Search

This retrospective study investigated whether repetitive exposure to intravenous iodinated contrast media (ICM) affects long-term renal function in patients who undergo curative surgery for early gastric cancer (EGC) collected from the Korean Health Insurance and Review Assessment (HIRA) database. Patients diagnosed with gastric cancer between January 2010 and December 2013 underwent regular computed tomography (CT) scans to monitor for extragastric recurrence. Patients who already had chronic kidney disease (CKD) before cancer diagnosis or had undergone chemotherapy or repeated surgery were excluded. A nested case–control study design was chosen to analyze the effect of repetitive ICM exposure to long-term renal function by comparing patients who developed CKD 2 years after cancer diagnosis and patients who did not. Among 59,971 patients collected according to inclusion and exclusion criteria, 1021 were diagnosed with CKD 2 years after cancer diagnosis. Using 1:5 matching after adjusting for age, sex and date of cancer diagnosis, 5097 control patients were matched to 1021 CKD patients. Conditional logistic regression showed that the number of CTs taken using ICM slightly increased the odds of CKD (odds ratio, 1.080; 95% confidence interval (CI): 1.059, 1.100; P < 0.0001). Thus, the administration of ICM might contribute to chronic renal function impairment. [ABSTRACT FROM AUTHOR]

Objectives: The aim of this retrospective study was to predict circumferential resection margin (CRM) involvement on preoperative CT, and prognostic impact of CRM assessment by CT (ctCRM) in patients with retroperitonealized colon cancer. Methods: This study included patients who underwent resection for ascending or descending colon cancer between July 2010 and February 2013. Positive ctCRM was defined as tumor distance to the retromesenteric plane of ≤ 1 mm. The origin of positive CRM was divided into primary tumor or other tumor components including lymph nodes, tumor deposits, or extramural venous invasions. Logistic regression analysis was performed to identify preoperative factors to predict pathologic CRM (pCRM). A Cox proportional hazards model was used in multivariable analysis to determine the preoperative factors affecting disease-free survival (DFS). Results: A total of 274 patients (mean age, 64.0 years ± 11.0 [standard deviation]; 157 men) with retroperitonealized colon cancer were evaluated. Of 274 patients, 67 patients (24.5%) had positive CRM on surgical pathology. The accuracy of preoperative CT in predicting pCRM was 79.6% (218/274). Among preoperative factors, only CRM assessment on CT was independently associated with pCRM (p < 0.001). Positive ctCRM by primary tumor was an independent factor for DFS (HR, 3.362 [1.714–6.593]) and systemic recurrence (HR, 3.715 [1.787–7.724], but not for local recurrence on multivariable analyses. Conclusions: Preoperative CT can accurately predict pCRM, and positive ctCRM by primary tumor is an independent risk factor for DFS and systemic recurrence, but not for local recurrence in retroperitonealized colon cancer. Key Points: • Preoperative CT can predict pathologic circumferential resection margin (CRM) with approximately 80% of accuracy in patients with retroperitonealized colon cancer. • Positive CRM by a primary tumor on preoperative CT is a poor prognostic factor for disease-free survival and systemic recurrence in patients with retroperitonealized colon cancer. • CRM involvement on CT was not associated with local recurrence in patients with retroperitonealized colon cancer. [ABSTRACT FROM AUTHOR]

Objective: To evaluate diagnostic performance of a radiomics model for classifying hepatic cyst, hemangioma, and metastasis in patients with colorectal cancer (CRC) from portal-phase abdominopelvic CT images. Methods: This retrospective study included 502 CRC patients who underwent contrast-enhanced CT and contrast-enhanced liver MRI between January 2005 and December 2010. Portal-phase CT images of training (n = 386) and validation (n = 116) cohorts were used to develop a radiomics model for differentiating three classes of liver lesions. Among multiple handcrafted features, the feature selection was performed using ReliefF method, and random forest classifiers were used to train the selected features. Diagnostic performance of the developed model was compared with that of four radiologists. A subgroup analysis was conducted based on lesion size. Results: The radiomics model demonstrated significantly lower overall and hemangioma- and metastasis-specific polytomous discrimination index (PDI) (overall, 0.8037; hemangioma-specific, 0.6653; metastasis-specific, 0.8027) than the radiologists (overall, 0.9622–0.9680; hemangioma-specific, 0.9452–0.9630; metastasis-specific, 0.9511–0.9869). For subgroup analysis, the PDI of the radiomics model was different according to the lesion size (< 10 mm, 0.6486; ≥ 10 mm, 0.8264) while that of the radiologists was relatively maintained. For classifying metastasis from benign lesions, the radiomics model showed excellent diagnostic performance, with an accuracy of 84.36% and an AUC of 0.9426. Conclusion: Albeit inferior to the radiologists, the radiomics model achieved substantial diagnostic performance when differentiating hepatic lesions from portal-phase CT images of CRC patients. This model was limited particularly to classifying hemangiomas and subcentimeter lesions. Key Points: • Albeit inferior to the radiologists, the radiomics model could differentiate cyst, hemangioma, and metastasis with substantial diagnostic performance using portal-phase CT images of colorectal cancer patients. • The radiomics model demonstrated limitations especially in classifying hemangiomas and subcentimeter liver lesions. [ABSTRACT FROM AUTHOR]

