Non-contrast CT Scan Research Articles

Emphysematous Pyelonephritis: Nephrectomy is not necessary, a narrative review

Emphysematous pyelonephritis (EPN) is a severe, life-endangering, necrotising, gas-producing infection affecting the kidney and perinephric tissue. It is associated with renal loss, high morbidity, and mortality. Diabetes, obstructive uropathy, female gender and hypertension are the most common risk factors. Escherichia coli, Klebsiella pneumoniae and Proteus mirabilis are the most common pathogens causing Emphysematous pyelonephritis. EPN presents as a clinical triad of fever, flank pain, and nausea. Non-contrast CT scan is the mainstay of early diagnosis of EPN. It is diagnosed by demonstrating gas in renal parenchyma and perinephric tissue. Huang and Sang's classification is used for the classification and management of EPN. Treatment of EPN progressed from early nephrectomy to minimally invasive treatments including newer and advanced antibiotics and percutaneous drainage along with Percutaneous nephrostomy and DJ stenting. Early nephrectomy was the treatment of choice in the last century however newer studies suggest conservative treatment is associated with renal preservation, decreased morbidity and mortality. The initial use of broad-spectrum antibiotics such as Third-or fourth generation cephalosporins and carbapenems is recommended. Emergency drainage is indicated with a larger-sized pigtail catheter as pus is thick and frequent blockage of tubes is common. Nephrectomy should be done for those patients who fail to respond to conservative therapy i.e., antibiotics and percutaneous drainage. This review aims to provide comprehensive, evidence-based recommendations for the diagnosis and management of this life-threatening urological infection Emphysematous pyelonephritis. Keywords: Emphysematous pyelonephritis, Diabetes, Antibiotics, Conservative, Percutaneous drainage, Nephrectomy

Zero-Contrast Transcatheter Aortic Valve Implantation vs. Standard Practice: Periprocedural and Long-Term Clinical Outcomes.

Background: We aimed to compare the procedural efficacy and long-term clinical results of a totally contrast-free Transcatheter Aortic Valve Implantation (TAVI) procedure (i.e., contrast dye was not used for either the pre-procedural assessment or during the procedure) to those of standard practice in patients with severe renal dysfunction. Methods: All consecutive patients with a glomerular filtration rate (GFR) ≤ 35 mL/min and severe aortic stenosis who were treated with transfemoral TAVI at our Institution were included in the registry. The zero-contrast patients underwent carbon dioxide angiography and a non-contrast CT scan for assessment of vascular access suitability, and aortic annulus sizing was performed by a TEE, and the procedural guidance was fluoroscopic and echocardiographic. Procedural outcomes were evaluated, and clinical long-term follow-up was performed for all included patients. Results: A total of 44 patients (median age, 85 (IQR, 80.75-87.00)) were included in the zero-contrast group (TEE guidance and general anesthesia in 37 (84%) patients), while 63 patients were included in the standard practice arm (82 ± 78 mL of contrast dye used). Procedural success was obtained in 100% of cases. There were no differences in procedural outcomes, including final mean aortic gradients (5.5 (IQR, 5.0-10.0) mmHg in the zero-contrast group vs. 6.0 (IQR, 5.0-10.0) mmHg in the standard practice group) and rate of at least a moderate paravalvular leak (0% vs. 1.6% in the zero-contrast and standard practice groups, respectively; p = 0.31). No differences in AKI during the hospital stay were observed. Over a median follow-up of 3.3 years, there was a significantly lower rate of AKI (1.2% vs. 25.9%, p < 0.001) and rehospitalizations (1.6% vs. 35.5%, p < 0.00) in standard practice group. Conclusions: We showed for the first time the feasibility and efficacy of a totally contrast-free strategy compared to standard practice in TAVI patients with severe renal dysfunction. Besides achieving comparable procedural results, the zero-contrast strategy showed a better long-term clinical outcome in reducing hospital readmissions for kidney function deterioration.

Comparative Prognostic Value of Coronary Calcium Score and Perivascular Fat Attenuation Index in Coronary Artery Disease.

