External validation of a convolutional neural network algorithm for opportunistically detecting vertebral fractures in routine CT scans.

The Convolutional Neural Network algorithm achieved a sensitivity of 94% and specificity of 93% in identifying scans with vertebral fractures (VFs). The external validation results suggest that the algorithm provides an opportunity to aid radiologists with the early identification of VFs in routine CT scans of abdomen and chest. PURPOSE: To evaluate the performance of a previously trained Convolutional Neural Network (CNN) model to automatically detect vertebral fractures (VFs) in CT scans in an external validation cohort. METHODS: Two Chinese studies and clinical data were used to retrospectively select CT scans of the chest, abdomen and thoracolumbar spine in men and women aged ≥50 years. The CT scans were assessed using the semiquantitative (SQ) Genant classification for prevalent VFs in a process blinded to clinical information. The performance of the CNN model was evaluated against reference standard readings by the area under the receiver operating characteristics curve (AUROC), accuracy, Cohen's kappa, sensitivity, and specificity. RESULTS: A total of 4,810 subjects were included, with a median age of 62 years (IQR 56-67), of which 2,654 (55.2%) were females. The scans were acquired between January 2013 and January 2019 on 16 different CT scanners from three different manufacturers. 2,773 (57.7%) were abdominal CTs. A total of 628 scans (13.1%) had ≥1 VF (grade 2-3), representing 899 fractured vertebrae out of a total of 48,584 (1.9%) visualized vertebral bodies. The CNN's performance in identifying scans with ≥1 moderate or severe fractures achieved an AUROC of 0.94 (95% CI: 0.93-0.95), accuracy of 93% (95% CI: 93%-94%), kappa of 0.75 (95% CI: 0.72-0.77), a sensitivity of 94% (95% CI: 92-96%) and a specificity of 93% (95% CI: 93-94%). CONCLUSION: The algorithm demonstrated excellent performance in the identification of vertebral fractures in a cohort of chest and abdominal CT scans of Chinese patients ≥50 years.

CT-defined sarcopenia predicts treatment response in primary central nervous system lymphomas.

OBJECTIVE: Body composition assessment derived from cross-sectional imaging has shown promising results as a prognostic biomarker in several tumor entities. Our aim was to analyze the role of low skeletal muscle mass (LSMM) and fat areas for prognosis of dose-limiting toxicity (DLT) and treatment response in patients with primary central nervous system lymphoma (PCNSL). METHODS: Overall, 61 patients (29 female patients, 47.5%) with a mean age of 63.8 ± 12.2 years, range 23-81 years, were identified in the data base between 2012 and 2020 with sufficient clinical and imaging data. Body composition assessment, comprising LSMM and visceral and subcutaneous fat areas, was performed on one axial slice on L3-height derived from staging computed tomography (CT) images. DLT was assessed during chemotherapy in clinical routine. Objective response rate (ORR) was measured on following magnetic resonance images of the head accordingly to the Cheson criteria. RESULTS: Twenty-eight patients had DLT (45.9%). Regression analysis revealed that LSMM was associated with objective response, OR = 5.19 (95% CI 1.35-19.94, p = 0.02) (univariable regression), and OR = 4.23 (95% CI 1.03- 17.38, p = 0.046) (multivariable regression). None of the body composition parameters could predict DLT. Patients with normal visceral to subcutaneous ratio (VSR) could be treated with more chemotherapy cycles compared to patients with high VSR (mean, 4.25 vs 2.94, p = 0.03). Patients with ORR had higher muscle density values compared to patients with stable and/or progressive disease (34.46 ± vs 28.18 ± HU, p = 0.02). CONCLUSIONS: LSMM is strongly associated with objective response in patients with PCNSL. Body composition parameters cannot predict DLT. CLINICAL RELEVANCE STATEMENT: Low skeletal muscle mass on computed tomography (CT) is an independent prognostic factor of poor treatment response in central nervous system lymphoma. Analysis of the skeletal musculature on staging CT should be implemented into the clinical routine in this tumor entity. KEY POINTS: • Low skeletal muscle mass is strongly associated with the objective response rate. • No body composition parameters could predict dose-limiting toxicity.

Value of deep learning reconstruction of chest low-dose CT for image quality improvement and lung parenchyma assessment on lung window.

OBJECTIVES: To explore the performance of low-dose computed tomography (LDCT) with deep learning reconstruction (DLR) for the improvement of image quality and assessment of lung parenchyma. METHODS: Sixty patients underwent chest regular-dose CT (RDCT) followed by LDCT during the same examination. RDCT images were reconstructed with hybrid iterative reconstruction (HIR) and LDCT images were reconstructed with HIR and DLR, both using lung algorithm. Radiation exposure was recorded. Image noise, signal-to-noise ratio, and subjective image quality of normal and abnormal CT features were evaluated and compared using the Kruskal-Wallis test with Bonferroni correction. RESULTS: The effective radiation dose of LDCT was significantly lower than that of RDCT (0.29 ± 0.03 vs 2.05 ± 0.65 mSv, p < 0.001). The mean image noise ± standard deviation was 33.9 ± 4.7, 39.6 ± 4.3, and 31.1 ± 3.2 HU in RDCT, LDCT HIR-Strong, and LDCT DLR-Strong, respectively (p < 0.001). The overall image quality of LDCT DLR-Strong was significantly better than that of LDCT HIR-Strong (p < 0.001) and comparable to that of RDCT (p > 0.05). LDCT DLR-Strong was comparable to RDCT in evaluating solid nodules, increased attenuation, linear opacity, and airway lesions (all p > 0.05). The visualization of subsolid nodules and decreased attenuation was better with DLR than with HIR in LDCT but inferior to RDCT (all p < 0.05). CONCLUSION: LDCT DLR can effectively reduce image noise and improve image quality. LDCT DLR provides good performance for evaluating pulmonary lesions, except for subsolid nodules and decreased lung attenuation, compared to RDCT-HIR. CLINICAL RELEVANCE STATEMENT: The study prospectively evaluated the contribution of DLR applied to chest low-dose CT for image quality improvement and lung parenchyma assessment. DLR can be used to reduce radiation dose and keep image quality for several indications. KEY POINTS: • DLR enables LDCT maintaining image quality even with very low radiation doses. • Chest LDCT with DLR can be used to evaluate lung parenchymal lesions except for subsolid nodules and decreased lung attenuation. • Diagnosis of pulmonary emphysema or subsolid nodules may require higher radiation doses.

