Multiphasic Computed Tomography Research Articles

A landmark-supervised registration framework for multi-phase CT images with cross-distillation.

Multi-phase computed tomography (CT) has become a leading modality for identifying hepatic tumors. Nevertheless, the presence of misalignment in the images of different phases poses a challenge in accurately identifying and analyzing the patient's anatomy. Conventional registration methods typically concentrate on either intensity-based features or landmark-based features in isolation, so imposing limitations on the accuracy of the registration process. In this paper, we establish a nonrigid cycle-registration network that leverages semi-supervised learning techniques, wherein a point distance term based on Euclidean distance between registered landmark points is introduced into the loss function. Additionally, a cross-distillation strategy is proposed in network training to further improve registration performance which incorporates response-based knowledge concerning the distances between feature points. We conducted experiments using multi-centered liver CT datasets to evaluate the performance of the proposed method. The results demonstrate that our method outperforms baseline methods in terms of Target Registration Error (TRE). Additionally, Dice scores of the warped tumor masks were calculated. Our method consistently achieved the highest scores among all the comparing methods. Specifically, it achieved scores of 82.9% and 82.5% in the hepatocellular carcinoma (HCC) and the intrahepatic cholangiocarcinoma (ICC) dataset, respectively. The superior registration performance indicates its potential to serve as an important tool in hepatic tumor identification and analysis.

Childhood Solid Pseudopapillary Neoplasm of the Pancreas: Acute Presentation of an Uncommon Tumor

AbstractSolid pseudopapillary tumor is a mixed solid and cystic neoplasm of the pancreas which occurs most commonly in females in second or third decade of life with few reported pediatric cases. It has a low malignancy potential with good prognosis after surgical resection.We present a case of an 11-year-old male child with solid pseudopapillary tumor of pancreas who presented with clinical features of acute pancreatitis. He was initially evaluated with ultrasound, followed by computed tomography (CT) and magnetic resonance imaging (MRI).Solid pseudopapillary tumor of pancreas presenting as acute pancreatitis is rarely reported in literature. This case is a unique presentation of a pancreatic tumor and highlights the value of multiphase CT imaging and MRI in the diagnosis.

Effect of contrast phase upon quantitative analysis of skeletal muscle and adipose tissue by computed tomography

ObjectiveSignificant variability exists in the contrast phases applied during computed tomography (CT) studies when assessing morphometric measurements of muscle area (CT-assessed sarcopenia), density (CT-assessed myosteatosis), and visceral adipose tissue area (CT-assessed visceral obesity). This study explored the impact of contrast phase timing on changes in morphometric measurements of body composition. Research Methods & ProceduresThis single-center retrospective cohort study included 459 patients undergoing a multiphase CT scan. Morphometric measurements were obtained at the third lumbar vertebra level. Patients were classified as sarcopenic, myosteatotic, or visceral obese using pre-defined cutoff values. The intra-class correlation coefficient was used to assess correlations across different enhancement phases, and Cohen's κ measured the inter-enhancement agreement for sarcopenia, myosteatosis, and visceral obesity. ResultsSignificant differences were observed in mean visceral adipose tissue area, muscle density, and muscle area (p<0.001). The intra-class correlation coefficient between unenhanced and arterial phases was 0.987 (95% confidence interval [CI], 0.759-0.996) for adipose tissue, 0.995 (95% CI, 0.989-0.997) for muscle area, and 0.850 (95% CI, 0.000-0.956) for muscle density. However, when morphometric measurements were categorized using predefined cutoffs, the κ agreement was considerably lower, particularly for CT-assessed myosteatosis, ranging from 0.635 (unenhanced to arterial) to 0.331 (unenhanced to late venous phase). ConclusionDifferent CT contrast phases induce small, but clinically significant, alterations in the measurements of muscle area, density, and visceral fat. Such minor changes can result in misclassification issues when fixed cutoff values are used to diagnose myosteatosis with computed tomography. This underscores the importance of reporting absolute values and the specific contrast phase used in future studies.