Purpose: We propose a deep learning method that classifies focal liver lesions (FLLs) into cysts, hemangiomas, and metastases from portal phase abdominal CT images. We propose a synthetic data augmentation process to alleviate the class imbalance and the Lesion INformation Augmented (LINA) patch to improve the learning efficiency. Methods: A dataset of 502 portal phase CT scans of 1,290 FLLs was used. First, to alleviate the class imbalance and to diversify the training data patterns, we suggest synthetic training data augmentation using DCGAN‐based lesion mask synthesis and pix2pix‐based mask‐to‐image translation. Second, to improve the learning efficiency of convolutional neural networks (CNNs) for the small lesions, we propose a novel type of input patch termed the LINA patch to emphasize the lesion texture information while also maintaining the lesion boundary information in the patches. Third, we construct a multi‐scale CNN through a model ensemble of ResNet‐18 CNNs trained on LINA patches of various mini‐patch sizes. Results: The experiments demonstrate that (a) synthetic data augmentation method shows characteristics different but complementary to those in conventional real data augmentation in augmenting data distributions, (b) the proposed LINA patches improve classification performance compared to those by existing types of CNN input patches due to the enhanced texture and boundary information in the small lesions, and (c) through an ensemble of LINA patch‐trained CNNs with different mini‐patch sizes, the multi‐scale CNN further improves overall classification performance. As a result, the proposed method achieved an accuracy of 87.30%, showing improvements of 10.81%p and 15.0%p compared to the conventional image patch‐trained CNN and texture feature‐trained SVM, respectively. Conclusions: The proposed synthetic data augmentation method shows promising results in improving the data diversity and class imbalance, and the proposed LINA patches enhance the learning efficiency compared to the existing input image patches. [ABSTRACT FROM AUTHOR]

Objectives: To compare the performance of current guidelines applicable to the diagnosis of hepatocellular carcinomas (HCCs) using gadoxetic acid–enhanced magnetic resonance imaging (MRI). Methods: Two hundred and forty-one hepatic lesions (149 HCCs, six other malignancies, 86 benign lesions) in 177 patients at risk of HCC without a history of previous treatment for hepatic malignancy in a tertiary center were retrospectively reviewed. Either histopathology results or long-term (> 24 months) follow-up images were used as a standard of reference. All lesions were categorized according to the Liver Imaging Reporting and Data System (LI-RADS), European Association for the Study of the Liver (EASL), Asian Pacific Association for the Study of the Liver (APASL), and Korean Liver Cancer Study Group-National Cancer Center (KLCSG-NCC) guidelines. The sensitivity and specificity thereof were assessed using a generalized estimation equation. Results: For gadoxetic acid–enhanced MRI, LI-RADS (95%, 95% confidence interval [CI] 88–98) and EASL (94%, 95% CI 86–97) yielded the highest specificity, while EASL yielded the lowest sensitivity (54% [95% CI 46–62]). APASL yielded the highest sensitivity (91% [95% CI 86–95]) with the lowest specificity (78% [95% CI 69–86]). KLCSG-NCC showed balanced sensitivity (85% [79–90]) and specificity (88% [95% CI 80–93]). Differences were more prominent in small nodules between 1 and 2 cm. Conclusion: The diagnostic performance of current guidelines for HCC on gadoxetic acid–enhanced MRI was significantly different, and a potential inverse association between sensitivity and specificity was observed. Key Points: • EASL and LI-RADS yielded the highest specificity with the lowest sensitivity, whereas APASL yielded the highest sensitivity with the lowest specificity. • Differences in the diagnostic performances of guidelines were prominent in small nodules between 1 and 2 cm. • Additional evaluation of CT findings improved the diagnostic sensitivity and accuracy of EASL and LI-RADS. Although doing so decreased specificity, it remained above 89–90%. [ABSTRACT FROM AUTHOR]