Improving radiomics reproducibility using deep learning-based image conversion of CT reconstruction algorithms in hepatocellular carcinoma patients.

OBJECTIVES: CT reconstruction algorithms affect radiomics reproducibility. In this study, we evaluate the effect of deep learning-based image conversion on CT reconstruction algorithms. METHODS: This study included 78 hepatocellular carcinoma (HCC) patients who underwent four-phase liver CTs comprising non-contrast, late arterial (LAP), portal venous (PVP), and delayed phase (DP), reconstructed using both filtered back projection (FBP) and advanced modeled iterative reconstruction (ADMIRE). PVP images were used to train a convolutional neural network (CNN) model to convert images from FBP to ADMIRE and vice versa. LAP, PVP, and DP images were used for validation and testing. Radiomic features were extracted for each patient with a semi-automatic segmentation tool. We used concordance correlation coefficients (CCCs) to evaluate the radiomics reproducibility for original FBP (oFBP) vs. original ADMIRE (oADMIRE), oFBP vs. converted FBP (cFBP), and oADMIRE vs. converted ADMIRE (cADMIRE). RESULTS: In the test group including 30 patients, the CCC and proportion of reproducible features (CCC ≥ 0.85) for oFBP vs. oADMIRE were 0.65 and 32.9% (524/1595) for LAP, 0.65 and 35.9% (573/1595) for PVP, and 0.69 and 43.8% (699/1595) for DP. For oFBP vs. cFBP, the values increased to 0.92 and 83.9% (1339/1595) for LAP, 0.89 and 71.0% (1133/1595) for PVP, and 0.90 and 79.7% (1271/1595) for DP. Similarly, for oADMIRE vs. cADMIRE, the values increased to 0.87 and 68.1% (1086/1595) for LAP, 0.91 and 82.1% (1309/1595) for PVP, and 0.89 and 76.2% (1216/1595) for DP. CONCLUSIONS: CNN-based image conversion between CT reconstruction algorithms improved the radiomics reproducibility of HCCs. CLINICAL RELEVANCE STATEMENT: This study demonstrates that using a CNN-based image conversion technique significantly improves the reproducibility of radiomic features in HCCs, highlighting its potential for enhancing radiomics research in HCC patients. KEY POINTS: Radiomics reproducibility of HCC was improved via CNN-based image conversion between two different CT reconstruction algorithms. This is the first clinical study to demonstrate improvements across a range of radiomic features in HCC patients. This study promotes the reproducibility and generalizability of different CT reconstruction algorithms in radiomics research.

CT acquisition parameter selection in the real world: impacts on radiation dose and variation amongst 155 institutions.

OBJECTIVE: Quantify the relationship between CT acquisition parameters and radiation dose, how often parameters are adjusted in real-world practice, and their degree of contribution to real-world dose distribution. Identify discrepancies between parameters that are impactful in theory and impactful in practice. METHODS: This study analyses 1.3 million consecutive adult routine abdomen exams performed between November 2015 and Jan 2021 included in the University of California, San Francisco International CT Dose Registry of 155 institutions. We calculated geometric standard deviation (gSD) for five parameters (kV, mAs, spiral pitch, number of phases, scan length) to assess variation in practice. A Gaussian mixed regression model was performed to predict the radiation dose-length product (DLP) using the parameters. Three conceptualizations of "impact" were computed for each parameter. To reflect the theoretical impact, we predict the increase in DLP per 10% (and 15%) increase in the parameter. To reflect the real-world practical impact, we predict the increase in DLP per gSD increase in the parameter. RESULTS: Among studied examinations, mAs, number of phases, and scan length were frequently manipulated (gSD 1.52-1.70); kV was rarely manipulated (gSD 1.07). Theoretically, kV is the most impactful parameter (29% increase in DLP per 10% increase in kV, versus 5-9% increase for other parameters). In real-world practice, kV is less impactful; for each gSD increase in kV, the DLP increases by 20%, versus 22-69% for other parameters. CONCLUSION: Despite the potential impact of kV on radiation dose, this parameter is rarely manipulated in common practice and this potential remains untapped. CLINICAL RELEVANCE STATEMENT: CT beam energy (kV) modulation has the potential to strongly reduce radiation over-dosage to the patient, theoretically more so than similar degrees of modulation in other CT acquisition parameters. Despite this, beam energy modulation rarely occurs in practice, leaving its potential untapped. KEY POINTS: • The relationship between CT acquisition parameter selection and radiation dose roughly coincided with established theoretical understanding. • CT acquisition parameters differ from each other in frequency and magnitude of manipulation, with beam energy (kV) being rarely manipulated. • Beam energy (kV) has the potential to substantially impact radiation dose, but because it is rarely manipulated, it is the least impactful CT acquisition parameter affecting radiation dose in practice.

ESR Essentials: Imaging of sacroiliitis-practice recommendations by ESSR.