Coronary artery disease (CAD) is the leading global cause of mortality, accounting for approximately 30% of all deaths. It is primarily characterized by the accumulation of atherosclerotic plaques within the coronary arteries, leading to reduced blood flow to the heart muscle. Early detection of atherosclerotic plaques is crucial to prevent major adverse cardiac events. Notably, recent studies have shown that 15% of myocardial infarctions occur in patients with non-obstructive CAD, underscoring the importance of comprehensive plaque assessment beyond merely identifying obstructive lesions. Cardiac Computed Tomography Angiography (CCTA) has emerged as a cost-effective and efficient technique for excluding obstructive CAD, particularly in patients with a low-to-intermediate clinical likelihood of the disease. Recent advancements in CCTA technology, such as improved resolution and reduced scan times, have mitigated many technical challenges, allowing for precise quantification and characterization of both calcified and non-calcified atherosclerotic plaques. This review focuses on two critical physiological aspects of atherosclerotic plaques: the burden of calcifications, assessed via the coronary artery calcium score (CACs), and perivascular fat attenuation index (pFAI), an emerging marker of vascular inflammation. The CACs, obtained through non-contrast CT scans, quantifies calcified plaque burden and is widely used to stratify cardiovascular risk, particularly in asymptomatic patients. Despite its prognostic value, the CACs does not provide information on non-calcified plaques or inflammatory status. In contrast, the pFAI, derived from CCTA, serves as an indirect marker of coronary inflammation and has shown potential in predicting adverse cardiac events. Combining both CACs and pFAI assessment could offer a comprehensive risk stratification approach, integrating the established calcification burden with novel inflammatory markers to enhance CAD prevention and management strategies.

Deep Learning Segmentation of Ascites on Abdominal CT Scans for Automatic Volume Quantification.

Purpose To evaluate the performance of an automated deep learning method in detecting ascites and subsequently quantifying its volume in patients with liver cirrhosis and patients with ovarian cancer. Materials and Methods This retrospective study included contrast-enhanced and noncontrast abdominal-pelvic CT scans of patients with cirrhotic ascites and patients with ovarian cancer from two institutions, National Institutes of Health (NIH) and University of Wisconsin (UofW). The model, trained on The Cancer Genome Atlas Ovarian Cancer dataset (mean age [±SD], 60 years ± 11; 143 female), was tested on two internal datasets (NIH-LC and NIH-OV) and one external dataset (UofW-LC). Its performance was measured by the F1/Dice coefficient, SDs, and 95% CIs, focusing on ascites volume in the peritoneal cavity. Results On NIH-LC (25 patients; mean age, 59 years ± 14; 14 male) and NIH-OV (166 patients; mean age, 65 years ± 9; all female), the model achieved F1/Dice scores of 85.5% ± 6.1 (95% CI: 83.1, 87.8) and 82.6% ± 15.3 (95% CI: 76.4, 88.7), with median volume estimation errors of 19.6% (IQR, 13.2%-29.0%) and 5.3% (IQR: 2.4%-9.7%), respectively. On UofW-LC (124 patients; mean age, 46 years ± 12; 73 female), the model had a F1/Dice score of 83.0% ± 10.7 (95% CI: 79.8, 86.3) and median volume estimation error of 9.7% (IQR, 4.5%-15.1%). The model showed strong agreement with expert assessments, with r2 values of 0.79, 0.98, and 0.97 across the test sets. Conclusion The proposed deep learning method performed well in segmenting and quantifying the volume of ascites in patients with cirrhosis and those with ovarian cancer, in concordance with expert radiologist assessments. Keywords: Abdomen/GI, Cirrhosis, Deep Learning, Segmentation Supplemental material is available for this article. © RSNA, 2024 See also commentary by Aisen and Rodrigues in this issue.

Lateral intraventricular epidermoid cyst: A case report

Background: Epidermoid cysts originate from ectopic embryonic epithelial cells and are a very common type of benign intracranial tumor. Epidermoid tumors located in the lateral ventricle are exceedingly rare. Case description: We present a rare case of a 55-year-old man complaining of a recurrent headache over a 2 year-period. A non-contrast CT scan and MRI of the head revealed a large mass within the right lateral ventricle. Gross total resection of the lesion was retained but unfortunately the patient refused surgery. The prognosis for patients who undergo surgery is excellent and depends on the quality of tumor removal.