Prognostic predictive value of urothelial carcinoma of the bladder after TURBT based on multiphase CT radiomics.

To investigate multiphase computed tomography (CT) radiomics-based combined with clinical factors to predict overall survival (OS) in patients with bladder urothelial carcinoma (BLCA) who underwent transurethral resection of bladder tumor (TURBT). Data were retrospectively collected from 114 patients with primary BLCA from February 2016 to February 2018. The regions of interest (ROIs) of the plain, arterial, and venous phase images were manually segmented. The Cox regression algorithm was used to establish 3 basic models for the plain phase (PP), arterial phase (AP), and venous phase (VP) and 2 combination models (AP + VP and PP + AP + VP). The highest-performing radiomics model was selected to calculate the radiomics score (Rad-score), and independent risk factors affecting patients' OS were analyzed using Cox regression. The Rad-score and clinical risk factors were combined to construct a joint model and draw a visualized nomogram. The combined model of PP + AP + VP showed the best performance with the Akaike Information Criterion (AIC) and Consistency Index (C-index) in the test group of 130.48 and 0.779, respectively. A combined model constructed with two independent risk factors (age and Ki-67 expression status) in combination with the Rad-score outperformed the radiomics model alone; AIC and C-index in the test group were 115.74 and 0.840, respectively. The calibration curves showed good agreement between the predicted probabilities of the joint model and the actual (p < 0.05). The decision curve showed that the joint model had good clinical application value within a large range of threshold probabilities. This new model can be used to predict the OS of patients with BLCA who underwent TURBT.

Deep learning of pretreatment multiphase CT images for predicting response to lenvatinib and immune checkpoint inhibitors in unresectable hepatocellular carcinoma

ObjectivesCombination therapy of lenvatinib and immune checkpoint inhibitors (CLICI) has emerged as a promising approach for managing unresectable hepatocellular carcinoma (HCC). However, the response to such treatment is observed in only a subset of patients, underscoring the pressing need for reliable methods to identify potential responders. Materials & methodsThis was a retrospective analysis involving 120 patients with unresectable HCC. They were divided into training (n = 72) and validation (n = 48) cohorts. We developed an interpretable deep learning model using multiphase computed tomography (CT) images to predict whether patients will respond or not to CLICI treatment, based on the Response Evaluation Criteria in Solid Tumors, version 1.1 (RECIST v1.1). We evaluated the models' performance and analyzed the impact of each CT phase. Critical regions influencing predictions were identified and visualized through heatmaps. ResultsThe multiphase model outperformed the best biphase and uniphase models, achieving an area under the curve (AUC) of 0.802 (95% CI = 0.780–0.824). The portal phase images were found to significantly enhance the model's predictive accuracy. Heatmaps identified six critical features influencing treatment response, offering valuable insights to clinicians. Additionally, we have made this model accessible via a web server at http://uhccnet.com/ for ease of use. ConclusionsThe integration of multiphase CT images with deep learning-generated heatmaps for predicting treatment response provides a robust and practical tool for guiding CLICI therapy in patients with unresectable HCC.

Automatic Liver Segmentation from Multiphase CT Using Modified SegNet and ASPP Module

Liver cancer is one of the dominant causes of cancer death worldwide. Computed Tomography (CT) is the commonly used imaging modality for diagnosing it. Computer-based liver cancer diagnosis systems can assist radiologists in image interpretation and improve diagnosis speed and accuracy. Since liver segmentation is crucial to such systems, researchers are relentlessly pursuing various segmentation approaches. A clinically viable computer-aided system requires examining multiphase CT images. However, most of the research focuses only on the portal venous phase. In this work, we developed an automatic and efficient Deep Learning (DL) method using SegNet, atrous spatial pyramid pooling module and leaky ReLU layers for liver segmentation from quadriphasic abdominal CT volumes. The proposed method was validated on two datasets, an internal institutional dataset consisting of multiphase CT and a public dataset of portal venous phase CT volumes. The Dice Coefficients (DC) obtained were greater than 96% for the latter dataset and the portal venous phase of the former. For arterial, delayed and plain CT phases of the former dataset, the DC achieved were 94.61%, 95.01% and 93.23%, respectively. Experiments showed that our model performed better than the other state-of-the-art DL models. Ablation studies have revealed that the proposed model leverages the strengths of all the three components that make it up. The promising performance of the proposed method suggests that it is appropriate for incorporation in hepatic cancer diagnosis systems.

