (n = 70) cohorts. The radiomics features of the primary tumor were extracted from the non-contrast enhanced computed tomography (NCE-CT), the enhanced computed tomography (CE-CT), the T2-weighted imaging (T2WI) and the gadolinium contrast-enhanced T1-weighted imaging (CE-TIWI) of each patient. Three kinds of models were constructed based on training cohort, including the Clinical model (based on the clinical features), the radiomics models (based on NCE-CT, CE-CT, T2WI, CE-T1WI, CT, MRI, CT combing with MRI) and the clinical-radiomics models (based on CT or MRI radiomics model combing with clinical data) and Clinical-IMG model (based on CT and MRI radiomics model combing with clinical data). The performances of the 11 models were evaluated via the area under the receiver operator characteristic curve (AUC), accuracy, sensitivity, and specificity in the training and validation cohort. Differences in the AUCs among the 11 models were compared using DeLong’s test. Finally, the optimal model (Clinical-IMG model) was selected to create a radiomics nomogram. The performance of the nomogram to evaluate clinical efficacy was verified by ROC curves and decision curve analysis (DCA).ResultsThe MRI radiomics model in the validation cohort significantly outperformed than CT radiomics model (AUC, 0.785 vs. 0.721, p<0.05). The Clinical-IMG nomogram had the highest prediction efficiency than all other predictive models (p<0.05), of which the AUC was 0.947, the sensitivity was 0.870 and the specificity was 0.884.ConclusionMRI radiomics model performed better than both CT radiomics model and Clinical model in predicting LNM of RC. The clinical-radiomics nomogram that combines the radiomics features obtained from both CT and MRI along with preoperative clinical characteristics exhibits the best diagnostic performance.

Hematuria: Is it useful in predicting renal or ureteral stones in patient presenting to emergency department with flank pain?

The objective of the study was to evaluate hematuria as a diagnostic test for renal and ureteral stones compared with a noncontrast-enhanced computed tomography (CT) scan (gold standard test) in emergency room patients with acute flank pain. In total, 604 patients treated in our emergency department from 2006 to 2011, with a history of flank pain and suspected urolithiasis were included in a retrospective review. All patients were evaluated with a noncontrast-enhanced CT scan and urine analysis. Using the noncontrast CT scan as the gold standard for the evaluation of the presence, number, size, and site (renal or ureteral [upper, middle, and lower]) of the stones, we calculated the sensitivity, specificity, and positive and negative predictive values of hematuria for diagnosing both renal and ureteral stones. Urolithiasis was diagnosed in 388 patients (64%) and 216 patients (36%) had no stones on a noncontrast-enhanced CT scan. The sensitivity, specificity, positive predictive value, and negative predictive value for microhematuria were 77%, 33%, 67%, and 45%, respectively. Microhematuria was more common in patients with ureteral stones only (139 patients) and had a sensitivity of 85% compared to patients with renal stones only (32 patients), with a sensitivity of 55% (P < 0.001). There were no significant differences in the specificity or positive or negative predictive values. Although microhematuria is more sensitive to ureteral stones, the absence of microhematuria does not exclude the possibility of urolithiasis and a noncontrast-enhanced CT scan should be the gold standard diagnostic tool.

Clinical Reproducibility of the Stone Volume Measurement: A "Kidney Stone Calculator" Study.

An accurate estimation of the stone burden is the key factor for predicting retrograde intra-renal surgical outcomes. Volumetric calculations better stratify stone burden than linear measurements. We developed a free software to assess the stone volume and estimate the lithotrity duration according to 3D-segmented stone volumes, namely the Kidney Stone Calculator (KSC). The present study aimed to validate the KSC's reproducibility in clinical cases evaluating its inter-observer and intra-observer correlations. Fifty patients that harbored renal stones were retrospectively selected from a prospective cohort. For each patient, three urologists with different experience levels in stone management made five measurements of the stone volume on non-contrast-enhanced computed tomography (NCCT) images using the KSC. the overall inter-observer correlation (Kendall's concordance coefficient) was 0.99 (p < 0.0001). All three paired analyses of the inter-observer reproducibility were superior to 0.8. The intra-observer variation coefficients varied from 4% to 6%, and Kendall's intra-observer concordance coefficient was found to be superior to 0.98 (p < 0.0001) for each participant. Subgroup analyses showed that the segmentation of complex stones seems to be less reproductible. The Kidney Stone Calculator is a reliable tool for the stone burden estimation. Its extension for calculating the lithotrity duration is of major interest and could help the practitioner in surgical planning.