Sacroiliitis is commonly seen in patients with axial spondyloarthritis, in whom timely diagnosis and treatment are crucial to prevent irreversible structural damage. Imaging has a prominent place in the diagnostic process and several new imaging techniques have been examined for this purpose. We present a summary of updated evidence-based practice recommendations for imaging of sacroiliitis. MRI remains the imaging modality of choice for patients with suspected sacroiliitis, using at least four sequences: coronal oblique T1-weighted and fluid-sensitive sequences, a perpendicular axial oblique sequence, and a sequence for optimal evaluation of the bone-cartilage interface. Both active inflammatory and structural lesions should be described in the report, indicating location and extent. Radiography and CT, especially low-dose CT, are reasonable alternatives when MRI is unavailable, as patients are often young. This is particularly true to evaluate structural lesions, at which CT excels. Dual-energy CT with virtual non-calcium images can be used to depict bone marrow edema. Knowledge of normal imaging features in children (e.g., flaring, blurring, or irregular appearance of the articular surface) is essential for interpreting sacroiliac joint MRI in children because these normal processes can simulate disease. CLINICAL RELEVANCE STATEMENT: Sacroiliitis is a potentially debilitating disease if not diagnosed and treated promptly, before structural damage to the sacroiliac joints occurs. Imaging has a prominent place in the diagnostic process. We present a summary of practice recommendations for imaging of sacroiliitis, including several new imaging techniques. KEY POINTS: • MRI is the modality of choice for suspected inflammatory sacroiliitis, including a joint-line-specific sequence for optimal evaluation of the bone-cartilage interface to improve detection of erosions. • Radiography and CT (especially low-dose CT) are reasonable alternatives when MRI is unavailable. • Knowledge of normal imaging features in children is mandatory for interpretation of MRI of pediatric sacroiliac joints.

CT of appendicoliths in adult appendicitis: clinical significance and characteristics of overlooked cases.

OBJECTIVES: Accurate computed tomography (CT) identification of appendicoliths in adults with acute appendicitis is crucial as it may preclude nonoperative management due to high risk of failure and complications. This investigation aimed to identify the significance of appendicoliths in acute appendicitis and to evaluate the performance of portovenous-phase (PVP) CT and the consequences of overlooked appendicoliths. METHODS: CT examinations of 324 consecutive patients (mean age 51.9 years, 112 men) with pathologically confirmed acute appendicitis were retrospectively included. Two radiologists independently reviewed the images, and disagreement was resolved by a consensus. RESULTS: Appendicoliths were identified in 134/324 patients, of which 75 had complicated appendicitis. Among 190 patients without appendicoliths, 52 had complicated appendicitis. An appendicolith was independently associated with complicated appendicitis (adjusted odds ratio 2.289; 95% CI: 1.343-3.902; p = 0.002). The larger minimum diameter was significantly associated with complication. The 4.5-/6.0-mm cutoffs for minimum and maximum diameters of appendicoliths demonstrated 82.7%/85.3% sensitivity and 35.6%/33.9% specificity in predicting complications. The PVP alone had 82.1-88.1% sensitivity, respectively per patient and per appendicolith, and a 100% specificity in the detection of appendicoliths, as compared with combined noncontrast and PVP. PVP overlooked 28/237 appendicoliths (11.8%) corresponding to 24/134 patients (17.9%). Of the 24 patients with overlooked appendicoliths, 16 had complicated appendicitis but 14 were correctly categorized by findings other than appendicoliths. In total, 2/127 patients (1.6%) with complicated appendicitis were misdiagnosed as having uncomplicated appendicitis. CONCLUSIONS: Appendicoliths in acute appendicitis were strongly associated with complications. While PVP overlooked some appendicoliths, only 1.6% of complicated appendicitis were misclassified when considering other CT findings. CLINICAL RELEVANCE STATEMENT: This study found a strong association between appendicoliths and complications. Its presence may preclude conservative management. Although portovenous-phase CT overlooked some appendicoliths, the combination with other CT findings allowed correct classification in a vast majority of cases. KEY POINTS: • Accurate identification of appendicoliths is crucial for nonoperative management decisions in adult acute appendicitis. • Appendicoliths are strongly associated with complications in adult acute appendicitis. • Portovenous-phase CT overlooked some appendicoliths, but only a small percentage of patients with complicated appendicitis were misclassified when considering other CT findings.

Large variation in radiation dose for routine abdomen CT: reasons for excess and easy tips for reduction.

OBJECTIVE: To characterize the use and impact of radiation dose reduction techniques in actual practice for routine abdomen CT. METHODS: We retrospectively analyzed consecutive routine abdomen CT scans in adults from a large dose registry, contributed by 95 hospitals and imaging facilities. Grouping exams into deciles by, first, patient size, and second, size-adjusted dose length product (DLP), we summarized dose and technical parameters and estimated which parameters contributed most to between-protocols dose variation. Lastly, we modeled the total population dose if all protocols with mean size-adjusted DLP above 433 or 645 mGy-cm were reduced to these thresholds. RESULTS: A total of 748,846 CTs were performed using 1033 unique protocols. When sorted by patient size, patients with larger abdominal diameters had increased dose and effective mAs (milliampere seconds), even after adjusting for patient size. When sorted by size-adjusted dose, patients in the highest versus the lowest decile in size-adjusted DLP received 6.4 times the average dose (1680 vs 265 mGy-cm) even though diameter was no different (312 vs 309 mm). Effective mAs was 2.1-fold higher, unadjusted CTDIvol 2.9-fold, and phase 2.5-fold for patients in the highest versus lowest size-adjusted DLP decile. There was virtually no change in kV (kilovolt). Automatic exposure control was widely used to modulate mAs, whereas kV modulation was rare. Phase was the strongest driver of between-protocols variation. Broad adoption of optimized protocols could result in total population dose reductions of 18.6-40%. CONCLUSION: There are large variations in radiation doses for routine abdomen CT unrelated to patient size. Modification of kV and single-phase scanning could result in substantial dose reduction. CLINICAL RELEVANCE: Radiation dose-optimization techniques for routine abdomen CT are routinely under-utilized leading to higher doses than needed. Greater modification of technical parameters and number of phases could result in substantial reduction in radiation exposure to patients. KEY POINTS: • Based on an analysis of 748,846 routine abdomen CT scans in adults, radiation doses varied tremendously across patients of the same size and optimization techniques were routinely under-utilized. • The difference in observed dose was due to variation in technical parameters and phase count. Automatic exposure control was commonly used to modify effective mAs, whereas kV was rarely adjusted for patient size. Routine abdomen CT should be performed using a single phase, yet multi-phase was common. • kV modulation by patient size and restriction to a single phase for routine abdomen indications could result in substantial reduction in radiation doses using well-established dose optimization approaches.

