Artificial Intelligence for Assessment of Endotracheal Tube Position on Chest Radiographs: Validation in Patients From Two Institutions.

BACKGROUND. Timely and accurate interpretation of chest radiographs obtained to evaluate endotracheal tube (ETT) position is important for facilitating prompt adjustment if needed. OBJECTIVE. The purpose of our study was to evaluate the performance of a deep learning (DL)-based artificial intelligence (AI) system for detecting ETT presence and position on chest radiographs in three patient samples from two different institutions. METHODS. This retrospective study included 539 chest radiographs obtained immediately after ETT insertion from January 1 to March 31, 2020, in 505 patients (293 men, 212 women; mean age, 63 years) from institution A (sample A); 637 chest radiographs obtained from January 1 to January 3, 2020, in 302 patients (157 men, 145 women; mean age, 66 years) in the ICU (with or without an ETT) from institution A (sample B); and 546 chest radiographs obtained from January 1 to January 20, 2020, in 83 patients (54 men, 29 women; mean age, 70 years) in the ICU (with or without an ETT) from institution B (sample C). A commercial DL-based AI system was used to identify ETT presence and measure ETT tip-to-carina distance (TCD). The reference standard for proper ETT position was TCD between greater than 3 cm and less than 7 cm, determined by human readers. Critical ETT position was separately defined as ETT tip below the carina or TCD of 1 cm or less. ROC analysis was performed. RESULTS. AI had sensitivity and specificity for identification of ETT presence of 100.0% and 98.7% (sample B) and 99.2% and 94.5% (sample C). AI had sensitivity and specificity for identification of improper ETT position of 72.5% and 92.0% (sample A), 78.9% and 100.0% (sample B), and 83.7% and 99.1% (sample C). At a threshold y-axis TCD of 2 cm or less, AI had sensitivity and specificity for critical ETT position of 100.0% and 96.7% (sample A), 100.0% and 100.0% (sample B), and 100.0% and 99.2% (sample C). CONCLUSION. AI identified improperly positioned ETTs on chest radiographs obtained after ETT insertion as well as on chest radiographs obtained of patients in the ICU at two institutions. CLINICAL IMPACT. Automated AI identification of improper ETT position on chest radiographs may allow earlier repositioning and thereby reduce complications.

A Deep Learning Model Using Chest Radiographs for Prediction of 30-Day Mortality in Patients With Community-Acquired Pneumonia: Development and External Validation.

BACKGROUND. Chest radiography is an essential tool for diagnosing community-acquired pneumonia (CAP), but it has an uncertain prognostic role in the care of patients with CAP. OBJECTIVE. The purpose of this study was to develop a deep learning (DL) model to predict 30-day mortality from diagnosis among patients with CAP by use of chest radiographs to validate the performance model in patients from different time periods and institutions. METHODS. In this retrospective study, a DL model was developed from data on 7105 patients from one institution from March 2013 to December 2019 (3:1:1 allocation to training, validation, and internal test sets) to predict the risk of all-cause mortality within 30 days after CAP diagnosis by use of patients' initial chest radiographs. The DL model was evaluated in a cohort of patients diagnosed with CAP during emergency department visits at the same institution from January 2020 to March 2020 (temporal test cohort [n = 947]) and in two additional cohorts from different institutions (external test cohort A [n = 467], January 2020 to December 2020; external test cohort B [n = 381], March 2019 to October 2021). AUCs were compared between the DL model and an established risk prediction tool based on the presence of confusion, blood urea nitrogen level, respiratory rate, blood pressure, and age 65 years or older (CURB-65 score). The combination of CURB-65 score and DL model was evaluated with a logistic regression model. RESULTS. The AUC for predicting 30-day mortality was significantly larger (p < .001) for the DL model than for CURB-65 score in the temporal test set (0.77 vs 0.67). The larger AUC for the DL model than for CURB-65 score was not significant (p > .05) in external test cohort A (0.80 vs 0.73) or external test cohort B (0.80 vs 0.72). In the three cohorts, the DL model, in comparison with CURB-65 score, had higher (p < .001) specificity (range, 61-69% vs 44-58%) at the sensitivity of CURB-65 score. The combination of DL model and CURB-65 score, in comparison with CURB-65 score, yielded a significant increase in AUC in the temporal test cohort (0.77, p < .001) and external test cohort B (0.80, p = .04) and a nonsignificant increase in AUC in external test cohort A (0.80, p = .16). CONCLUSION. A DL-based model consisting of initial chest radiographs was predictive of 30-day mortality among patients with CAP with improved performance over CURB-65 score. CLINICAL IMPACT. The DL-based model may guide clinical decision-making in the care of patients with CAP.