Institutional clinical indication-based typical dose values of multiphasic abdominopelvic computed tomography examinations.

Our study aimed to obtain clinical indication-based typical dose values and size-specific dose estimates (SSDEs) for multiphasic abdominopelvic computed tomography (CT) examinations and to review our data with published diagnostic reference levels (DRLs). In this retrospective study, multiphasic liver, kidney, pancreas, and mesenteric ischemia protocol CT scans performed at our center between January 2018 and December 2021 were analyzed. The clinical indications were hepatocellular carcinoma, renal cell carcinoma, pancreas adenocarcinoma, and mesenteric ischemia. The computed tomography dose index volume (CTDIvol) and dose-length product (DLP) values were recorded, and the SSDE and effective dose (ED) values were calculated. The water-equivalent diameter (Dw) value required for the SSDE calculation was measured using the automated calculation of the Dw program. The total number of patients was 514, with 86 patients excluded from this study. The dose values were calculated for 426 patients (183 female and 243 male; 111 liver, 120 kidney, 85 pancreas, and 110 mesenteric). The median values for the CTDIvol, DLP, SSDE, and ED were 6.86 mGy, 683.02 mGy. cm, 8.75 mGy, and 10.45 mSv for the liver CT; 8.37 mGy, 908.37 mGy.cm, 10.37 mGy, and 13.89 mSv for the kidney CT; 7.82 mGy, 517.98 mGy.cm, 10.01 mGy, and 7.92 mSv for the pancreas CT; and 9.48 mGy, 983.68 mGy.cm, 12.78 mGy, and 13.86 mSv for the mesenteric CT, respectively. All dose values were lower than the published DRLs. The literature reveals large differences in the multiphasic abdominopelvic CT protocols, especially in the number of phases and scan length. This situation makes comparing dose values difficult. Dose studies revealing the protocol parameters in detail are needed so that institutions can compare and optimize their own protocols. Additionally, users should periodically check the dose values in their own institutions.

An interpretable machine learning model based on contrast-enhanced CT parameters for predicting treatment response to conventional transarterial chemoembolization in patients with hepatocellular carcinoma.

To explore the potential of pre-therapy computed tomography (CT) parameters in predicting the treatment response to initial conventional TACE (cTACE) in intermediate-stage hepatocellular carcinoma (HCC) and develop an interpretable machine learning model. This retrospective study included 367 patients with intermediate-stage HCC who received cTACE as first-line therapy from three centers. We measured the mean attenuation values of target lesions on multi-phase contrast-enhanced CT and further calculated three CT parameters, including arterial (AER), portal venous (PER), and arterial portal venous (APR) enhancement ratios. We used logistic regression analysis to select discriminative features and trained three machine learning models via 5-fold cross-validation. The performance in predicting treatment response was evaluated in terms of discrimination, calibration, and clinical utility. Afterward, a Shapley additive explanation (SHAP) algorithm was leveraged to interpret the outputs of the best-performing model. The mean diameter, ECOG performance status, and cirrhosis were the important clinical predictors of cTACE treatment response, by multiple logistic regression. Adding the CT parameters to clinical variables showed significant improvement in performance (net reclassification index, 0.318, P < 0.001). The Random Forest model (hereafter, RF-combined model) integrating CT parameters and clinical variables demonstrated the highest performance on external validation dataset (AUC of 0.800). The decision curve analysis illustrated the optimal clinical benefits of RF-combined model. This model could successfully stratify patients into responders and non-responders with distinct survival (P = 0.001). The RF-combined model can serve as a robust and interpretable tool to identify the appropriate crowd for cTACE sessions, sparing patients from receiving ineffective and unnecessary treatments.