Objectives: Contrast-induced acute kidney injury (CI-AKI) is one of the leading causes of new-onset renal failure in hospitalized patients. Statin has been evaluated for its protective effect against CI-AKI but rarely in patients receiving intravenous (IV) administrations of iodine-based contrast media for enhanced computed tomography (CT).Methods: In total, 12,371 patients who underwent contrast-enhanced abdominopelvic CT were retrospectively reviewed and stratified into statin users and statin nonusers. Subgroup analyses comparing high-intensity statins with low- to moderate-intensity statins were conducted within statin users and similar comparisons were performed within statin users stratified based on baseline eGFR.Results: Overall, CI-AKI events did not occur less in statin users compared with non-statin users (p = 0.342). Within statin users, CI-AKI events did not decrease in high-intensity statin users compared with low- to moderate-intensity statin users (p = 0.355). Moreover, no significant difference in CI-AKI events was found between high-intensity statin users and low- to moderate-intensity statin users even after stratifying the patients with baseline eGFR.Conclusions: Collectively, statin was not significantly associated with the risk of CI-AKI events in patients undergoing contrast-enhanced abdominopelvic CT and high-intensity statins did not show significant association with CI-AKI over low- to moderate-intensity statins in the subgroup analysis.Key Points: • Statin is not associated with risk of CI-AKI events in patients undergoing intravenous administration of contrast-enhanced CT. • CI-AKI incidence among high-intensity statin users was not significantly different from that of low- to moderate-intensity statin users. [ABSTRACT FROM AUTHOR]

Deep learning has recently attracted widespread interest as a means of reducing noise in low-dose CT (LDCT) images. Deep convolutional neural networks (CNNs) are typically trained to transfer high-quality image features of normal-dose CT (NDCT) images to LDCT images. However, existing deep learning approaches for denoising LDCT images often overlook the statistical property of CT images. In this paper, we propose an approach to statistical image restoration for LDCT using deep learning. We introduce a loss function to incorporate the noise property in image domain derived from the noise statistics in sinogram domain. In order to capture the spatially-varying statistics of CT images, we increase the receptive fields of the neural network to cover full-size CT slices. In addition, the proposed network utilizes $z$ -directional correlation by taking multiple consecutive CT slices as input. For performance evaluation, the proposed networks are trained and validated with a public dataset consisting of LDCT-NDCT image pairs. We also perform a retrospective study by testing the networks with clinical LDCT images. The experimental results show that the denoising networks successfully reduce the noise level and restore the image details without adding artifacts. This study demonstrates that the statistical deep learning approach can restore the image quality of LDCT without loss of anatomical information. [ABSTRACT FROM AUTHOR]

Training a convolutional neural network (CNN) to reduce noise in low-dose CT (LDCT) images typically relies on supervised learning, which requires input–target pairs of noisy LDCT and corresponding full-dose CT (FDCT) images. Although previous approaches have shown promising results in LDCT image denoising, it is difficult to acquire clinical datasets of LDCT-FDCT image pairs, which require additional and unnecessary radiation dose delivery to patients. In this paper, we propose a self-supervised learning approach to training a CNN-based denoiser with LDCT images alone. As a means of self-supervision, the proposed approach searches inter-pixel correlation of LDCT images in z -direction as well as in-plane direction. To regularize the CNN-based denoiser, thicker LDCT slices are used as image priors during the self-supervised training process in our approach. For efficient self-supervised learning, we adopt a two-stage training strategy with offline pretraining and online finetuning. The proposed approach is thoroughly evaluated with public and private clinical LDCT datasets. Both image quality measures and clinical assessments indicate that the self-supervised denoising model simultaneously reduces noise level and restores anatomical information in LDCT images from the images alone. The experimental results also show that our online finetuning scheme can improve the denoising performance of supervised learning models as well as self-supervised learning models at test time. • Training a CNN-based denoiser with LDCT images and without FDCT references. • Inter- and intra-slice correlation incorporated into self-supervised learning. • Online finetuning to further improve denoising performance. • Retrospective study with LDCT patients to explore clinical applicability. [ABSTRACT FROM AUTHOR]