Aorta and main pulmonary artery segmentation using stacked U-Net and localization on non-contrast-enhanced computed tomography images.

The contact between the aorta, main pulmonary artery (MPA), main pulmonary vein, vena cava (VC), and esophagus affects segmentation of the aorta and MPA in non-contrast-enhanced computed tomography (NCE-CT) images. A two-stage stacked U-Net and localization of the aorta and MPA were developed for the segmentation of the aorta and MPA in NCE-CT images. Normal-dose NCE-CT images of 24 subjects with chronic thromboembolic pulmonary hypertension (CTEPH) and low-dose NCE-CT images of 100 subjects without CTEPH were used in this study. The aorta is in contact with the ascending aorta (AA) and MPA, the AA with the VC, the aortic arch (AR) with the VC and esophagus, and the descending aorta (DA) with the esophagus. These contact surfaces were manually annotated. The contact surfaces were quantified using the contact surface ratio (CSR). Segmentation of the aorta and MPA in NCE-CT images was performed by localization of the aorta and MPA and a two-stage stacked U-Net. Localization was performed by extracting and processing the trachea and main bronchus. The first stage of the stacked U-Net consisted of a 2D U-Net, 2D U-Net with a pre-trained VGG-16 encoder, and 2D attention U-Net. The second stage consisted of a 3D U-Net with four input channels: the CT volume and three segmentation results of the first stage. The model was trained and tested using 10-fold cross-validation. Segmentation of the entire volume was evaluated using the Dice similarity coefficient (DSC). Segmentation of the contact area was also assessed using the mean surface distance (MSD). The statistical analysis of the evaluation underwent a multi-comparison correction. CTEPH and non-CTEPH cases were classified based on the vessel diameters measured from the segmented MPA. For the noncontact surfaces of AA, the MSD of stacked U-Net was 0.31±0.10mm (p<0.05) and 0.32±0.13mm (p<0.05) for non-CTEPH and CTEPH cases, respectively. For contact surfaces with a CSR of 0.4 or greater in AA, the MSD was 0.52±0.23mm (p<0.05), and 0.68±0.29mm (p>0.05) for non-CTEPH and CTEPH cases, respectively. MSDs were lower than those of 2D and 3D U-Nets for contact and noncontact surfaces; moreover, MSDs increased slightly with larger CSRs. However, the stacked U-Net achieved MSDs of approximately 1pixel for a wide contact surface. The area under the receiver operating characteristic curve for CTEPH and non-CTEPH classification using the right main pulmonary artery (RMPA) diameter was 0.97 (95% confidence interval [CI]: 0.94-1.00). Segmentation of the aorta and MPA on NCE-CT images were affected by vascular and esophageal contact. The application of stacked U-Net and localization techniques for non-CTEPH and CTEPH cases mitigated the impact of contact, suggesting its potential for diagnosing CTEPH.

Prevalence of malignant or possibly malignant renal masses among homogeneous low-attenuation masses that are too small to characterize at computed tomography.

Homogeneous low-attenuation renal masses that are too small to characterize (tstc) are considered clinically insignificant; however, based primarily on expert opinion. To determine the prevalence of malignant or possibly malignant masses among homogeneous low-attenuation renal masses that are tstc. This retrospective cross-sectional study evaluated 75 patients with 104 tstc who underwent renal CT and MRI between Jan 2016 and Jul 2022. Low-attenuation renal masses measuring < 1cm in size were identified and, independently evaluated by two blinded radiologists measuring attenuation (Hounsfield Units, HU) at non-contrast enhanced CT (NECT) and nephrographic phase contrast-enhanced (CE)-CT when possible. Reference standard for benign cyst was MRI and for other renal masses was pathology or MRI showing enhancement. Average tstc size was 6 ± 2 (range 2-10) mm. Considering only incidental tstc (CT performed for another reason), 100% (98/98, 95%CI 96-100%) tstc were benign. Overall, considering both incidental and tstc referred for further characterization, there were 94% (98/104; 95% Confidence Intervals [CIs] 88-98%) benign cysts and 6% (6/104; 95%CI 2-12%) other masses (1 Bosniak 2F cystic mass, 2 probable renal cell carcinoma (RCC), three metastases). Pseudoenhancement, attenuation change > 10 HU or > 20 HU, was present in 29% (15/59) and 12% (7/59) benign cysts. All six other masses enhanced by > 20 HU. CECT threshold of ≤ 30 HU correctly classified 62% of benign cysts (61/98). All six other masses measured > 30 HU at CECT. The prevalence of malignant or possibly malignant renal masses among homogeneous low-attenuation too small to characterize masses among incidental tstc masses is near zero. Attenuation measurements misclassify a substantial proportion of these cysts, likely due to their small size.