Customised weight-based volume contrast media protocol for multiphase abdominal computed tomography.

Multiphase computed tomography (CT) using fixed volume contrast media may lead to high radiation exposure and toxicity in patients with low body weight. We evaluated a customised weight-based protocol for multiphase CT in terms of radiation exposure, image quality and cost savings. A total of 224 patients were recruited. An optimised CT protocol was applied using 100 kV and 1 mL/kg of contrast media dosing. The image quality and radiation dose exposure of this CT protocol were compared to those of a standard 120 kV, 80 mL fixed volume protocol. The radiation dose information and CT Hounsfield units were recorded. The signal-to-noise ratio, contrast-to-noise ratio (CNR) and figure of merit (FOM) were used as comparison metrics. The images were assessed for contrast opacification and visual quality by two radiologists. The renal function, contrast media volume and cost were also evaluated. The median effective dose was lowered by 16% in the optimised protocol, while the arterial phase images achieved significantly higher CNR and FOM. The radiologists' evaluation showed more than 97% absolute agreement with no significant differences in image quality. No significant differences were found in the pre- and post-CT estimated glomerular filtration rate. However, contrast media usage was significantly reduced by 1,680 mL, with an overall cost savings of USD 421 in the optimised protocol. The optimised weight-based protocol is cost-efficient and lowers radiation dose while maintaining overall contrast enhancement and image quality.

Systematic training of LI-RADS CT v2018 improves interobserver agreements and performances in LR categorization for focal liver lesions.

To retrospectively explored whether systematic training in the use of Liver Imaging Reporting and Data System (LI-RADS) v2018 on computed tomography (CT) can improve the interobserver agreements and performances in LR categorization for focal liver lesions (FLLs) among different radiologists. A total of 18 visiting radiologists and the liver multiphase CT images of 70 hepatic observations in 63 patients at high risk of HCC were included in this study. The LI-RADS v2018 training procedure included three thematic lectures, with an interval of 1 month. After each seminar, the radiologists had 1 month to adopt the algorithm into their daily work. The interobserver agreements and performances in LR categorization for FLLs among the radiologists before and after training were compared. After training, the interobserver agreements in classifying the LR categories for all radiologists were significantly increased for most LR categories (P < 0.001), except for LR-1 (P = 0.053). After systematic training, the areas under the curve (AUCs) for LR categorization performance for all participants were significantly increased for most LR categories (P < 0.001), except for LR-1 (P = 0.062). Systematic training in the use of the LI-RADS can improve the interobserver agreements and performances in LR categorization for FLLs among radiologists with different levels of experience.

Enhancing preoperative diagnosis of microvascular invasion in hepatocellular carcinoma: domain-adaptation fusion of multi-phase CT images

ObjectivesIn patients with hepatocellular carcinoma (HCC), accurately predicting the preoperative microvascular invasion (MVI) status is crucial for improving survival rates. This study proposes a multi-modal domain-adaptive fusion model based on deep learning methods to predict the preoperative MVI status in HCC.Materials and methodsFrom January 2008 to May 2022, we collected 163 cases of HCC from our institution and 42 cases from another medical facility, with each case including Computed Tomography (CT) images from the pre-contrast phase (PCP), arterial phase (AP), and portal venous phase (PVP). We divided our institution’s dataset (n=163) into training (n=119) and test sets (n=44) in an approximate 7:3 ratio. Additionally, we included cases from another institution (n=42) as an external validation set (test1 set). We constructed three single-modality models, a simple concatenated multi-modal model, two current state-of-the-art image fusion model and a multi-modal domain-adaptive fusion model (M-DAFM) based on deep learning methods. We evaluated and analyzed the performance of these constructed models in predicting preoperative MVI using the area under the receiver operating characteristic curve (AUC), decision curve analysis (DCA), and net reclassification improvement (NRI) methods.ResultsIn comparison with all models, M-DAFM achieved the highest AUC values across the three datasets (0.8013 for the training set, 0.7839 for the test set, and 0.7454 for the test1 set). Notably, in the test set, M-DAFM’s Decision Curve Analysis (DCA) curves consistently demonstrated favorable or optimal net benefits within the 0-0.65 threshold probability range. Additionally, the Net Reclassification Improvement (NRI) values between M-DAFM and the three single-modal models, as well as the simple concatenation model, were all greater than 0 (all p &amp;lt; 0.05). Similarly, the NRI values between M-DAFM and the two current state-of-the-art image fusion models were also greater than 0. These findings collectively indicate that M-DAFM effectively integrates valuable information from multi-phase CT images, thereby enhancing the model’s preoperative predictive performance for MVI.ConclusionThe M-DAFM proposed in this study presents an innovative approach to improve the preoperative predictive performance of MVI.