Precise Image-level Localization of Intracranial Hemorrhage on Head CT Scans with Deep Learning Models Trained on Study-level Labels.

"Just Accepted" papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content. Purpose To develop a highly generalizable weakly supervised model to automatically detect and localize image- level intracranial hemorrhage (ICH) using study-level labels. Materials and Methods In this retrospective study, the proposed model was pretrained on the image-level RSNA dataset and fine-tuned on a local dataset using attention-based bidirectional long-short-term memory networks. This local training dataset included 10,699 noncontrast head CT scans from 7469 patients with ICH study-level labels extracted from radiology reports. Model performance was compared with that of two senior neuroradiologists on 100 random test scans using the McNemar test, and its generalizability was evaluated on an external independent dataset. Results The model achieved a positive predictive value (PPV) of 85.7% (95% CI: [84.0%, 87.4%]) and an AUC of 0.96 (95% CI: [0.96, 0.97]) on the held-out local test set (n = 7243, 3721 female) and 89.3% (95% CI: [87.8%, 90.7%]) and 0.96 (95% CI: [0.96, 0.97]), respectively, on the external test set (n = 491, 178 female). For 100 randomly selected samples, the model achieved performance on par with two neuroradiologists, but with a significantly faster (P < .05) diagnostic time of 5.04 seconds per scan (versus 86 seconds and 22.2 seconds for the two neuroradiologists, respectively). The model's attention weights and heatmaps visually aligned with neuroradiologists' interpretations. Conclusion The proposed model demonstrated high generalizability and high PPVs, offering a valuable tool for expedited ICH detection and prioritization while reducing false-positive interruptions in radiologists' workflows. ©RSNA, 2024.

Influence of contrast medium on long-term renal function and outcomes in patients with septic acute kidney injury: A propensity-matched cohort study

PurposeTo investigate the relationship between contrast medium administration and long-term mortality and renal function in patients with septic acute kidney injury (AKI). Materials and methodsWe performed a retrospective, propensity-matched cohort study involving 1521 adult patients admitted with septic shock. Patients with septic AKI who underwent contrast or non-contrast CT scans were enrolled. The primary outcomes were the rates of 90-day mortality and dialysis within 90 days. The secondary outcomes included worsening of AKI, in-hospital mortality, and maintenance of dialysis after 90 days. ResultsDuring the study period, 609 patients with septic AKI were identified; 220 (36.1%) underwent contrast CT and 389 (63.9%) underwent non-contrast CT. After propensity score matching, 133 pairs were obtained. There were no significant differences between the contrast and non-contrast CT groups in 90-day mortality (54.9% vs. 58.6%, P = 0.579), dialysis within 90 days (6.8% vs. 8.3%, P = 0.655), worsening AKI (2.3% vs. 3.0%, P = 0.706), in-hospital mortality (10.6% vs. 14.4%, P = 0.369), or maintenance of dialysis after 90 days (0.0% vs. 0.8%, P > 0.99). ConclusionsThe administration of intravenous contrast medium was not associated with long-term mortality, deterioration of renal function, or dialysis in patients with septic AKI.

A 3D boundary-guided hybrid network with convolutions and Transformers for lung tumor segmentation in CT images