Chest Radiography of Tuberculosis: Determination of Activity Using Deep Learning Algorithm.

BACKGROUND: Inactive or old, healed tuberculosis (TB) on chest radiograph (CR) is often found in high TB incidence countries, and to avoid unnecessary evaluation and medication, differentiation from active TB is important. This study develops a deep learning (DL) model to estimate activity in a single chest radiographic analysis. METHODS: A total of 3,824 active TB CRs from 511 individuals and 2,277 inactive TB CRs from 558 individuals were retrospectively collected. A pretrained convolutional neural network was fine-tuned to classify active and inactive TB. The model was pretrained with 8,964 pneumonia and 8,525 normal cases from the National Institute of Health (NIH) dataset. During the pretraining phase, the DL model learns the following tasks: pneumonia vs. normal, pneumonia vs. active TB, and active TB vs. normal. The performance of the DL model was validated using three external datasets. Receiver operating characteristic analyses were performed to evaluate the diagnostic performance to determine active TB by DL model and radiologists. Sensitivities and specificities for determining active TB were evaluated for both the DL model and radiologists. RESULTS: The performance of the DL model showed area under the curve (AUC) values of 0.980 in internal validation, and 0.815 and 0.887 in external validation. The AUC values for the DL model, thoracic radiologist, and general radiologist, evaluated using one of the external validation datasets, were 0.815, 0.871, and 0.811, respectively. CONCLUSION: This DL-based algorithm showed potential as an effective diagnostic tool to identify TB activity, and could be useful for the follow-up of patients with inactive TB in high TB burden countries.

CT analysis of thoracolumbar body composition for estimating whole-body composition.

BACKGROUND: To evaluate the correlation between single- and multi-slice cross-sectional thoracolumbar and whole-body compositions. METHODS: We retrospectively included patients who underwent whole-body PET-CT scans from January 2016 to December 2019 at multiple institutions. A priori-developed, deep learning-based commercially available 3D U-Net segmentation provided whole-body 3D reference volumes and 2D areas of muscle, visceral fat, and subcutaneous fat at the upper, middle, and lower endplate of the individual T1-L5 vertebrae. In the derivation set, we analyzed the Pearson correlation coefficients of single-slice and multi-slice averaged 2D areas (waist and T12-L1) with the reference values. We then built prediction models using the top three correlated levels and tested the models in the validation set. RESULTS: The derivation and validation datasets included 203 (mean age 58.2 years; 101 men) and 239 patients (mean age 57.8 years; 80 men). The coefficients were distributed bimodally, with the first peak at T4 (coefficient, 0.78) and the second peak at L2-3 (coefficient 0.90). The top three correlations in the abdominal scan range were found for multi-slice waist averaging (0.92) and single-slice L3 and L2 (0.90, each), while those in the chest scan range were multi-slice T12-L1 averaging (0.89), single-slice L1 (0.89), and T12 (0.86). The model performance at the top three levels for estimating whole-body composition was similar in the derivation and validation datasets. CONCLUSIONS: Single-slice L2-3 (abdominal CT range) and L1 (chest CT range) analysis best correlated with whole-body composition around 0.90 (coefficient). Multi-slice waist averaging provided a slightly higher correlation of 0.92.

Artificial Intelligence-Based Identification of Normal Chest Radiographs: A Simulation Study in a Multicenter Health Screening Cohort.

OBJECTIVE: This study aimed to investigate the feasibility of using artificial intelligence (AI) to identify normal chest radiography (CXR) from the worklist of radiologists in a health-screening environment. MATERIALS AND METHODS: This retrospective simulation study was conducted using the CXRs of 5887 adults (mean age ± standard deviation, 55.4 ± 11.8 years; male, 4329) from three health screening centers in South Korea using a commercial AI (Lunit INSIGHT CXR3, version 3.5.8.8). Three board-certified thoracic radiologists reviewed CXR images for referable thoracic abnormalities and grouped the images into those with visible referable abnormalities (identified as abnormal by at least one reader) and those with clearly visible referable abnormalities (identified as abnormal by at least two readers). With AI-based simulated exclusion of normal CXR images, the percentages of normal images sorted and abnormal images erroneously removed were analyzed. Additionally, in a random subsample of 480 patients, the ability to identify visible referable abnormalities was compared among AI-unassisted reading (i.e., all images read by human readers without AI), AI-assisted reading (i.e., all images read by human readers with AI assistance as concurrent readers), and reading with AI triage (i.e., human reading of only those rendered abnormal by AI). RESULTS: Of 5887 CXR images, 405 (6.9%) and 227 (3.9%) contained visible and clearly visible abnormalities, respectively. With AI-based triage, 42.9% (2354/5482) of normal CXR images were removed at the cost of erroneous removal of 3.5% (14/405) and 1.8% (4/227) of CXR images with visible and clearly visible abnormalities, respectively. In the diagnostic performance study, AI triage removed 41.6% (188/452) of normal images from the worklist without missing visible abnormalities and increased the specificity for some readers without decreasing sensitivity. CONCLUSION: This study suggests the feasibility of sorting and removing normal CXRs using AI with a tailored cut-off to increase efficiency and reduce the workload of radiologists.