Diagnostic Sensitivity of Unenhanced CT for Cerebral Venous Thrombosis: Can Clot Density Measurement Replace CT Venogram?

Objectives Cerebral venous sinus thrombosis is an important cause of stroke in young adults. Noncontrast-enhanced CT head (NECT) is almost always the first investigation. Our objectives were as follows: 1. How accurately does venous sinus density on NECT predict the presence of clot on CT venogram (CTV)? 2. Whether repeated measurements changed the confidence? 3. How many venous sinus thrombus would be missed if we do not do a CTV? 4. Can clot density measurement replace CTV? Methods Multicenter case-control study was designed with data from seven hospitals. Inclusion criteria: all CT and magnetic resonance imaging venograms with a prior NECT, performed between 1.1.2018 and 31.12.2018 (12 months), were included. Hounsfield unit (HU) values were calculated at the site of highest density on the NECT. Logistic regression analysis was performed using STATA. Result Two-hundred seventy-seven cases met the criteria with 33 positive cerebral venous thrombosis (density on NECT 60-92 HU) and 244 negative examinations (density on NECT 31-68 HU). Area under the curve for average clot density on NECT was 0.9984. Conclusion We found a strong relationship between sinus density on NECT and outcome of CTV. Repeating density measurements did not add any predictive value or changed outcome. Advances in Knowledge Density 70 HU or higher on NECT always resulted in a positive CTV but would miss a fifth of the positives. Cutoff at 60 HU would not miss any but result in significant false positives. An efficient option could be to limit CTV to sinus densities 60 to 70 HU only. However, a larger study would be required for such change in practice.

Abstract TP98: Deep Learning-based Automatic Ischemic Core Estimation Based On Non-contrastenhanced Computed Tomography

Background/Purpose: With the spread of thrombus retrieval procedures, the demand for immediate diagnosis and severity assessment of ischemic stroke is increasing, but current standard method using ASPECTS is semi-quantitative, and has low consistency among physicians. We develop and validate the fully automatic machine learning-based ischemic core segmentation model based only on non-contrast-enhanced computed tomography (NCCT). Materials and Methods: This multicenter retrospective study included patients with anterior circulation acute ischemic stroke who received both CT and MRI before thrombolysis or recanalization treatment between 2013 to 2019. The ischemic core on CT was manually delineated referencing the DWI and ADC map. A deep learning-based ischemic core segmentation model was developed using the derivation cohort data from 3 institutions (n=313), and the model performance was validated on the external validation cohort data from 3 institutions (n=106). Results: On the validation cohort, the median time between CT and MRI was 18 min. The ischemic core volume calculated by the DL model was significantly correlated with the reference standard (Pearson r=0.91, p&lt;.001). The correlation was significant both in early time window (&lt;4.5hr from onset; Pearson r=0.91, p&lt;.0.01) and the late time window (&gt;4.5hr from onset; Pearson r=0.93, p&lt;0.01). The median difference between model and reference standard was 4.7 mL (IQR, 0.8-12.4 mL). The DL model displayed high accuracy on discriminating large ischemic core (&gt;70mL) with a sensitivity of 84.2%, specificity of 97.7%, and AUC of 0.91. Conclusions: A segmentation model of cerebral ischemic core volume of the anterior circulation based on NCCT using deep learning showed good correlation with the correct labels generated with reference to MRI.

Cephalhaematoma Mimicking an Extradural Haematoma due to MirrorImage Artifact on Sonography in a Term Neonate: A Case Report

Cephalhaematomas and subgaleal haematomas are among the most common birth injuries and are associated with birth trauma, forceps, and vacuum-assisted deliveries. They present as scalp swelling and are usually identified shortly after birth. During sonographic examination, if an ultrasound beam scatters off a mirror-like interface, it creates mirror-image artifacts that can cause a diagnostic dilemma. In this case report, a six-day-old neonate presented with a right-side parietal cephalhaematoma that appeared to resemble an epidural haematoma on routine sonographic examination. Gray scale ultrasound revealed an anechoic structure resembling an epidural haematoma in the right parietal region. However, a non-contrast-enhanced computed tomography (NECT) scan of the brain showed a cephalhaematoma without an underlying epidural haematoma. Further evaluation using colour Doppler sonography revealed normal vascular findings within an anechoic space, and gray scale imaging in the sagittal plane showed normal cerebral parenchyma without midline shift. These findings helped identify the observed structure as a mirror-image artifact. It is important to note that these artifacts can lead to diagnostic errors, resulting in additional investigations and causing anxiety for parents. Understanding and being aware of these artifacts can help avoid unnecessary imaging and reduce radiation exposure.