External validation, radiological evaluation, and development of deep learning automatic lung segmentation in contrast-enhanced chest CT.

OBJECTIVES: There is a need for CT pulmonary angiography (CTPA) lung segmentation models. Clinical translation requires radiological evaluation of model outputs, understanding of limitations, and identification of failure points. This multicentre study aims to develop an accurate CTPA lung segmentation model, with evaluation of outputs in two diverse patient cohorts with pulmonary hypertension (PH) and interstitial lung disease (ILD). METHODS: This retrospective study develops an nnU-Net-based segmentation model using data from two specialist centres (UK and USA). Model was trained (n = 37), tested (n = 12), and clinically evaluated (n = 176) on a diverse 'real-world' cohort of 225 PH patients with volumetric CTPAs. Dice score coefficient (DSC) and normalised surface distance (NSD) were used for testing. Clinical evaluation of outputs was performed by two radiologists who assessed clinical significance of errors. External validation was performed on heterogenous contrast and non-contrast scans from 28 ILD patients. RESULTS: A total of 225 PH and 28 ILD patients with diverse demographic and clinical characteristics were evaluated. Mean accuracy, DSC, and NSD scores were 0.998 (95% CI 0.9976, 0.9989), 0.990 (0.9840, 0.9962), and 0.983 (0.9686, 0.9972) respectively. There were no segmentation failures. On radiological review, 82% and 71% of internal and external cases respectively had no errors. Eighteen percent and 25% respectively had clinically insignificant errors. Peripheral atelectasis and consolidation were common causes for suboptimal segmentation. One external case (0.5%) with patulous oesophagus had a clinically significant error. CONCLUSION: State-of-the-art CTPA lung segmentation model provides accurate outputs with minimal clinical errors on evaluation across two diverse cohorts with PH and ILD. CLINICAL RELEVANCE: Clinical translation of artificial intelligence models requires radiological review and understanding of model limitations. This study develops an externally validated state-of-the-art model with robust radiological review. Intended clinical use is in techniques such as lung volume or parenchymal disease quantification. KEY POINTS: • Accurate, externally validated CT pulmonary angiography (CTPA) lung segmentation model tested in two large heterogeneous clinical cohorts (pulmonary hypertension and interstitial lung disease). • No segmentation failures and robust review of model outputs by radiologists found 1 (0.5%) clinically significant segmentation error. • Intended clinical use of this model is a necessary step in techniques such as lung volume, parenchymal disease quantification, or pulmonary vessel analysis.

A contrast-enhanced CT radiomics-based model to identify candidates for deintensified chemoradiotherapy in locoregionally advanced nasopharyngeal carcinoma patients.

OBJECTIVES: To develop a contrast-enhanced CT (CECT) radiomics-based model to identify locoregionally advanced nasopharyngeal carcinoma (LA-NPC) patients who would benefit from deintensified chemoradiotherapy. METHODS: LA-NPC patients who received low-dose concurrent cisplatin therapy (cumulative: 150 mg/m2), were randomly divided into training and validation groups. 107 radiomics features based on the primary nasopharyngeal tumor were extracted from each pre-treatment CECT scan. Through Cox regression analysis, a radiomics model and patients' corresponding radiomics scores were created with predictive independent radiomics features. T stage (T) and radiomics score (R) were compared as predictive factors. Combining the N stage (N), a clinical model (T + N), and a substitution model (R + N) were constructed. RESULTS: Training and validation groups consisted of 66 and 33 patients, respectively. Three significant independent radiomics features (flatness, mean, and gray level non-uniformity in gray level dependence matrix (GLDM-GLN)) were found. The radiomics score showed better predictive ability than the T stage (concordance index (C-index): 0.67 vs. 0.61, AUC: 0.75 vs. 0.60). The R + N model had better predictive performance and more effective risk stratification than the T + N model (C-index: 0.77 vs. 0.68, AUC: 0.80 vs. 0.70). The R + N model identified a low-risk group as deintensified chemoradiotherapy candidates in which no patient developed progression within 3 years, with 5-year progression-free survival (PFS) and overall survival (OS) both 90.7% (hazard ratio (HR) = 4.132, p = 0.018). CONCLUSION: Our radiomics-based model combining radiomics score and N stage can identify specific LA-NPC candidates for whom de-escalation therapy can be performed without compromising therapeutic efficacy. CLINICAL RELEVANCE STATEMENT: Our study shows that the radiomics-based model (R + N) can accurately stratify patients into different risk groups, with satisfactory prognosis in the low-risk group when treated with low-dose concurrent chemotherapy, providing new options for individualized de-escalation strategies. KEY POINTS: • A radiomics score, consisting of 3 predictive radiomics features (flatness, mean, and GLDM-GLN) integrated with the N stage, can identify specific LA-NPC populations for deintensified treatment. • In the selection of LA-NPC candidates for de-intensified treatment, radiomics score extracted from primary nasopharyngeal tumors based on CECT can be superior to traditional T stage classification as a predictor.

Deep learning-based assessment of CT markers of sarcopenia and myosteatosis for outcome assessment in patients with advanced pancreatic cancer after high-intensity focused ultrasound treatment.