—Accurate lung tumor segmentation from Computed Tomography (CT) scans is crucial for lung cancer diagnosis. Since the 2D methods lack the volumetric information of lung CT images, 3D convolution-based and Transformer-based methods have recently been applied in lung tumor segmentation tasks using CT imaging. However, most existing 3D methods cannot effectively collaborate the local patterns learned by convolutions with the global dependencies captured by Transformers, and widely ignore the important boundary information of lung tumors. To tackle these problems, we propose a 3D boundary-guided hybrid network using convolutions and Transformers for lung tumor segmentation, named BGHNet. In BGHNet, we first propose the Hybrid Local-Global Context Aggregation (HLGCA) module with parallel convolution and Transformer branches in the encoding phase. To aggregate local and global contexts in each branch of the HLGCA module, we not only design the Volumetric Cross-Stripe Window Transformer (VCSwin-Transformer) to build the Transformer branch with local inductive biases and large receptive fields, but also design the Volumetric Pyramid Convolution with transformer-based extensions (VPConvNeXt) to build the convolution branch with multi-scale global information. Then, we present a Boundary-Guided Feature Refinement (BGFR) module in the decoding phase, which explicitly leverages the boundary information to refine multi-stage decoding features for better performance. Extensive experiments were conducted on two lung tumor segmentation datasets, including a private dataset (HUST-Lung) and a public benchmark dataset (MSD-Lung). Results show that BGHNet outperforms other state-of-the-art 2D or 3D methods in our experiments, and it exhibits superior generalization performance in both non-contrast and contrast-enhanced CT scans.

Development and validation of a model predicting adrenal lipid-poor adenoma based on the minimum attenuation value from non-contrast CT: a dual-center retrospective study

ObjectiveThe early differentiation of adrenal lipid-poor adenomas from non-adenomas is a crucial step in reducing excessive examinations and treatments. This study seeks to construct an eXtreme Gradient Boosting (XGBoost) predictive model utilizing the minimum attenuation values (minAVs) from non-contrast CT (NCCT) scans to identify lipid-poor adenomas.Materials and methodsRetrospective analysis encompassed clinical data, minAVs, CT histogram (CTh), mean attenuation values (meanAVs), and lesion diameter from patients with pathologically or clinically confirmed adrenal lipid-poor adenomas across two medical institutions, juxtaposed with non-adenomas. Variable selection transpired in Institution A (training set), with XGBoost models established based on minAVs and CTh separately. Institution B (validation set) corroborated the diagnostic efficacy of the two models. Receiver operator characteristic (ROC) curve analysis, calibration curves, and Brier scores assessed the diagnostic performance and calibration of the models, with the Delong test gauging differences in the area under the curve (AUC) between models. SHapley Additive exPlanations (SHAP) values elucidated and visualized the models.ResultsThe training set comprised 136 adrenal lipid-poor adenomas and 126 non-adenomas, while the validation set included 46 and 40 instances, respectively. In the training set, there were substantial inter-group differences in minAVs, CTh, meanAVs, diameter, and body mass index (BMI) (p < 0.05 for all). The AUC for the minAV and CTh models were 0.912 (95% confidence interval [CI]: 0.866–0.957) and 0.916 (95% CI: 0.873–0.958), respectively. Both models exhibited good calibration, with Brier scores of 0.141 and 0.136. In the validation set, the AUCs were 0.871 (95% CI: 0.792–0.951) and 0.878 (95% CI: 0.794–0.962), with Brier scores of 0.156 and 0.165, respectively. The Delong test revealed no statistically significant differences in AUC between the models (p > 0.05 for both). SHAP value analysis for the minAV model suggested that minAVs had the highest absolute weight (AW) and negative contribution.ConclusionThe XGBoost predictive model based on minAVs demonstrates effective discrimination between adrenal lipid-poor adenomas and non-adenomas. The minAV variable is easily obtainable, and its diagnostic performance is comparable to that of the CTh model. This provides a basis for patient diagnosis and treatment plan selection.

Accuracy and time efficiency of a novel deep learning algorithm for Intracranial Hemorrhage detection in CT Scans