Automated estimation of ischemic core volume on noncontrast-enhanced CT via machine learning.

Accurate estimation of ischemic core on baseline imaging has treatment implications in patients with acute ischemic stroke (AIS). Machine learning (ML) algorithms have shown promising results in estimating ischemic core using routine noncontrast computed tomography (NCCT). We used an ML-trained algorithm to quantify ischemic core volume on NCCT in a comparative analysis to pretreatment magnetic resonance imaging (MRI) diffusion-weighted imaging (DWI) in patients with AIS. Patients with AIS who had both pretreatment NCCT and MRI were enrolled. An automatic segmentation ML approach was applied using Brainomix software (Oxford, UK) to segment the ischemic voxels and calculate ischemic core volume on NCCT. Ischemic core volume was also calculated on baseline MRI DWI. Comparative analysis was performed using Bland-Altman plots and Pearson correlation. A total of 72 patients were included. The time-to-stroke onset time was 134.2/89.5 minutes (mean/median). The time difference between NCCT and MRI was 64.8/44.5 minutes (mean/median). In patients who presented within 1 hour from stroke onset, the ischemic core volumes were significantly (p = 0.005) underestimated by ML-NCCT. In patients presented beyond 1 hour, the ML-NCCT estimated ischemic core volumes approximated those obtained by MRI-DWI and with significant correlation (r = 0.56, p < 0.001). The ischemic core volumes calculated by the described ML approach on NCCT approximate those obtained by MRI in patients with AIS who present beyond 1 hour from stroke onset.

Artificial intelligence-driven ASPECTS for the detection of early stroke changes in non-contrast CT: a systematic review and meta-analysis

BackgroundRecent advances in machine learning have enabled development of the automated Alberta Stroke Program Early CT Score (ASPECTS) prediction algorithms using non-contrast enhanced computed tomography (NCCT) scans. The applicability of...

CT radiomics to predict Deauville score 4 positive and negative Hodgkin lymphoma manifestations

18F-FDG-PET/CT is standard to assess response in Hodgkin lymphoma by quantifying metabolic activity with the Deauville score. PET/CT, however, is time-consuming, cost-extensive, linked to high radiation and has a low availability. As an alternative, we investigated radiomics from non-contrast-enhanced computed tomography (NECT) scans. 75 PET/CT examinations of 43 patients on two different scanners were included. Target lesions were classified as Deauville score 4 positive (DS4+) or negative (DS4−) based on their SUVpeak and then segmented in NECT images. From these segmentations, 107 features were extracted with PyRadiomics. All further statistical analyses were then performed scanner-wise: differences between DS4+ and DS4− manifestations were assessed with the Mann–Whitney-U-test and single feature performances with the ROC-analysis. To further verify the reliability of the results, the number of features was reduced using different techniques. The feature median showed a high sensitivity for DS4+ manifestations on both scanners (scanner A: 0.91, scanner B: 0.85). It furthermore was the only feature that remained in both datasets after applying different feature reduction techniques. The feature median from NECT concordantly has a high sensitivity for DS4+ Hodgkin manifestations on two different scanners and thus could provide a surrogate for increased metabolic activity in PET/CT.

A delta-radiomics model for preoperative prediction of invasive lung adenocarcinomas manifesting as radiological part-solid nodules.