OBJECTIVES: To evaluate the prognostic value of CT-based markers of sarcopenia and myosteatosis in comparison to the Eastern Cooperative Oncology Group (ECOG) score for survival of patients with advanced pancreatic cancer treated with high-intensity focused ultrasound (HIFU). MATERIALS AND METHODS: For 142 retrospective patients, the skeletal muscle index (SMI), skeletal muscle radiodensity (SMRD), fatty muscle fraction (FMF), and intermuscular fat fraction (IMFF) were determined on superior mesenteric artery level in pre-interventional CT. Each marker was tested for associations with sex, age, body mass index (BMI), and ECOG. The prognostic value of the markers was examined in Kaplan-Meier analyses with the log-rank test and in uni- and multivariable Cox proportional hazards (CPH) models. RESULTS: The following significant associations were observed: Male patients had higher BMI and SMI. Patients with lower ECOG had lower BMI and SMI. Patients with BMI lower than 21.8 kg/m2 (median) also showed lower SMI and IMFF. Patients younger than 63.3 years (median) were found to have higher SMRD, lower FMF, and lower IMFF. In the Kaplan-Meier analysis, significantly lower survival times were observed in patients with higher ECOG or lower SMI. Increased patient risk was observed for higher ECOG, lower BMI, and lower SMI in univariable CPH analyses for 1-, 2-, and 3-year survival. Multivariable CPH analysis for 1-year survival revealed increased patient risk for higher ECOG, lower SMI, lower IMFF, and higher FMF. In multivariable analysis for 2- and 3-year survival, only ECOG and FMF remained significant. CONCLUSION: CT-based markers of sarcopenia and myosteatosis show a prognostic value for assessment of survival in advanced pancreatic cancer patients undergoing HIFU therapy. CLINICAL RELEVANCE STATEMENT: The results indicate a greater role of myosteatosis for additional risk assessment beyond clinical scores, as only FMF was associated with long-term survival in multivariable CPH analyses along ECOG and also showed independence to ECOG in group analysis. KEY POINTS: • This study investigates the prognostic value of CT-based markers of sarcopenia and myosteatosis for patients with pancreatic cancer treated with high-intensity focused ultrasound. • Markers for sarcopenia and myosteatosis showed a prognostic value besides clinical assessment of the physical status by the Eastern Cooperative Oncology Group score. In contrast to muscle size measurements, the myosteatosis marker fatty muscle fraction demonstrated independence to the clinical score. • The results indicate that myosteatosis might play a greater role for additional patient risk assessments beyond clinical assessments of physical status.

Feasibility of spectral CT-derived extracellular volume fraction for differentiating aldosterone-producing from nonfunctioning adrenal nodules.

OBJECTIVE: To assess the feasibility of spectral CT-derived extracellular volume (ECV) for differentiating aldosterone-producing nodules (APN) from nonfunctioning adrenal nodules (NFN). METHODS: Sixty-nine patients with biochemically and histologically confirmed unilateral APN (34) and NFN (35) as well as 23 patients with bilateral APN (19) and NFN (27) confirmed biochemically and by adrenal vein sampling (AVS) were enrolled in this retrospective study from October 2020 to April 2022. All patients underwent contrast-enhanced spectral CT of the adrenal glands with a 10-min delayed phase. The haematocrit level was measured within 2 days of CT. An iodine density map was derived from the delayed CT. The ECV fractions of the APN and NFN were calculated and compared in the test cohort of 69 patients with unilateral adrenal nodules. The optimal cut-off value was determined to evaluate the diagnostic efficacy of the ECV fraction for differentiating APN from NFN in the validation cohort of 23 patients with bilateral adrenal nodules. RESULTS: The ECV fractions of the APN (11.17 ± 4.57%) were significantly lower (p < 0.001) than that of the NFN (24.79 ± 6.01%) in the test cohort. At cut-off ECV value of 17.16%, the optimal area under the receiver operating characteristic curve was 0.974 (95% confidence interval: 0.942-1) with 91.4% sensitivity, 93.9% specificity, and 92.8% accuracy in the test cohort and 89.5% sensitivity, 96.3% specificity, and 93.5% accuracy in the validation cohort for differentiating APN from NFN. CONCLUSION: The spectral CT-derived ECV fraction can differentiate APN from NFN with high diagnostic performance. CLINICAL RELEVANCE STATEMENT: Spectral CT-derived extracellular volume fraction could accurately differentiate between adrenal aldosterone-producing nodules and nonfunctioning nodules. It might serve as a noninvasive alternative to adrenal vein sampling in primary aldosteronism patients with bilateral adrenal nodules. KEY POINTS: • Conventional CT cannot differentiate aldosterone-producing adrenal nodules from nonfunctioning nodules. • Extracellular volume of adrenal aldosterone-producing nodules was significantly lower than that of nonfunctioning nodules and normal adrenal glands. It can accurately differentiate between aldosterone-producing and nonfunctioning adrenal nodules. • Extracellular volume may be a novel, noninvasive biomarker alternative to adrenal vein sampling for determining the functional status of bilateral adrenal nodules in patients with primary aldosteronism.

An automatic texture feature analysis framework of renal tumor: surgical, pathological, and molecular evaluation based on multi-phase abdominal CT.

OBJECTIVES: To determine whether the texture feature analysis of multi-phase abdominal CT can provide a robust prediction of benign and malignant, histological subtype, pathological stage, nephrectomy risk, pathological grade, and Ki67 index in renal tumor. METHODS: A total of 1051 participants with renal tumor were split into the internal cohort (850 patients from four different hospitals) and the external testing cohort (201 patients from another local hospital). The proposed framework comprised a 3D-kidney and tumor segmentation model by 3D-UNet, a feature extractor for the regions of interest based on radiomics and image dimension reduction, and the six classifiers by XGBoost. A quantitative model interpretation method called SHAP was used to explore the contribution of each feature. RESULTS: The proposed multi-phase abdominal CT model provides robust prediction for benign and malignant, histological subtype, pathological stage, nephrectomy risk, pathological grade, and Ki67 index in the internal validation set, with the AUROC values of 0.88 ± 0.1, 0.90 ± 0.1, 0.91 ± 0.1, 0.89 ± 0.1, 0.84 ± 0.1, and 0.88 ± 0.1, respectively. The external testing set also showed impressive results, with AUROC values of 0.83 ± 0.1, 0.83 ± 0.1, 0.85 ± 0.1, 0.81 ± 0.1, 0.79 ± 0.1, and 0.81 ± 0.1, respectively. The radiomics feature including the first-order statistics, the tumor size-related morphology, and the shape-related tumor features contributed most to the model predictions. CONCLUSIONS: Automatic texture feature analysis of abdominal multi-phase CT provides reliable predictions for multi-tasks, suggesting the potential usage of clinical application. CLINICAL RELEVANCE STATEMENT: The automatic texture feature analysis framework, based on multi-phase abdominal CT, provides robust and reliable predictions for multi-tasks. These valuable insights can serve as a guiding tool for clinical diagnosis and treatment, making medical imaging an essential component in the process. KEY POINTS: • The automatic texture feature analysis framework based on multi-phase abdominal CT can provide more accurate prediction of benign and malignant, histological subtype, pathological stage, nephrectomy risk, pathological grade, and Ki67 index in renal tumor. • The quantitative decomposition of the prediction model was conducted to explore the contribution of the extracted feature. • The study involving 1051 patients from 5 medical centers, along with a heterogeneous external data testing strategy, can be seamlessly transferred to various tasks involving new datasets.