Localization-adjusted diagnostic performance and assistance effect of a computer-aided detection system for pneumothorax and consolidation.

While many deep-learning-based computer-aided detection systems (CAD) have been developed and commercialized for abnormality detection in chest radiographs (CXR), their ability to localize a target abnormality is rarely reported. Localization accuracy is important in terms of model interpretability, which is crucial in clinical settings. Moreover, diagnostic performances are likely to vary depending on thresholds which define an accurate localization. In a multi-center, stand-alone clinical trial using temporal and external validation datasets of 1,050 CXRs, we evaluated localization accuracy, localization-adjusted discrimination, and calibration of a commercially available deep-learning-based CAD for detecting consolidation and pneumothorax. The CAD achieved image-level AUROC (95% CI) of 0.960 (0.945, 0.975), sensitivity of 0.933 (0.899, 0.959), specificity of 0.948 (0.930, 0.963), dice of 0.691 (0.664, 0.718), moderate calibration for consolidation, and image-level AUROC of 0.978 (0.965, 0.991), sensitivity of 0.956 (0.923, 0.978), specificity of 0.996 (0.989, 0.999), dice of 0.798 (0.770, 0.826), moderate calibration for pneumothorax. Diagnostic performances varied substantially when localization accuracy was accounted for but remained high at the minimum threshold of clinical relevance. In a separate trial for diagnostic impact using 461 CXRs, the causal effect of the CAD assistance on clinicians' diagnostic performances was estimated. After adjusting for age, sex, dataset, and abnormality type, the CAD improved clinicians' diagnostic performances on average (OR [95% CI] = 1.73 [1.30, 2.32]; p < 0.001), although the effects varied substantially by clinical backgrounds. The CAD was found to have high stand-alone diagnostic performances and may beneficially impact clinicians' diagnostic performances when used in clinical settings.

Deep neural network for automatic volumetric segmentation of whole-body CT images for body composition assessment.

BACKGROUND & AIMS: Body composition analysis on CT images is a valuable tool for sarcopenia assessment. We aimed to develop and validate a deep neural network applicable to whole-body CT images of PET-CT scan for the automatic volumetric segmentation of body composition. METHODS: For model development, one hundred whole-body or torso 18F-fluorodeoxyglucose PET-CT scans of 100 patients were retrospectively included. Two radiologists semi-automatically labeled the following seven body components in every CT image slice, providing a total of 46,967 image slices from the 100 scans for training the 3D U-Net (training, 39,268 slices; tuning, 3116 slices; internal validation, 4583 slices): skin, bone, muscle, abdominal visceral fat, subcutaneous fat, internal organs with vessels, and central nervous system. The segmentation accuracy was assessed using reference masks from three external datasets: two Korean centers (4668 and 4796 image slices from 20 CT scans, each) and a French public dataset (3763 image slices from 24 CT scans). The 3D U-Net-driven values were clinically validated using bioelectrical impedance analysis (BIA) and by assessing the model's diagnostic performance for sarcopenia in a community-based elderly cohort (n = 522). RESULTS: The 3D U-Net achieved accurate body composition segmentation with an average dice similarity coefficient of 96.5%-98.9% for all masks and 92.3%-99.3% for muscle, abdominal visceral fat, and subcutaneous fat in the validation datasets. The 3D U-Net-derived torso volume of skeletal muscle and fat tissue and the average area of those tissues in the waist were correlated with BIA-derived appendicular lean mass (correlation coefficients: 0.71 and 0.72, each) and fat mass (correlation coefficients: 0.95 and 0.93, each). The 3D U-Net-derived average areas of skeletal muscle and fat tissue in the waist were independently associated with sarcopenia (P < .001, each) with adjustment for age and sex, providing an area under the curve of 0.858 (95% CI, 0.815 to 0.901). CONCLUSIONS: This deep neural network model enabled the automatic volumetric segmentation of body composition on whole-body CT images, potentially expanding adjunctive sarcopenia assessment on PET-CT scan and volumetric assessment of metabolism in whole-body muscle and fat tissues.