Computed tomography for visible haematuria - a single nephrographic phase is sufficient for detecting renal cell carcinoma.

No previous studies have compared two computed tomography (CT) protocols in patients presenting with visible haematuria, and most patients undergo a multiphase CT in order todetect upper tract malignancies. We aimed to prospectively compare the diagnostic performance ofsingle- and four-phase CT for detecting renal cell carcinoma (RCC) in patients with visible haematuria. 'A Prospective Trial for Examining Hematuria using Computed Tomography' (PROTEHCT) was a single-centre prospective paired diagnostic study in patients referred for CT due to painless visible haematuria between September 2019 and June 2021. All patients underwent four-phase CT (control) from which a single nephrographic phase dual energy CT (experimental) was extracted. Both were independently assessed for RCC by randomised radiologists. Histologically verified RCC defined a positive reference standard. Follow-up ascertainment of RCC diagnosis was completed in May 2022. Descriptive statistics were used to calculate the accuracies. Inter-reader agreement was assessed by kappa statistics. A total of 308 patients (median age, 68 years [interquartile range 53-77, range 18-96], 250 males) were included for analysis. RCC was diagnosed in seven (2.3%) patients during a median follow-up time of 19 months (interquartile range: 15-25). For the control and experimental CT, sensitivity was 100% versus 100%, specificity was 97% versus 98% and accuracy 97% versus 97%. The positive predictive value was 44% versus 50%, and the negative predictive value was 100% versus 100%. The agreement between the control and experimental CT was 98% (k = 0.79). A single nephrographic phase dual energy CT is sufficient for detecting RCC in patients with visible haematuria.

Multi-phase features interaction transformer network for liver tumor segmentation and microvascular invasion assessment in contrast-enhanced CT

&lt;abstract&gt; &lt;p&gt;Precise segmentation of liver tumors from computed tomography (CT) scans is a prerequisite step in various clinical applications. Multi-phase CT imaging enhances tumor characterization, thereby assisting radiologists in accurate identification. However, existing automatic liver tumor segmentation models did not fully exploit multi-phase information and lacked the capability to capture global information. In this study, we developed a pioneering multi-phase feature interaction Transformer network (MI-TransSeg) for accurate liver tumor segmentation and a subsequent microvascular invasion (MVI) assessment in contrast-enhanced CT images. In the proposed network, an efficient multi-phase features interaction module was introduced to enable bi-directional feature interaction among multiple phases, thus maximally exploiting the available multi-phase information. To enhance the model's capability to extract global information, a hierarchical transformer-based encoder and decoder architecture was designed. Importantly, we devised a multi-resolution scales feature aggregation strategy (MSFA) to optimize the parameters and performance of the proposed model. Subsequent to segmentation, the liver tumor masks generated by MI-TransSeg were applied to extract radiomic features for the clinical applications of the MVI assessment. With Institutional Review Board (IRB) approval, a clinical multi-phase contrast-enhanced CT abdominal dataset was collected that included 164 patients with liver tumors. The experimental results demonstrated that the proposed MI-TransSeg was superior to various state-of-the-art methods. Additionally, we found that the tumor mask predicted by our method showed promising potential in the assessment of microvascular invasion. In conclusion, MI-TransSeg presents an innovative paradigm for the segmentation of complex liver tumors, thus underscoring the significance of multi-phase CT data exploitation. The proposed MI-TransSeg network has the potential to assist radiologists in diagnosing liver tumors and assessing microvascular invasion.&lt;/p&gt; &lt;/abstract&gt;