Intra-patient variability of iodine quantification across different dual-energy CT platforms: assessment of normalization techniques.

OBJECTIVES: To investigate intra-patient variability of iodine concentration (IC) between three different dual-energy CT (DECT) platforms and to test different normalization approaches. METHODS: Forty-four patients who underwent portal venous phase abdominal DECT on a dual-source (dsDECT), a rapid kVp switching (rsDECT), and a dual-layer detector platform (dlDECT) during cancer follow-up were retrospectively included. IC in the liver, pancreas, and kidneys and different normalized ICs (NICPV:portal vein; NICAA:abdominal aorta; NICALL:overall iodine load) were compared between the three DECT scanners for each patient. A longitudinal mixed effects analysis was conducted to elucidate the effect of the scanner type, scan order, inter-scan time, and contrast media amount on normalized iodine concentration. RESULTS: Variability of IC was highest in the liver (dsDECT vs. dlDECT 28.96 (14.28-46.87) %, dsDECT vs. rsDECT 29.08 (16.59-62.55) %, rsDECT vs. dlDECT 22.85 (7.52-33.49) %), and lowest in the kidneys (dsDECT vs. dlDECT 15.76 (7.03-26.1) %, dsDECT vs. rsDECT 15.67 (8.86-25.56) %, rsDECT vs. dlDECT 10.92 (4.92-22.79) %). NICALL yielded the best reduction of IC variability throughout all tissues and inter-scanner comparisons, yet did not reduce the variability between dsDECT vs. dlDECT and rsDECT, respectively, in the liver. The scanner type remained a significant determinant for NICALL in the pancreas and the liver (F-values, 12.26 and 23.78; both, p < 0.0001). CONCLUSIONS: We found tissue-specific intra-patient variability of IC across different DECT scanner types. Normalization mitigated variability by reducing physiological fluctuations in iodine distribution. After normalization, the scanner type still had a significant effect on iodine variability in the pancreas and liver. CLINICAL RELEVANCE STATEMENT: Differences in iodine quantification between dual-energy CT scanners can partly be mitigated by normalization, yet remain relevant for specific tissues and inter-scanner comparisons, which should be taken into account at clinical routine imaging. KEY POINTS: • Iodine concentration showed the least variability between scanner types in the kidneys (range 10.92-15.76%) and highest variability in the liver (range 22.85-29.08%). • Normalizing tissue-specific iodine concentrations against the overall iodine load yielded the greatest reduction of variability between scanner types for 2/3 inter-scanner comparisons in the liver and for all (3/3) inter-scanner comparisons in the kidneys and pancreas, respectively. • However, even after normalization, the dual-energy CT scanner type was found to be the factor significantly influencing variability of iodine concentration in the liver and pancreas.

Net water uptake as a predictive neuroimaging marker for acute ischemic stroke outcomes: a meta-analysis.

OBJECTIVE: To assess the role of net water uptake (NWU) in predicting outcomes in acute ischemic stroke (AIS) patients. METHODS: A systematic review and meta-analysis were performed, adhering to established guidelines. The search covered PubMed, Scopus, Web of Science, and Embase databases until July 1, 2023. Eligible studies reporting quantitative ischemic lesion NWU in admission CT scans of AIS patients, stratified based on outcomes, were included. Data analysis was performed using R software version 4.2.1. RESULTS: Incorporating 17 original studies with 2217 AIS patients, NWU was significantly higher in patients with poor outcomes compared to those with good outcomes (difference of medians: 5.06, 95% CI: 3.00-7.13, p < 0.001). Despite excluding one outlier study, considerable heterogeneity persisted among the included studies (I2 = 90.8%). The meta-regression and subgroup meta-analyses demonstrated significantly higher NWU in patients with poor functional outcome, as assessed by modified Rankin Scale (difference of medians: 3.83, 95% CI: 1.98-5.68, p < 0.001, I2 = 72.9%), malignant edema/infarct (difference of medians: 8.30, 95% CI: 4.01-12.58, p < 0.001, I2 = 95.6%), and intracranial hemorrhage (difference of medians: 5.43, 95% CI: 0.44-10.43, p = 0.03, I2 = 91.1%). CONCLUSION: NWU on admission CT scans shows promise as a predictive marker for outcomes in AIS patients. Prospective, multicenter trials with standardized, automated NWU measurement are crucial for robustly predicting diverse clinical outcomes. CLINICAL RELEVANCE STATEMENT: The potential of net water uptake as a biomarker for predicting outcomes in acute ischemic stroke patients holds significant promise. Further validation through additional research could lead to its integration into clinical practice, potentially improving the accuracy of clinical decision-making and allowing for the development of more precise patient care strategies. KEY POINTS: • Net water uptake, a CT-based biomarker, quantifies early brain edema after acute ischemic stroke. • Net water uptake is significantly higher in poor outcome acute ischemic stroke patients. • Net water uptake on CT scans holds promise in predicting diverse acute ischemic stroke outcomes.