PurposeTo evaluate a deep learning-based pipeline using a Dense-UNet architecture for the assessment of acute intracranial hemorrhage (ICH) on non-contrast computed tomography (NCCT) head scans after traumatic brain injury (TBI).Materials and methodsThis retrospective study was conducted using a prototype algorithm that evaluated 502 NCCT head scans with ICH in context of TBI.Four board-certified radiologists evaluated in consensus the CT scans to establish the standard of reference for hemorrhage presence and type of ICH. Consequently, all CT scans were independently analyzed by the algorithm and a board-certified radiologist to assess the presence and type of ICH. Additionally, the time to diagnosis was measured for both methods.ResultsA total of 405/502 patients presented ICH classified in the following types: intraparenchymal (n = 172); intraventricular (n = 26); subarachnoid (n = 163); subdural (n = 178); and epidural (n = 15). The algorithm showed high diagnostic accuracy (91.24%) for the assessment of ICH with a sensitivity of 90.37% and specificity of 94.85%. To distinguish the different ICH types, the algorithm had a sensitivity of 93.47% and a specificity of 99.79%, with an accuracy of 98.54%. To detect midline shift, the algorithm had a sensitivity of 100%. In terms of processing time, the algorithm was significantly faster compared to the radiologist’s time to first diagnosis (15.37 ± 1.85 vs 277 ± 14 s, p < 0.001).ConclusionA novel deep learning algorithm can provide high diagnostic accuracy for the identification and classification of ICH from unenhanced CT scans, combined with short processing times. This has the potential to assist and improve radiologists’ ICH assessment in NCCT scans, especially in emergency scenarios, when time efficiency is needed.

A Fully Automated Pipeline Using Swin Transformers for Deep Learning-Based Blood Segmentation on Head Computed Tomography Scans After Aneurysmal Subarachnoid Hemorrhage

Accurate volumetric assessment of spontaneous aneurysmal subarachnoid hemorrhage (aSAH) is a labor-intensive task performed with current manual and semiautomatic methods that might be relevant for its clinical and prognostic implications. In the present research, we sought to develop and validate an artificial intelligence-driven, fully automated blood segmentation tool for subarachnoid hemorrhage (SAH) patients via noncontrast computed tomography (NCCT) scans employing a transformer-based Swin-UNETR architecture. We retrospectively analyzed NCCT scans from patients with confirmed aSAH utilizing the Swin-UNETR for segmentation. The performance of the proposed method was evaluated against manually segmented ground truth data using metrics such as Dice score, intersection over union, volumetric similarity index , symmetric average surface distance , sensitivity, and specificity. A validation cohort from an external institution was included to test the generalizability of the model. The model demonstrated high accuracy with robust performance metrics across the internal and external validation cohorts. Notably, it achieved high Dice coefficient (0.873±0.097), intersection over union (0.810±0.092), volumetric similarity index (0.840±0.131), sensitivity (0.821±0.217), and specificity (0.996±0.004) values and a low symmetric average surface distance (1.866±2.910), suggesting proficiency in segmenting blood in SAH patients. The model's efficiency was reflected in its processing speed, indicating potential for real-time applications. Our Swin UNETR-based model offers significant advances in the automated segmentation of blood in SAH patients on NCCT images. Despite the computational demands, the model operates effectively on standard hardware with a user-friendly interface, facilitating broader clinical adoption. Further validation across diverse datasets is warranted to confirm its clinical reliability.

Development of a non-invasive diagnostic model for high-risk esophageal varices based on radiomics of spleen CT.

To evaluate the diagnostic performance of radiomics models derived from multi-phase spleen CT for high-risk esophageal varices (HREV) in cirrhotic patients. We retrospectively selected cirrhotic patients with esophageal varices from two hospitals from September 2019 to September 2023. Patients underwent non-contrast and contrast-enhanced CT scans and were categorized into HREV and non-HREV groups based on endoscopic evaluations. Radiomics features were extracted from spleen CT images in non-contrast, arterial, and portal venous phases, with feature selection via lasso regression and Pearson's correlation. Ten machine learning models were developed to diagnose HREV, evaluated by area under the curve (AUC). The AUC values of the three groups of models were statistically compared by the Kruskal-Wallis H test and Bonferroni-corrected Mann-Whitney U test. A p-value less than 0.05 was considered statistically significant. Among 233 patients, 11, 6, and 11 features were selected from non-contrast, arterial, and portal venous phases, respectively. Significant differences in AUC values were observed across phases (p < 0.05), and the arterial phase models showed the highest AUC values. The best model in arterial phase was the logical regression model, whose AUC value was 0.85, sensitivity was 83.3%, specificity was 80% and F1 score was 0.81. Radiomics models based on spleen CT, especially the arterial phase models, demonstrate high diagnostic accuracy for HREV, offering the potential for early detection and intervention.