Evaluation of a deep learning-based computer-aided detection algorithm on chest radiographs: Case-control study.

ABSTRACT: Along with recent developments in deep learning techniques, computer-aided diagnosis (CAD) has been growing rapidly in the medical imaging field. In this work, we evaluate the deep learning-based CAD algorithm (DCAD) for detecting and localizing 3 major thoracic abnormalities visible on chest radiographs (CR) and to compare the performance of physicians with and without the assistance of the algorithm. A subset of 244 subjects (60% abnormal CRs) was evaluated. Abnormal findings included mass/nodules (55%), consolidation (21%), and pneumothorax (24%). Observer performance tests were conducted to assess whether the performance of physicians could be enhanced with the algorithm. The area under the receiver operating characteristic (ROC) curve (AUC) and the area under the jackknife alternative free-response ROC (JAFROC) were measured to evaluate the performance of the algorithm and physicians in image classification and lesion detection, respectively. The AUCs for nodule/mass, consolidation, and pneumothorax were 0.9883, 1.000, and 0.9997, respectively. For the image classification, the overall AUC of the pooled physicians was 0.8679 without DCAD and 0.9112 with DCAD. Regarding lesion detection, the pooled observers exhibited a weighted JAFROC figure of merit (FOM) of 0.8426 without DCAD and 0.9112 with DCAD. DCAD for CRs could enhance physicians' performance in the detection of 3 major thoracic abnormalities.

Correction: Performance of a deep-learning algorithm for referable thoracic abnormalities on chest radiographs: A multicenter study of a health screening cohort.

[This corrects the article DOI: 10.1371/journal.pone.0246472.].

Performance of a deep-learning algorithm for referable thoracic abnormalities on chest radiographs: A multicenter study of a health screening cohort.

PURPOSE: This study evaluated the performance of a commercially available deep-learning algorithm (DLA) (Insight CXR, Lunit, Seoul, South Korea) for referable thoracic abnormalities on chest X-ray (CXR) using a consecutively collected multicenter health screening cohort. METHODS AND MATERIALS: A consecutive health screening cohort of participants who underwent both CXR and chest computed tomography (CT) within 1 month was retrospectively collected from three institutions' health care clinics (n = 5,887). Referable thoracic abnormalities were defined as any radiologic findings requiring further diagnostic evaluation or management, including DLA-target lesions of nodule/mass, consolidation, or pneumothorax. We evaluated the diagnostic performance of the DLA for referable thoracic abnormalities using the area under the receiver operating characteristic (ROC) curve (AUC), sensitivity, and specificity using ground truth based on chest CT (CT-GT). In addition, for CT-GT-positive cases, three independent radiologist readings were performed on CXR and clear visible (when more than two radiologists called) and visible (at least one radiologist called) abnormalities were defined as CXR-GTs (clear visible CXR-GT and visible CXR-GT, respectively) to evaluate the performance of the DLA. RESULTS: Among 5,887 subjects (4,329 males; mean age 54±11 years), referable thoracic abnormalities were found in 618 (10.5%) based on CT-GT. DLA-target lesions were observed in 223 (4.0%), nodule/mass in 202 (3.4%), consolidation in 31 (0.5%), pneumothorax in one 1 (<0.1%), and DLA-non-target lesions in 409 (6.9%). For referable thoracic abnormalities based on CT-GT, the DLA showed an AUC of 0.771 (95% confidence interval [CI], 0.751-0.791), a sensitivity of 69.6%, and a specificity of 74.0%. Based on CXR-GT, the prevalence of referable thoracic abnormalities decreased, with visible and clear visible abnormalities found in 405 (6.9%) and 227 (3.9%) cases, respectively. The performance of the DLA increased significantly when using CXR-GTs, with an AUC of 0.839 (95% CI, 0.829-0.848), a sensitivity of 82.7%, and s specificity of 73.2% based on visible CXR-GT and an AUC of 0.872 (95% CI, 0.863-0.880, P <0.001 for the AUC comparison of GT-CT vs. clear visible CXR-GT), a sensitivity of 83.3%, and a specificity of 78.8% based on clear visible CXR-GT. CONCLUSION: The DLA provided fair-to-good stand-alone performance for the detection of referable thoracic abnormalities in a multicenter consecutive health screening cohort. The DLA showed varied performance according to the different methods of ground truth.