CT-based deep learning model for predicting hospital discharge outcome in spontaneous intracerebral hemorrhage.

OBJECTIVES: To predict the functional outcome of patients with intracerebral hemorrhage (ICH) using deep learning models based on computed tomography (CT) images. METHODS: A retrospective, bi-center study of ICH patients was conducted. Firstly, a custom 3D convolutional model was built for predicting the functional outcome of ICH patients based on CT scans from randomly selected ICH patients in H training dataset collected from H hospital. Secondly, clinical data and radiological features were collected at admission and the Extreme Gradient Boosting (XGBoost) algorithm was used to establish a second model, named the XGBoost model. Finally, the Convolution model and XGBoost model were fused to build the third "Fusion model." Favorable outcome was defined as modified Rankin Scale score of 0-3 at discharge. The prognostic predictive accuracy of the three models was evaluated using an H test dataset and an external Y dataset, and compared with the performance of ICH score and ICH grading scale (ICH-GS). RESULTS: A total of 604 patients with ICH were included in this study, of which 450 patients were in the H training dataset, 50 patients in the H test dataset, and 104 patients in the Y dataset. In the Y dataset, the areas under the curve (AUCs) of the Convolution model, XGBoost model, and Fusion model were 0.829, 0.871, and 0.905, respectively. The Fusion model prognostic performance exceeded that of ICH score and ICH-GS (p = 0.043 and p = 0.045, respectively). CONCLUSIONS: Deep learning models have good accuracy for predicting functional outcome of patients with spontaneous intracerebral hemorrhage. CLINICAL RELEVANCE STATEMENT: The proposed deep learning Fusion model may assist clinicians in predicting functional outcome and developing treatment strategies, thereby improving the survival and quality of life of patients with spontaneous intracerebral hemorrhage. KEY POINTS: • Integrating clinical presentations, CT images, and radiological features to establish deep learning model for functional outcome prediction of patients with intracerebral hemorrhage. • Deep learning applied to CT images provides great help in prognosing functional outcome of intracerebral hemorrhage patients. • The developed deep learning model performs better than clinical prognostic scores in predicting functional outcome of patients with intracerebral hemorrhage.

Semiquantitative index of symptomatic minor instability of the lateral elbow at CT arthrography (SMILE index): clinical applicability and reproducibility study.

OBJECTIVE: To assess the applicability of a semiquantitative index for symptomatic minor instability of the lateral elbow (SMILE). MATERIALS AND METHODS: CT arthrograms of consecutive patients with lateral elbow pain who underwent ultrasound-guided CT arthrography at our orthopedic center between April 2019 and May 2022 were included. Images were acquired at 100 kVp and 80 mAs. An expert radiologist (R1) and a radiology resident (R2) retrospectively performed an independent, blinded evaluation of the arthrograms to assess the presence of imaging findings suggestive of elbow instability. The SMILE index (0-8) was obtained adding (I) radial head chondromalacia (0 - 1); (II) humeral capitellum chondromalacia (0 - 1); (III) humeral trochlear ridge chondromalacia (0 - 1); (IV) annular ligament laxity (0 - 2); (V) synovial thickening (0 - 1); (VI) humeroradial joint asymmetry (0 - 1); and (VII) capsular tear (0 - 1). R1 repeated the assessment after 14 days. Cohen's weighted κ statistic and raw concordance were used to appraise reproducibility. RESULTS: Eighty patients (median age 49 years, interquartile range 40-53 years, 49, 61% males) underwent CT arthrography at our center, and 10 (12%) of them underwent bilateral elbow examination, leading to 90 included CT arthrograms. Median SMILE index was 4 (IQR: 2-5) for R1, 4 (IQR: 2-5) for R2, and 4 (IQR: 2-5) for the second assessment by R1. Intra-reader agreement was excellent (κ = 0.94, concordance 87%), while inter-reader agreement was substantial (κ = 0.75, concordance 67%). CONCLUSION: The proposed SMILE index showed good reproducibility; further studies are warranted to correlate our index with clinical and surgical data. CLINICAL RELEVANCE STATEMENT: Our scoring system allows a standardized evaluation of patients with lateral elbow pain and instability suitable for application into clinical practice, complementing the orthopedic surgeon's clinical diagnosis with imaging findings that may aid treatment choices. KEY POINTS: • Lateral elbow pain is often interpreted clinically as lateral epicondylitis, but it can also encompass intra-articular pathology. • The proposed arthrographic index allows comprehensive quantification of lateral elbow pathology with good reproducibility and application times. • Our index provides the orthopedic surgeon with information regarding intra-articular findings, aiding treatment choices.

Prediction of high Ki-67 proliferation index of gastrointestinal stromal tumors based on CT at non-contrast-enhanced and different contrast-enhanced phases.

OBJECTIVES: To evaluate and analyze radiomics models based on non-contrast-enhanced computed tomography (CT) and different phases of contrast-enhanced CT in predicting Ki-67 proliferation index (PI) among patients with pathologically confirmed gastrointestinal stromal tumors (GISTs). METHODS: A total of 383 patients with pathologically proven GIST were divided into a training set (n = 218, vendor 1) and 2 validation sets (n = 96, vendor 2; n = 69, vendors 3-5). Radiomics features extracted from the most recent non-contrast-enhanced and three contrast-enhanced CT scan prior to pathological examination. Random forest models were trained for each phase to predict tumors with high Ki-67 proliferation index (Ki-67>10%) and were evaluated using the area under the receiver operating characteristic curve (AUC) and other metrics on the validation sets. RESULTS: Out of 107 radiomics features extracted from each phase of CT images, four were selected for analysis. The model trained using the non-contrast-enhanced phase achieved an AUC of 0.792 in the training set and 0.822 and 0.711 in the two validation sets, similar to models trained on different contrast-enhanced phases (p > 0.05). Several relevant features, including NGTDM Busyness and tumor size, remained predictive in non-contrast-enhanced and different contrast-enhanced images. CONCLUSION: The results of this study indicate that a radiomics model based on non-contrast-enhanced CT matches that of models based on different phases of contrast-enhanced CT in predicting the Ki-67 PI of GIST. GIST may exhibit similar radiological patterns irrespective of the use of contrast agent, and such radiomics features may help quantify these patterns to predict Ki-67 PI of GISTs. CLINICAL RELEVANCE STATEMENT: GIST may exhibit similar radiomics patterns irrespective of contrast agent; thus, radiomics models based on non-contrast-enhanced CT could be an alternative for risk stratification in GIST patients with contraindication to contrast agent. KEY POINTS: • Performance of radiomics models in predicting Ki-67 proliferation based on different CT phases is evaluated. • Non-contrast-enhanced CT-based radiomics models performed similarly to contrast-enhanced CT in risk stratification in GIST patients. • NGTDM Busyness remains stable to contrast agents in GISTs in radiomics models.