A Deep Learning-Based Framework for Predicting Intracerebral Hemorrhage Hematoma Expansion Using Head Non-contrast CT Scan

Rationale and ObjectivesHematoma expansion (HE) in intracerebral hemorrhage (ICH) is a critical factor affecting patient outcomes, yet effective clinical tools for predicting HE are currently lacking. We aim to develop a fully automated framework based on deep learning for predicting HE using only clinical non-contrast CT (NCCT) scans. Materials and MethodsA large retrospective dataset (n=2,484) was collected from 84 centers, while a prospective dataset (n=500) was obtained from 26 additional centers. Baseline NCCT scans and follow-up NCCT scans were conducted within 6 hours and 48 hours from symptom onset, respectively. HE was defined as a volume increase of more than 6 mL on the follow-up NCCT. The retrospective dataset was divided into a training set (n=1,876) and a validation set (n=608) by patient inclusion time. A two-stage framework was trained to predict HE, and its performance was evaluated on both the validation and prospective sets. Receiver operating characteristics area under the curve (AUC), sensitivity, and specificity were leveraged. ResultsOur two-stage framework achieved an AUC of 0.760 (95% CI 0.724-0.799) on the retrospective validation set and 0.806 (95% CI 0.750-0.859) on the prospective set, outperforming the commonly used BAT score, which had AUCs of 0.582 and 0.699, respectively. ConclusionOur framework can automatically and robustly identify ICH patients at high risk of HE using admission head NCCT scans, providing more accurate predictions than the BAT score.

Prediction of anemia with thoracic computed tomography findings.

This study explored the potential of non-contrast thoracic computed tomography (CT) to predict anemia by correlating CT parameters with hemoglobin (Hb) levels in patients who underwent non-contrast thoracic CT for various indications. This retrospective study included 150 patients who underwent non-contrast thoracic CT scans and complete blood counts within 24 hours at our center between January and June 2023. Exclusion criteria included acute bleeding, iron accumulation disorders, recent transfusions, pregnancy, and certain thoracic CT artifacts. Hounsfield Unit (HU) measurements were obtained from the ascending aorta, left ventricular cavity, and descending aorta, and compared with Hb and hematocrit (Htc) values. Anemia indicators such as the 'Aortic Ring Sign (ARS)' and the 'Hyperdense Septum Sign (HSS)' were also evaluated. Anemic patients (48%) exhibited significantly lower HU measurements at all three CT scan locations and higher instances of ARS and HSS compared to non-anemic patients. Notably, the presence of HSS and ARS was strongly associated with anemia. Thresholds for HU measurements corresponding to anemia were determined using receiver operating characteristic curve analysis, which also revealed strong positive correlations between HU measurements and Hb/Htc levels. The study concludes that non-contrast thoracic CT parameters, particularly HU measurements and the presence of ARS and HSS, are significantly associated with anemia. These CT indicators could serve as reliable, non-invasive markers for predicting anemia in patients, potentially aiding in the early diagnosis and management of the condition.

Radiodensities of Skeletal Muscle and Visceral Adipose Tissues Are Prognostic Factors in Clear-Cell Renal Cell Carcinoma.

Body composition may be related to survival in patients with clear-cell renal cell carcinoma (ccRCC), but studies have not simultaneously considered adipose and muscle tissue quantity and radiodensity. We analyzed data from 1,022 patients with ccRCC who underwent nephrectomy between 2000 and 2020 at Memorial Sloan Kettering Cancer Center. Skeletal muscle, visceral adipose tissue, and subcutaneous adipose tissue indexes (cm2/m2) and radiodensities [Hounsfield units (HU)] were assessed from noncontrast presurgical CT scans; clinical and demographic characteristics were available from the time of surgery. HRs and confidence intervals were estimated for overall (OS) and disease-free survival (DFS) through March 2023 in multivariable models that simultaneously accounted for all body composition measures. The median age of the patients was 58 years, 69% were male, and 90% were White. There were 169 OS events over 8,392 person-years and 253 DFS events over 7,753 person-years of follow-up. In adjusted analyses, poor OS was associated with lower skeletal muscle radiodensity [-10 HU, HR (95% confidence interval), 1.37 (1.05-1.77)] and greater visceral adipose tissue radiodensity [+10 HU, 1.66 (1.06-2.59)], with similar findings for DFS. Poor survival was also associated with greater visceral adipose tissue index [+40 cm2/m2, OS: 1.32 (0.97, 1.79); DFS: 1.33 (1.04, 1.71)]. Associations with skeletal muscle radiodensity were limited to patients with stage 1/2 disease. Radiodensities of skeletal muscle and visceral adipose tissues may be novel presurgical prognostic factors for patients with ccRCC. The findings underscore the importance of evaluating the full range of body composition features simultaneously in multivariable models.