Chest Radiographs and CT Findings during Healthcare Workers' Tuberculosis Screening Using Interferon-Gamma Release Assay: Retrospective Observational Study.

Purpose: To investigate the incidence of tuberculosis (TB) in healthcare workers (HCWs) with positive interferon-gamma release assay (IGRA) results based on chest X-ray (CXR) and CT findings and determine the role of imaging in the diagnosis of TB. Materials and Methods: Among 1976 hospital personnel screened for TB using IGRA, IGRA-positive subjects were retrospectively investigated. Clustered nodular and/or linear streaky opacities in the upper lung zone were considered positive on CXR. The CT findings were classified as active, indeterminate, inactive, or normal. The active or indeterminate class was considered CT-positive. Results: IGRA was positive in 255 subjects (12.9%). CXR and CT were performed in 249 (99.2%) and 113 subjects (45.0%), respectively. CXR- and CT-positive findings were found in 7 of 249 (2.8%) and 9 of 113 (8.0%) patients, respectively. Among the nine CT-positive subjects, active and indeterminate TB findings were found in 6 (5.3%) and 3 (2.7%) patients, respectively. Microbiological tests, including acid-fast bacilli staining, culture, and polymerase chain reaction for TB, were negative in all nine CT-positive subjects. Empirical anti-TB medications were administered to 9 CT-positive subjects, and 3 of these nine subjects were CXR-negative for pulmonary TB. Conclusion: CT helped diagnose asymptomatic TB in IGRA-positive HCWs.

Incidence Lung Cancer after a Negative CT Screening in the National Lung Screening Trial: Deep Learning-Based Detection of Missed Lung Cancers.

We aimed to analyse the CT examinations of the previous screening round (CTprev) in NLST participants with incidence lung cancer and evaluate the value of DL-CAD in detection of missed lung cancers. Thoracic radiologists reviewed CTprev in participants with incidence lung cancer, and a DL-CAD analysed CTprev according to NLST criteria and the lung CT screening reporting & data system (Lung-RADS) classification. We calculated patient-wise and lesion-wise sensitivities of the DL-CAD in detection of missed lung cancers. As per the NLST criteria, 88% (100/113) of CTprev were positive and 74 of them had missed lung cancers. The DL-CAD reported 98% (98/100) of the positive screens as positive and detected 95% (70/74) of the missed lung cancers. As per the Lung-RADS classification, 82% (93/113) of CTprev were positive and 60 of them had missed lung cancers. The DL-CAD reported 97% (90/93) of the positive screens as positive and detected 98% (59/60) of the missed lung cancers. The DL-CAD made false positive calls in 10.3% (27/263) of controls, with 0.16 false positive nodules per scan (41/263). In conclusion, the majority of CTprev in participants with incidence lung cancers had missed lung cancers, and the DL-CAD detected them with high sensitivity and a limited false positive rate.

Cone-Beam Computed Tomography-Hepatic Arteriography as a Diagnostic Tool for Small Hypervascular Hepatocellular Carcinomas: Method and Clinical Implications.

OBJECTIVE: This study proposes a novel reference standard for hypervascular hepatocellular carcinomas (HCCs), established by cone-beam computed tomography-hepatic arteriography (CBCT-HA) and two-year imaging follow-up, and discusses its clinical implication on tumor staging and understanding the intrahepatic distant recurrence (IDR) in relation to dynamic computed tomography (CT). MATERIALS AND METHODS: In this retrospective study, 99 patients were enrolled, who underwent CBCT-HA during initial chemoembolization for HCC suspected on CT. All patients underwent chemoembolization and regular clinical and imaging follow-up for two years. If IDR appeared on follow-up imaging, initial CBCT-HA images were reviewed to determine if a hypervascular focus pre-existed at the site of recurrence. Pre-existing hypervascular foci on CBCT-HA were regarded as HCCs in initial presentation. Initial HCCs were classified into three groups according to their mode of detection (Group I, detected on CT and CBCT-HA; Group II, additionally detected on CBCT-HA; Group III, confirmed by interval growth). We assessed the influence of CBCT-HA and two-year follow-up on initial tumor stage and calculated the proportion of IDR that pre-existed in initial CBCT-HA. RESULTS: A total of 405 nodules were confirmed as HCCs, and 297 nodules initially pre-existed. Of the initial 297 HCCs, 149 (50.2%) lesions were in Group I, 74 (24.9%) lesions were in Group II, and the remaining 74 (24.9%) lesions were in Group III. After applying CBCT-HA findings, 11 patients upstaged in T stage, and 4 patients had a change in Milan criteria. Our reference standard for HCC indicated that 120 of 148 (81.1%) one-year IDR and 148 of 256 (57.8%) two-year IDR existed on initial CBCT-HA. CONCLUSION: The proposed method enabled the confirmation of many sub-centimeter-sized, faintly vascularized HCC nodules that pre-existed initially but clinically manifested as IDR. Our reference standard for HCC helped in understanding the nature of IDR and the early development of HCC as well as the clinical impact of tumor staging and treatment decision.