Effects of a comprehensive brain computed tomography deep learning model on radiologist detection accuracy.

OBJECTIVES: Non-contrast computed tomography of the brain (NCCTB) is commonly used to detect intracranial pathology but is subject to interpretation errors. Machine learning can augment clinical decision-making and improve NCCTB scan interpretation. This retrospective detection accuracy study assessed the performance of radiologists assisted by a deep learning model and compared the standalone performance of the model with that of unassisted radiologists. METHODS: A deep learning model was trained on 212,484 NCCTB scans drawn from a private radiology group in Australia. Scans from inpatient, outpatient, and emergency settings were included. Scan inclusion criteria were age ≥ 18 years and series slice thickness ≤ 1.5 mm. Thirty-two radiologists reviewed 2848 scans with and without the assistance of the deep learning system and rated their confidence in the presence of each finding using a 7-point scale. Differences in AUC and Matthews correlation coefficient (MCC) were calculated using a ground-truth gold standard. RESULTS: The model demonstrated an average area under the receiver operating characteristic curve (AUC) of 0.93 across 144 NCCTB findings and significantly improved radiologist interpretation performance. Assisted and unassisted radiologists demonstrated an average AUC of 0.79 and 0.73 across 22 grouped parent findings and 0.72 and 0.68 across 189 child findings, respectively. When assisted by the model, radiologist AUC was significantly improved for 91 findings (158 findings were non-inferior), and reading time was significantly reduced. CONCLUSIONS: The assistance of a comprehensive deep learning model significantly improved radiologist detection accuracy across a wide range of clinical findings and demonstrated the potential to improve NCCTB interpretation. CLINICAL RELEVANCE STATEMENT: This study evaluated a comprehensive CT brain deep learning model, which performed strongly, improved the performance of radiologists, and reduced interpretation time. The model may reduce errors, improve efficiency, facilitate triage, and better enable the delivery of timely patient care. KEY POINTS: • This study demonstrated that the use of a comprehensive deep learning system assisted radiologists in the detection of a wide range of abnormalities on non-contrast brain computed tomography scans. • The deep learning model demonstrated an average area under the receiver operating characteristic curve of 0.93 across 144 findings and significantly improved radiologist interpretation performance. • The assistance of the comprehensive deep learning model significantly reduced the time required for radiologists to interpret computed tomography scans of the brain.

Added value of arterial enhancement fraction derived from dual-energy computed tomography for preoperative diagnosis of cervical lymph node metastasis in papillary thyroid cancer: initial results.

OBJECTIVES: To explore the added value of arterial enhancement fraction (AEF) derived from dual-energy computed tomography CT (DECT) to conventional image features for diagnosing cervical lymph node (LN) metastasis in papillary thyroid cancer (PTC). METHODS: A total of 273 cervical LNs (153 non-metastatic and 120 metastatic) were recruited from 92 patients with PTC. Qualitative image features of LNs were assessed. Both single-energy CT (SECT)-derived AEF (AEFS) and DECT-derived AEF (AEFD) were calculated. Correlation between AEFD and AEFS was determined using Pearson's correlation coefficient. Multivariate logistic regression analysis with the forward variable selection method was used to build three models (conventional features, conventional features + AEFS, and conventional features + AEFD). Diagnostic performances were evaluated using receiver operating characteristic (ROC) curve analyses. RESULTS: Abnormal enhancement, calcification, and cystic change were chosen to build model 1 and the model provided moderate diagnostic performance with an area under the ROC curve (AUC) of 0.675. Metastatic LNs demonstrated both significantly higher AEFD (1.14 vs 0.48; p < 0.001) and AEFS (1.08 vs 0.38; p < 0.001) than non-metastatic LNs. AEFD correlated well with AEFS (r = 0.802; p < 0.001), and exhibited comparable performance with AEFS (AUC, 0.867 vs 0.852; p = 0.628). Combining CT image features with AEFS (model 2) and AEFD (model 3) could significantly improve diagnostic performances (AUC, 0.865 vs 0.675; AUC, 0.883 vs 0.675; both p < 0.001). CONCLUSIONS: AEFD correlated well with AEFS, and exhibited comparable performance with AEFS. Integrating qualitative CT image features with both AEFS and AEFD could further improve the ability in diagnosing cervical LN metastasis in PTC. CLINICAL RELEVANCE STATEMENT: Arterial enhancement fraction (AEF) values, especially AEF derived from dual-energy computed tomography, can help to diagnose cervical lymph node metastasis in patients with papillary thyroid cancer, and complement conventional CT image features for improved clinical decision making. KEY POINTS: • Metastatic cervical lymph nodes (LNs) demonstrated significantly higher arterial enhancement fraction (AEF) derived from dual-energy computed tomography (DECT) and single-energy CT (SECT)-derived AEF (AEFS) than non-metastatic LNs in patients with papillary thyroid cancer. • DECT-derived AEF (AEFD) correlated significantly with AEFS, and exhibited comparable performance with AEFS. • Integrating qualitative CT images features with both AEFS and AEFD could further improve the differential ability.