This study aims to explore the value of the delta-radiomics (DelRADx) model in predicting the invasiveness of lung adenocarcinoma manifesting as radiological part-solid nodules (PSNs). A total of 299 PSNs histopathologically confirmed as lung adenocarcinoma (training set, n = 209; validation set, n = 90) in our hospital were retrospectively analyzed from January 2017 to December 2021. All patients underwent diagnostic noncontrast-enhanced CT (NCECT) and contrast-enhanced CT (CECT) before surgery. After image preprocessing and ROI segmentation, 740 radiomic features were extracted from NCECT and CECT, respectively, resulting in 740 DelRADx. A DelRADx model was constructed using the least absolute shrinkage and selection operator logistic (LASSO-logistic) algorithm based on the training cohort. The conventional radiomics model based on NCECT was also constructed following the same process for comparison purposes. The prediction performance was assessed using area under the ROC curve (AUC). To provide an easy-to-use tool, a radiomics-based integrated nomogram was constructed and evaluated by integrated discrimination increment (IDI), calibration curves, decision curve analysis (DCA), and clinical impact plot. The DelRADx signature, which consisted of nine robust selected features, showed significant differences between the AIS/MIA group and IAC group (p < 0.05) in both training and validation sets. The DelRADx signature showed a significantly higher AUC (0.902) compared to the conventional radiomics model based on NCECT (AUC = 0.856) in the validation set. The IDI was significant at 0.0769 for the integrated nomogram compared with the DelRADx signature. The calibration curve of the integrated nomogram demonstrated favorable agreement both in the training set and validation set with a mean absolute error of 0.001 and 0.019, respectively. Decision curve analysis and clinical impact plot indicated that if the threshold probability was within 90%, the integrated nomogram showed a high clinical application value. The DelRADx method has the potential to assist doctors in predicting the invasiveness for patients with PSNs. The integrated nomogram incorporating the DelRADx signature with the radiographic features could facilitate the performance and serve as an alternative way for determining management.

Abstract 11285: Synthetic-Contrast-Enhanced Computed Tomography Creating by Machine Learning Without Contrast Media

Introduction: Contrast-enhanced computed tomography (CECT) is the gold standard for the diagnosis of aortic dissection (AD), at the same time, CECT has risks of nephropathy and allergic reactions caused by contrast media. Hypothesis: Synthetic-contrast-enhanced computed tomography (SECT) created from non-contrast-enhanced computed tomography (NECT) by machine learning without contrast agents could be an alternative to CECT in the diagnosis of AD. Methods: 50 patients with type A AD and 200 patients without AD who underwent NECT and CECT in pairs at Kobe university hospital were randomly selected and used for model training and testing. 3120 CT slices were used for training and 379 slices were used for testing. AD detection algorithms for NECT or CECT were developed by convolutional neural networks using NECT datasets or CECT datasets respectively. A SECT creating algorithm was developed by u-net using the pairs of CT datasets. To provide visual explanations of diagnosis, the gradient-weighted class activation mapping (Grad-CAM) method was adapted. Results: The provided sensitivities of AD detection algorithms for NECT or SECT were 90 % and 97 %, and the specificities were 100 % and 100 % respectively. When the developed AD detection algorithms for CECT were adapted to SECT, the provided sensitivity of 100 %, and sensitivity of 95 %. Most slices SECT of patients with AD demonstrated intimal flap and/or thrombosed false lumen as the almost same as CECT. The Grad-CAM visualization revealed that the AD detection algorithm using SECT images identified intimal flap or thrombosed false lumen in images more clearly than the AD detection algorithm using NECT images. Conclusions: The SECT creating algorithm improved diagnostic sensitivity for detecting AD using NECT datasets, which suggests the potential of the proposed method to support the clinical practice by reducing complications of contrast agents.

Radiomics models based on CT at different phases predicting lymph node metastasis of esophageal squamous cell carcinoma (GASTO-1089).

To investigate the value of radiomics models based on CT at different phases (non-contrast-enhanced and contrast-enhanced images) in predicting lymph node (LN) metastasis in esophageal squamous cell carcinoma (ESCC). Two hundred and seventy-four eligible patients with ESCC were divided into a training set (n =193) and a validation set (n =81). The least absolute shrinkage and selection operator algorithm (LASSO) was used to select radiomics features. The predictive models were constructed with radiomics features and clinical factors through multivariate logistic regression analysis. The predictive performance and clinical application value of the models were evaluated by area under receiver operating characteristic curve (AUC) and decision curve analysis (DCA). The Delong Test was used to evaluate the differences in AUC among models. Sixteen and eighteen features were respectively selected from non-contrast-enhanced CT (NECT) and contrast-enhanced CT (CECT) images. The model established using only clinical factors (Model 1) has an AUC value of 0.655 (95%CI 0.552-0.759) with a sensitivity of 0.585, a specificity of 0.725 and an accuracy of 0.654. The models contained clinical factors with radiomics features of NECT or/and CECT (Model 2,3,4) have significantly improved prediction performance. The values of AUC of Model 2,3,4 were 0.766, 0.811 and 0.809, respectively. It also achieved a great AUC of 0.800 in the model built with only radiomics features derived from NECT and CECT (Model 5). DCA suggested the potential clinical benefit of model prediction of LN metastasis of ESCC. A comparison of the receiver operating characteristic (ROC) curves using the Delong test indicated that Models 2, 3, 4, and 5 were superior to Model 1(P< 0.05), and no difference was found among Model 2, 3, 4 and Model 5(P > 0.05). Radiomics models based on CT at different phases could accurately predict the lymph node metastasis in patients with ESCC, and their predictive efficiency was better than the clinical model based on tumor size criteria. NECT-based radiomics model could be a reasonable option for ESCC patients due to its lower price and availability for renal failure or allergic patients.