Serial assessment of computed tomography angiography for pulmonary and systemic arteries using a reduced contrast agent dose for the diagnosis of systemic artery-to-pulmonary artery shunts

PurposeTo evaluate the diagnostic performance and feasibility of a modified computed tomography (CT) scan protocol, we performed a serial assessment of the computed tomography angiography for pulmonary artery (CTA-P) and systemic artery (CTA-S) (CTA-PS) using a reduced contrast agent dose to diagnose systemic artery-to-pulmonary artery shunts (SPSs).Materials and methodsTwenty-five patients who underwent multiphase contrast-enhanced chest CT and conventional chest angiography were included. Three image sets (CTA-P, CTA-S, and CTA-PS) were evaluated by two board-certified radiologists. The visualization of the CT image findings associated with SPSs, such as filling defects and enhancement in the pulmonary arteries, was evaluated using a 5-point scale.ResultsThe diagnostic performance (sensitivity, specificity, and accuracy) of CT imaging findings associated with SPSs in CTA-P and CTA-PS were as follows: CTA-P, 57.1%, 87.5%, and 62.0%; CTA-PS, 81.0%, 100.0%, and 84.0%. CT findings associated with SPSs in CTA-P were significantly sensitive to the CTA-PS protocol. There were no significant differences between the CTA-S and CTA-PS protocols. The area under the curve (AUC) of the CT imaging findings associated with SPSs in the CTA-P and CTA-PS groups was 0.835 and 0.911, respectively (P = 0.191). The AUC of the CT imaging findings associated with SPSs in CTA-S and CTA-PS were 0.891 and 0.926, respectively (P = 0.373).ConclusionCTA-PS using a reduced contrast agent dose protocol could improve the overall diagnostic confidence of SPSs, owing to better visualization of CT imaging findings associated with SPSs compared to individual assessments of CTA-P or CTA-S. Therefore, CTA-PS can be used as an alternative preembolization evaluation modality to conventional angiography in patients with hemoptysis suspected of having SPSs.

Convolutional Neural Network Model for Segmentation and Classification of Clear Cell Renal Cell Carcinoma Based on Multiphase CT Images.

(1) Background: Computed tomography (CT) imaging challenges in diagnosing renal cell carcinoma (RCC) include distinguishing malignant from benign tissues and determining the likely subtype. The goal is to show the algorithm's ability to improve renal cell carcinoma identification and treatment, improving patient outcomes. (2) Methods: This study uses the European Deep-Health toolkit's Convolutional Neural Network with ECVL, (European Computer Vision Library), and EDDL, (European Distributed Deep Learning Library). Image segmentation utilized U-net architecture and classification with resnet101. The model's clinical efficiency was assessed utilizing kidney, tumor, Dice score, and renal cell carcinoma categorization quality. (3) Results: The raw dataset contains 457 healthy right kidneys, 456 healthy left kidneys, 76 pathological right kidneys, and 84 pathological left kidneys. Preparing raw data for analysis was crucial to algorithm implementation. Kidney segmentation performance was 0.84, and tumor segmentation mean Dice score was 0.675 for the suggested model. Renal cell carcinoma classification was 0.885 accurate. (4) Conclusion and key findings: The present study focused on analyzing data from both healthy patients and diseased renal patients, with a particular emphasis on data processing. The method achieved a kidney segmentation accuracy of 0.84 and mean Dice scores of 0.675 for tumor segmentation. The system performed well in classifying renal cell carcinoma, achieving an accuracy of 0.885, results which indicates that the technique has the potential to improve the diagnosis of kidney pathology.