Growth of thymic epithelial tumors and thymic cysts: Differential radiological points.

BACKGROUND: The growth rate of thymic epithelial tumors (TETs) and thymic cysts was investigated to determine whether they can be differentiated and clinico-radiological predictors of interval growth was identified. METHODS: This retrospective study included 122 patients with pathologically proven thymic cysts (n = 56) or TETs (n = 66) who underwent two serial chest computed tomography scans at least eight weeks apart. Average diameters and attenuation were measured, volume-doubling times (VDTs) were calculated, and clinical characteristics were recorded. VDTs were compared using the log-rank test. Predictors of growth were analyzed using the log-rank test and Cox regression analysis. RESULTS: The frequency of growth did not differ significantly between TETs and thymic cysts (P = 0.279). The VDT of thymic cysts (median 324 days) was not significantly different from that of the TETs (median 475 days; P = 0.808). Water attenuation (≤ 20 Hounsfield units) predicted growth in thymic cysts (P = 0.016; hazard ratio 13.2, 95% confidence interval 1.6-107.3), while lesion size (> 17.2 mm) predicted growth in TETs (P = 0.008 for size, P = 0.029 for size*time). For the growing lesions, the positive and negative predictive values of water attenuation for thymic cysts were 93% and 80%, respectively. CONCLUSION: The frequencies of interval growth and VDTs were indistinguishable between TETs and thymic cysts. Water attenuation and lesion size predicted growth in thymic cysts and TETs, respectively. Among the growing lesions, water attenuation was a differential feature of thymic cysts.

Diagnostic accuracy of contrast-enhanced dynamic CT for small hypervascular hepatocellular carcinoma and assessment of dynamic enhancement patterns: Results of two-year follow-up using cone-beam CT hepatic arteriography.

OBJECTIVE: To evaluate the accuracy of CT for small, hypervascular hepatocellular carcinomas (HCCs) and assess the enhancement patterns on CT. MATERIALS AND METHODS: Ninety-nine patients who underwent cone-beam CT hepatic arteriography (CBCT-HA) during initial chemoembolization for HCC suspected on CT were enrolled in this study. A total of 297 hypervascular HCCs (142 ≥ 1 cm, 155 < 1 cm) were confirmed as HCCs based on two-year follow-up CT and CBCT-HA images. During the two-year follow-up, pre-existing hypervascular foci on CBCT-HA were regarded as HCCs at the initial presentation. Two radiologists categorized HCCs according to the following enhancement patterns on CT: type I, arterial enhancement and washout; type II, arterial enhancement without washout; and type III, no arterial enhancement. Two blinded reviewers rated the possibility of HCC. RESULTS: For the 297 HCCs, the enhancement patterns according to size were as follows: type I ≥1 cm in 114 HCCs; type I <1 cm in 40 HCCs; type II ≥1 cm in 16 HCCs; type II <1 cm in 37 HCCs; type III ≥1 cm in 12 HCCs; and type III <1 cm in 10 HCCs. The remaining 68 HCCs (22.9%) were not detected on CT. The detection rates of HCCs ≥ 1 cm were 83.1%, 76.8%, and 83.1% in the formal report for reviewer 1 and reviewer 2. In comparison, the detection rates of HCCs < 1 cm were 20.6%, 17.4%, and 17.4% in the formal report for reviewer 1 and reviewer 2. CONCLUSION: Many subcentimeter sized hypervascular HCCs were frequently missed or not evident on CT at the initial diagnostic workup. CT has limitations for diagnosing HCCs that are <1 cm in size or have atypical enhancement patterns.

Acute invasive fungal rhinosinusitis: MR imaging features and their impact on prognosis.