Contrast Medium Use in Computed Tomography for Patients Presenting with Headache: 4-year Retrospective Two-Center Study in Central and Western Regions of Ghana.

Background Contrast medium (CM) administration during computed tomography (CT) enhances the accuracy in the detection and interpretation of abnormalities. Evidence from literature also validate the essence of CM in imaging studies. CT, by virtue of its ubiquity, ease of use, speed, and lower financial footprint, is usually the first investigation in cases of headache. Through a multicenter retrospective analysis, we compared findings of contrast-enhanced CT (CECT) to noncontrast-enhanced CT (NCECT) head examinations among patients presenting with headache. Methods A multicenter retrospective analysis of four years' CT head examination data at two radiology centers located in Central and Western Regions of Ghana were reviewed. Records of patients who presented with headache as principal complaint between January 2017 and December 2020 were reviewed. A total of 477 records of patients with headache were identified, retrieved and evaluated. A Chi-square test and Fisher exact test were used to compare the CECT and NCECT groups. Binary logistic regression analysis was computed to assess association between CECT and each CT findings. Statistical significance was considered at p < 0.05 with a 95% confidence interval. Results A significant proportion of the patients was females (51.8% in CECT and 60% in NCECT). The NCECT group (40.06 ± 14.76 years) was relatively older than the CECT group (38.43 ± 17.64 years). There was a significant difference between the CECT and NCECT in terms of age (p=0.002) and facility CT was performed (p < 0.0001). The rate of abnormalities was higher in CECT (43.5%, 166/382) compared NCECT (37.9%, 36/95). There was no significant association between CT head findings and contrast enhancement. Conclusion CECT examination accounted for 5.6% increase in the detection of head abnormalities. Efforts required to establish local standard operation procedures (SOPs) for contrast medium use especially in CT head examinations. Further studies to improve the knowledge of agents, mechanism of action, and safety of contrast media used among practitioners in Ghana is recommended.

The adding value of contrast-enhanced CT radiomics: Differentiating tuberculosis from non-tuberculous infectious lesions presenting as solid pulmonary nodules or masses.

PurposeTo compare the value of contrast-enhanced CT (CECT) and non-contrast-enhanced CT (NCECT) radiomics models in differentiating tuberculosis (TB) from non-tuberculous infectious lesions (NTIL) presenting as solid pulmonary nodules or masses, and develop a combine radiomics model (RM).Materials and methodsThis study was a retrospective analysis of 101 lesions in 95 patients, including 49 lesions (from 45 patients) in the TB group and 52 lesions (from 50 patients) in the NTIL group. Lesions were randomly divided into training and test sets in the ratio of 7:3. Conventional imaging features were used to construct a conventional imaging model (IM). Radiomics features screening and NCECT or CECT RM construction were carried out by correlation analysis and gradient boosting decision tree, and logistic regression. Finally, conventional IM, NCECT RM, and CECT RM were used for combine RM construction. Additionally, we recruited three radiologists for independent diagnosis. The differential diagnostic performance of each model was assessed using the areas under the receiver operating characteristic curve (AUCs).ResultsThe CECT RM (training AUC, 0.874; test AUC, 0.796) outperformed the conventional IM (training AUC, 0.792; test AUC, 0.708), the NCECT RM (training AUC, 0.835; test AUC, 0.704), and three radiologists. The diagnostic efficacy of the combine RM (training AUC, 0.922; test AUC, 0.833) was best in the training and test sets.ConclusionsThe diagnostic efficacy of the CECT RM was superior to that of the NCECT RM in identifying TB from NTIL presenting as solid pulmonary nodules or masses. The combine RM had the best performance and may outperform expert radiologists.