Hepatocellular Carcinoma from a Hepatologist's Perspective.

Hepatocellular carcinoma (HCC), the most common type of primary liver cancer, represents a growing health challenge worldwide. The incidence of HCC is rising, which, in turn, has led to a corresponding increase in the associated number of deaths. HCC will become the third leading cause of cancer-related deaths in the United States by 2030. HCC usually develops in the setting of chronic liver disease. Individuals at increased risk of HCC are recommended to undergo surveillance with ultrasound every 6 months along with serum α-fetoprotein testing. Computed tomography (CT) and magnetic resonance imaging (MRI) are considered alternatives based on specific patient factors. Lesions suspicious for HCC are recommended to undergo a diagnostic testing, which includes contrast-enhanced multiphase CT or MRI and liver biopsy when findings are indeterminate. The Barcelona Clinic Liver Cancer prognosis and treatment strategy is the most used assessment for patients with HCC ( Fig. 2 ). Curative therapies include resection, liver transplantation, and ablation. Locoregional therapies, such as transarterial chemoembolization and radioembolization, can be used for patients with intermediate-stage HCC. For patients with advanced-stage HCC, systemic therapy is often used. This review aims to provide an overview of HCC from a hepatologist's perspective, including epidemiology, screening, surveillance, diagnosis, and management.

Wavelet radiomics features from multiphase CT images for screening hepatocellular carcinoma: analysis and comparison

Early detection of liver malignancy based on medical image analysis plays a crucial role in patient prognosis and personalized treatment. This task, however, is challenging due to several factors, including medical data scarcity and limited training samples. This paper presents a study of three important aspects of radiomics feature from multiphase computed tomography (CT) for classifying hepatocellular carcinoma (HCC) and other focal liver lesions: wavelet-transformed feature extraction, relevant feature selection, and radiomics features-based classification under the inadequate training samples. Our analysis shows that combining radiomics features extracted from the wavelet and original CT domains enhance the classification performance significantly, compared with using those extracted from the wavelet or original domain only. To facilitate the multi-domain and multiphase radiomics feature combination, we introduce a logistic sparsity-based model for feature selection with Bayesian optimization and find that the proposed model yields more discriminative and relevant features than several existing methods, including filter-based, wrapper-based, or other model-based techniques. In addition, we present analysis and performance comparison with several recent deep convolutional neural network (CNN)-based feature models proposed for hepatic lesion diagnosis. The results show that under the inadequate data scenario, the proposed wavelet radiomics feature model produces comparable, if not higher, performance metrics than the CNN-based feature models in terms of area under the curve.

Incidental Indeterminate Renal Lesions: Distinguishing Non-Enhancing from Potential Enhancing Renal Lesions Using Iodine Quantification on Portal Venous Dual-Layer Spectral CT.

The purpose of our study is to determine a threshold for iodine quantification to distinguish definitely non-enhancing benign renal lesions from potential enhancing masses on portal venous dual-layer spectral computed tomography (CT) to reduce the need for additional multiphase CT. In this single-center retrospective study, patients (≥18 years) scanned between April 2021 and January 2023 following the local renal CT protocol were included. Exclusion criteria were patients without renal lesions, lesions smaller than 10 mm, only fat-containing lesions, abscesses or infarction, follow-up after radiofrequent ablation, wrong scan protocol, or artefacts. Scans were performed on a dual layer detector-based spectral CT (CT 7500, Philips Healthcare, Best, The Netherlands). Iodine concentration (mgI/mL) in renal lesions was determined using spectral data. Analyses were performed for all lesions and for lesions of >30 HU on portal venous CT. Enhancement on multiphase CT (≥20 ΔHU from true unenhanced (TUE) to portal venous phase (PVP) CT) was used as reference standard. To determine thresholds for iodine concentration receiver operating characteristic (ROC) curves, area under the curve (AUC) and 95% confidence intervals were calculated. To obtain thresholds for definite (non-)enhancement, 100% sensitivity with maximum specificity and 100% specificity with maximum sensitivity were noted. Data were measured using one reader. To assess interobserver agreement, a second reader performed measurements on the PVP CT scans. A total of 103 patients (62 years ± 14, 68 men) were included. We measured 328 renal lesions, 56 enhancing lesions (17%) in 38 patients and 272 non-enhancing lesions (83%) in 86 patients. The threshold for non-enhancing lesions was 0.76 mgI/mL or lower (100% sensitivity, 76% specificity). The threshold for a definite enhancing mass was 1.69 mgI/mL or higher (100% specificity, 78% sensitivity). A total of 77% of indeterminate lesions (>30 HU on PVP CT) in our study could be definitely characterized. Renal lesions can be definitively classified as non-enhancing or enhancing on PVP spectral CT using thresholds of 0.76 mgI/mL or 1.69 mgI/mL, respectively, eliminating the need for multiphase imaging.