PURPOSE: Acute invasive fungal rhinosinusitis (AIFRS) is a life-threatening disease that is difficult to diagnose. Its overall imaging features have not been evaluated and the prognostic impact is unclear. The purpose of our study was to present MR imaging features and their impact on prognosis of AIFRS. METHODS: MR images and clinical records of 23 patients with AIFRS were retrospectively evaluated to identify the imaging features and to determine the factors affecting patients' survival. A multivariable Cox proportional hazard model was used to estimate the hazard ratio of the prognostic factors, and Kaplan-Meier survival curves were compared by using a log-rank test. RESULTS: All cases showed extra-sinonasal involvement and the orbit was the most common (65.2%, 15/23) location. The lesion enhancement pattern was classified into lack of contrast enhancement (LoCE) (47.8%, 11/23) and homogeneous (34.8%, 8/23) and heterogeneous (17.4%, 4/23) enhancement. Although LoCE showed variable signal intensity (SI), homogeneously or heterogeneously enhancing lesions showed exclusively low SI (100%, 12/12) on T2WI. Among various clinical and imaging factors, LoCE was correlated with coagulation necrosis, probably provoked by numerous fungal hyphae, and was found to be a sole independent prognostic factor for disease-specific mortality (hazard ratio = 16.819; 95% CI, 1.646-171.841, p = 0.017). In addition, patients with LoCE showed worse survival than patients without LoCE (p = 0.008). CONCLUSION: AIFRS showed frequent extra-sinonasal involvement and variable MR enhancement patterns. An enhancement pattern of LoCE was seen in about half of the cases and was a unique prognostic factor among the various clinico-radiologic factors.

Therapeutic response assessment using 3D ultrasound for hepatic metastasis from colorectal cancer: Application of a personalized, 3D-printed tumor model using CT images.

BACKGROUND & AIMS: To evaluate accuracy and reliability of three-dimensional ultrasound (3D US) for response evaluation of hepatic metastasis from colorectal cancer (CRC) using a personalized 3D-printed tumor model. METHODS: Twenty patients with liver metastasis from CRC who underwent baseline and after chemotherapy CT, were retrospectively included. Personalized 3D-printed tumor models using CT were fabricated. Two radiologists measured volume of each 3D printing model using 3D US. With CT as a reference, we compared difference between CT and US tumor volume. The response evaluation was based on Response Evaluation Criteria in Solid Tumors (RECIST) criteria. RESULTS: 3D US tumor volume showed no significant difference from CT volume (7.18 ± 5.44 mL, 8.31 ± 6.32 mL vs 7.42 ± 5.76 mL in CT, p>0.05). 3D US provided a high correlation coefficient with CT (r = 0.953, r = 0.97) as well as a high inter-observer intraclass correlation (0.978; 0.958-0.988). Regarding response, 3D US was in agreement with CT in 17 and 18 out of 20 patients for observer 1 and 2 with excellent agreement (κ = 0.961). CONCLUSIONS: 3D US tumor volume using a personalized 3D-printed model is an accurate and reliable method for the response evaluation in comparison with CT tumor volume.

Association between Image Characteristics on Chest CT and Severe Pleural Adhesion during Lung Cancer Surgery.

The aim of this study was to investigate the association between image characteristics on preoperative chest CT and severe pleural adhesion during surgery in lung cancer patients. We included consecutive 124 patients who underwent lung cancer surgeries. Preoperative chest CT was retrospectively reviewed to assess pleural thickening or calcification, pulmonary calcified nodules, active pulmonary inflammation, extent of emphysema, interstitial pneumonitis, and bronchiectasis in the operated thorax. The extent of pleural thickening or calcification was visually estimated and categorized into two groups: localized and diffuse. We measured total size of pulmonary calcified nodules. The extent of emphysema, interstitial pneumonitis, and bronchiectasis was also evaluated with a visual scoring system. The occurrence of severe pleural adhesion during lung cancer surgery was retrospectively investigated from the electrical medical records. We performed logistic regression analysis to determine the association of image characteristic on chest CT with severe pleural adhesion. Localized pleural thickening was found in 8 patients (6.5%), localized pleural calcification in 8 (6.5%), pulmonary calcified nodules in 28 (22.6%), and active pulmonary inflammation in 22 (17.7%). There was no patient with diffuse pleural thickening or calcification in this study. Trivial, mild, and moderate emphysema was found in 31 (25.0%), 21 (16.9%), and 12 (9.7%) patients, respectively. Severe pleural adhesion was found in 31 (25.0%) patients. The association of localized pleural thickening or calcification on CT with severe pleural adhesion was not found (P = 0.405 and 0.107, respectively). Size of pulmonary calcified nodules and extent of emphysema were significant variables in a univariate analysis (P = 0.045 and 0.005, respectively). In a multivariate analysis, moderate emphysema was significantly associated with severe pleural adhesion (odds ratio of 11.202, P = 0.001). In conclusion, severe pleural adhesion might be found during lung cancer surgery, provided that preoperative chest CT shows substantial pulmonary calcified nodules or emphysema.