Deep learning-based automatic ASPECTS calculation can improve diagnosis efficiency in patients with acute ischemic stroke: a multicenter study.

The Alberta Stroke Program Early CT Score (ASPECTS), a systematic method for assessing ischemic changes in acute ischemic stroke using non-contrast computed tomography (NCCT), is often interpreted relying on expert experience and can vary between readers. This study aimed to develop a clinically applicable automatic ASPECTS system employing deep learning (DL). This study enrolled 1987 NCCT scans that were retrospectively collected from four centers between January 2017 and October 2021. A DL-based system for automated ASPECTS assessment was trained on a development cohort (N = 1767) and validated on an independent test cohort (N = 220). The consensus of experienced physicians was regarded as a reference standard. The validity and reliability of the proposed system were assessed against physicians' readings. A real-world prospective application study with 13,399 patients was used for system validation in clinical contexts. The DL-based system achieved an area under the receiver operating characteristic curve (AUC) of 84.97% and an intraclass correlation coefficient (ICC) of 0.84 for overall-level analysis on the test cohort. The system's diagnostic sensitivity was 94.61% for patients with dichotomized ASPECTS at a threshold of ≥ 6, with substantial agreement (ICC = 0.65) with expert ratings. Combining the system with physicians improved AUC from 67.43 to 89.76%, reducing diagnosis time from 130.6 ± 66.3 s to 33.3 ± 8.3 s (p < 0.001). During the application in clinical contexts, 94.0% (12,591) of scans successfully processed by the system were utilized by clinicians, and 96% of physicians acknowledged significant improvement in work efficiency. The proposed DL-based system could accurately and rapidly determine ASPECTS, which might facilitate clinical workflow for early intervention. The deep learning-based automated ASPECTS evaluation system can accurately and rapidly determine ASPECTS for early intervention in clinical workflows, reducing processing time for physicians by 74.8%, but still requires validation by physicians when in clinical applications. The deep learning-based system for ASPECTS quantification has been shown to be non-inferior to expert-rated ASPECTS. This system improved the consistency of ASPECTS evaluation and reduced processing time to 33.3 seconds per scan. 94.0% of scans successfully processed by the system were utilized by clinicians during the prospective clinical application.

Sociodemographic biases in a commercial AI model for intracranial hemorrhage detection.

To evaluate whether a commercial AI tool for intracranial hemorrhage (ICH) detection on head CT exhibited sociodemographic biases. Our retrospective study reviewed 9736 consecutive, adult non-contrast head CT scans performed between November 2021 and February 2022 in a single healthcare system. Each CT scan was evaluated by a commercial ICH AI tool and a board-certified neuroradiologist; ground truth was defined as final radiologist determination of ICH presence/absence. After evaluating the AI tool's aggregate diagnostic performance, sub-analyses based on sociodemographic groups (age, sex, race, ethnicity, insurance status, and Area of Deprivation Index [ADI] scores) assessed for biases. χ2 test or Fisher's exact tests evaluated for statistical significance with p ≤ 0.05. Our patient population was 50% female (mean age 60 ± 19 years). The AI tool had an aggregate accuracy of 93% [9060/9736], sensitivity of 85% [1140/1338], specificity of 94% [7920/ 8398], positive predictive value (PPV) of 71% [1140/1618] and negative predictive value (NPV) of 98% [7920/8118]. Sociodemographic biases were identified, including lower PPV for patients who were females (67.3% [62,441/656] vs. 72.7% [699/962], p = 0.02), Black (66.7% [454/681] vs. 73.2% [686/937], p = 0.005), non-Hispanic/non-Latino (69.7% [1038/1490] vs. 95.4% [417/437]), p = 0.009), and who had Medicaid/Medicare (69.9% [754/1078]) or Private (66.5% [228/343]) primary insurance (p = 0.003). Lower sensitivity was seen for patients in the third quartile of national (78.8% [241/306], p = 0.001) and state ADI scores (79.0% [22/287], p = 0.001). In our healthcare system, a commercial AI tool had lower performance for ICH detection than previously reported and demonstrated several sociodemographic biases.