Segmental analysis of liver cirrhosis with different etiologies: a study based on iodine mixed imaging in port-venous phase.

Computed tomography (CT) in port-venous phase can display the intra-hepatic vessels, and may provide the possibility for segment function evaluation for cirrhosis. To assess the value of iodine mixed imaging of dual-source dual-energy CT in port-venous phase in segmental evaluation of liver cirrhosis with different etiologies. Patients diagnosed with liver cirrhosis were enrolled. Patients without cirrhosis were included as a control group. Each patient underwent iodine-contrast enhanced multi-phase dual-energy CT scanning. Parameters were analyzed by SPSS, version 22.0, and Medcalc. In total, 256 patients were investigated, including 114 Child-Pugh A, 51 Child-Pugh B, 41 Child-Pugh C and 50 control patients. Total iodine content (ICt)/body surface area (BSA) in the cirrhosis group was significantly lower than the control group (P < 0.05) and the standardized-iodine parameter (SI) of each segment decreased with cirrhosis progression. In Child-Pugh A and B, SI increased more significantly in the caudal and lateral segment in A (alcholism) than in the V (virus-related) and N (non-alcoholic steatohepatitis) groups (P < 0.001). ICt/BSA showed the best diagnosis power of cirrhosis with an area under the curve of 0.765, sensitivity of 76.0% and specificity of 71.8%. Blood flow compensated in the left lateral and caudal lobe in the early stage of liver cirrhosis. The compensation in alcoholism in the middle and early stages is significantly higher than that of V and N cirrhosis. Iodine mixed imaging in portal phase may provide the possibility of an incremental value in segmented blood flow perfusion and functional evaluation of liver cirrhosis on a morphological basis.

Deep learning to predict respiratory lung-tissue displacement from multi-phase pulmonary computed tomography images

Lung surgery, particularly lung puncture, is challenging because respiration motion significantly displaces lung tissues (e.g., blood vessels, trachea), including targeted lesions. To improve the success of lung surgery, it is essential to determine how lung tissues move during respiration. Early work used simple mathematical models (e.g., linear models) to predict lung-tissue displacement. However, the quality and quantity of pulmonary computed tomography (CT) images used to build such models were limited, preventing their universal application. Moreover, such mathematical models are too simple to describe how lung tissues are displaced during respiration. In this work, we analyzed the relationship between body-surface and lung-tissue displacement in a large number of multi-phase pulmonary CT images to create a model predicting lung-tissue displacement. First, a systematic method based on unsupervised learning approaches was designed to calculate the displacement of the body surface and lung tissues. Second, a deep neural network model was developed to predict lung-tissue displacement, using the body-surface displacement observed in CT imaging data as input. Multi-phase pulmonary CT images from 199 patients were collected to train and test our model. The validation experiment demonstrated that the prediction accuracy based on our deep learning model reached 79.8%, which outperformed traditional machine learning approaches like the linear model (79.2%), linear support vector regression model (79.2%), and support vector regression model with radial basis function (67.2%). These results demonstrated that our method could predict lung-tissue displacement based on body-surface displacement during respiration, potentially improving the success of lung puncture procedures.