Measurement Variability of Persistent Pulmonary Subsolid Nodules on Same-Day Repeat CT: What Is the Threshold to Determine True Nodule Growth during Follow-Up?

PURPOSE: To assess the measurement variability of subsolid nodules (SSNs) in follow-up situations and to compare the degree of variability between measurement metrics. METHODS: Two same-day repeat-CT scans of 69 patients (24 men and 45 women) with 69 SSNs were randomly assigned as initial or follow-up scans and were read by the same (situation 1) or different readers (situation 2). SSN size and solid portion size were measured in both situations. Measurement variability was calculated and coefficients of variation were used for comparisons. RESULTS: Measurement variability for the longest and average diameter of SSNs was ±1.3 mm (±13.0%) and ±1.3 mm (±14.4%) in situation 1, and ±2.2 mm (±21.0%) and ±2.1 mm (±21.3%) in situation 2, respectively. For solid portion, measurement variability on lung and mediastinal windows was ±1.2 mm (±27.1%) and ±0.8 mm (±24.0%) in situation 1, and ±3.7 mm (±61.0%) and ±1.5 mm (±47.3%) in situation 2, respectively. There were no significant differences in the degree of variability between the longest and average diameters and between the lung and mediastinal window settings (p>0.05). However, measurement variability significantly increased when the follow-up and initial CT readers were different (p<0.001). CONCLUSIONS: A cutoff of ±2.2 mm can be reliably used to determine true nodule growth on follow-up CT. Solid portion measurements were not reliable in evaluating SSNs' change when readers of initial and follow-up CT were different.

Differentiation of large (≥ 5 cm) gastrointestinal stromal tumors from benign subepithelial tumors in the stomach: radiologists' performance using CT.

PURPOSE: To identify significant CT findings for the differentiation of large (≥ 5 cm) gastric gastrointestinal stromal tumors (GIST) from benign subepithelial tumors and to assess whether radiologists' performance in differentiation is improved with knowledge of significant CT criteria. MATERIALS AND METHODS: One-hundred twenty patients with pathologically proven large (≥ 5 cm) GISTs (n=99), schwannomas (n=16), and leiomyomas (n=5) who underwent CT were enrolled. Two radiologists (A and B) retrospectively reviewed their CT images in consensus for the location, size, degree and pattern of enhancement, contour, growth pattern and the presence of calcification, necrosis, surface ulceration, or enlarged lymph nodes. CT findings considered significant for differentiation were determined using uni- and multivariate statistical analyses. Thereafter, two successive review sessions for the differentiation of GIST from non-GIST were independently performed by two other reviewers (C and D) with different expertise of 2 and 9 years using a 5-point confidence scale. At the first session, reviewers interpreted CT images without knowledge of significant CT findings. At the second session, the results of statistical analyses were provided to the reviewers. To assess improvement in radiologists' performance, a pairwise comparison of receiver operating curves (ROC) was performed. RESULTS: Heterogeneous enhancement, presence of necrosis, absence of lymph nodes, and mean size of ≥ 6 cm were found to be significant for differentiating GIST from schwannoma (P<0.05). Non-cardial location, heterogeneous enhancement, and presence of necrosis were differential CT features of GIST from leiomyoma (P<0.05). Multivariate analyses indicated that absence of enlarged LNs was the only statistically significant variable for GIST differentiating from schwannoma. The area under the curve of both reviewers obtained using ROC significantly increased from 0.682 and 0.613 to 0.903 and 0.904, respectively, with information of the significant CT findings differentiating GISTs from non-GISTs (P<0.001). CONCLUSION: Non-cardial location, heterogeneous enhancement, presence of necrosis, larger lesion size, and absence of lymphadenopathy are highly suggestive CT findings for large GISTs in differentiation from schwannomas or leiomyomas. Regardless of radiologists' expertise, diagnostic performance in differentiation can be significantly improved with knowledge of these CT findings.