Author Response: Deep learning-assisted detection and segmentation of intracranial hemorrhage in noncontrast computed tomography scans of acute stroke patients: a systematic review and meta-analysis.

Author Response: Deep learning-assisted detection and segmentation of intracranial hemorrhage in noncontrast computed tomography scans of acute stroke patients: a systematic review and meta-analysis.

Differential Impact of Subcutaneous and Visceral Fat on Bone Changes after Gastrectomy.

Osteoporosis and fragility fractures are crucial musculoskeletal complications in long-term survivors of gastric cancer. However, the relationship between changes in body composition after gastrectomy and bone loss has not been investigated. Therefore, this study aimed to explore whether computed tomography (CT)-derived body composition parameters are associated with bone loss after gastrectomy in patients with gastric cancer. We retrospectively reviewed medical records and abdomen CT scans of patients who underwent gastrectomy at Yonsei University Severance Hospital between 2009 and 2018. Patients with non-metastatic gastric adenocarcinoma and preoperative and postoperative non-contrast CT scans were analyzed. Section area of skeletal muscle (SMA), visceral fat (VFA), and subcutaneous fat (SFA) were assessed using semi-automatic segmentation software. Changes in trabecular bone attenuation of L1 mid-vertebra level (L1 Hounsfield units [HU]) were measured. Fifty-seven patients (mean age, 65.5±10.6; 70.2% males) were analyzed, and the median duration was 31 months. Fortyseven patients (82.5%) lost weight after gastrectomy. Baseline SMA and VFA did not differ between the bone loss and preserved groups; however, baseline SFA was significantly higher in the bone preserved group than in the bone loss group (P=0.020). In a multivariable linear regression model adjusted for confounding factors, one standard deviation higher VFA at baseline was associated with greater annualized L1 HU loss (%) (P=0.034). However, higher preoperative SFA was associated with protection against bone loss after gastrectomy (P=0.025). Higher preoperative SFA exhibited a protective effect against bone loss after gastrectomy in patients with non-metastatic gastric cancer, whereas VFA exhibited a negative effect.

Exploring the relationship between epicardial fat and coronary plaque burden and characteristics: insights from cardiac ct imaging.

Epicardial adipose tissue (EAT) may enhance the risk of coronary artery disease (CAD). We investigated the relationship between EAT density (a maker of local inflammation) and coronary plaque characteristics in stable CAD patients. This study included 123 individuals who underwent coronary artery calcium scan and coronary CT angiography to evaluate CAD. Plaque characteristics were analyzed by semi-automated software (QAngio, Leiden, Netherlands). Non-contrast CT scans were used to measure EAT density (HU) and volume (cc) (Philips, Cleveland, OH). Multivariate regression models were used to evaluate the association of EAT density and volume with different plaque types. The mean (SD) age was 59.4±10.1 years, 53% were male, the mean (SD) EAT density was -77.2±4.6 HU and the volume was 118.5±41.2 cc. After adjustment for cardiovascular risk factors, EAT density was associated with fibrous fatty (FF) plaque (p<0.03). A 1 unit increase in HU was associated with a 7% higher FF plaque, and lower EAT density is independently associated to FF plaque. The association between EAT density and fibrous (p=0.08), and total noncalcified (p=0.09) plaque trended toward but did not reach significance. There was no association between EAT volume and any plaque type. These results suggest that inflammatory EAT may promote coronary atherosclerosis. Therefore, non-contrast cardiac CT evaluation of EAT quality can help better assess cardiovascular risk.