Classification Of Focal Liver Lesions Research Articles

Training on contrast-enhanced ultrasound LI-RADS classification for resident radiologists: a retrospective comparison of performance after training

ObjectivesTo evaluate the effects and benefits of training radiology residents on contrast-enhanced ultrasound (CEUS) according to the Liver Imaging Reporting and Data System (LI-RADS).MethodsIn total, 234 patients at high risk of hepatocellular carcinoma (HCC) who underwent CEUS were enrolled, including 27 lesions in the education set and 207 lesions in the test sets (a–d). Forty-five radiology residents and 4 radiology experts involved in CEUS LI-RADS training individually reviewed the test sets before, immediately after, and 3-months after training. The consistency with kappa values of the description of CEUS features, the classification of focal liver lesions (FLLs), and the diagnostic performance were evaluated.ResultsThe level of agreement between the radiology experts and residents improved after training (all p < 0.05), while there were no significant differences between the post-training and 3-months post-training results (all p > 0.05). The sensitivity, specificity, positive predictive value, and area under the curve (AUC) based on the CEUS LI-RADS classification of the radiology experts in the diagnosis of HCC were 62.9%, 96.4%, 96.3%, and 0.796, respectively. The diagnostic performance of the radiology residents significantly improved after training (all p < 0.05). Misunderstanding of definitions and subjective interpretation of images were the main reasons for disagreement with multiple responses.ConclusionDedicated CEUS LI-RADS training improved the performance of radiology residents in diagnosing FLLs and their agreement with radiology experts on CEUS features. Images and videos to explain typical features of the training were essential to improve agreement between the radiology experts and residents.Critical relevance statementAgreement on lesion descriptors between radiology experts and residents can improve with training.Key PointsThe diagnostic performance of less experienced radiologists for diagnosing HCC could be improved by training.Images and videos to explain typical features during training were essential.Agreement on lesion descriptors between radiology experts and residents improved after training.Graphical

Deep Learning Based Classification of Focal Liver Lesions with 3 and 4 Phase Contrast-Enhanced CT Protocols

Deep Learning Based Classification of Focal Liver Lesions with 3 and 4 Phase Contrast-Enhanced CT Protocols

Classification of focal liver lesions in CT images using convolutional neural networks with lesion information augmented patches and synthetic data augmentation.

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. 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. 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. 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.

Abdominal T2-Weighted Imaging and T2 Mapping Using a Variable Flip Angle Radial Turbo Spin-Echo Technique.

T2 mapping is of great interest in abdominal imaging but current methods are limited by low resolution, slice coverage, motion sensitivity, or lengthy acquisitions. Develop a radial turbo spin-echo technique with refocusing variable flip angles (RADTSE-VFA) for high spatiotemporal T2 mapping and efficient slice coverage within a breath-hold and compare to the constant flip angle counterpart (RADTSE-CFA). Prospective technical efficacy. Testing performed on agarose phantoms and 12 patients. Focal liver lesion classification tested on malignant (N=24) and benign (N=11) lesions. 1.5 T/RADTSE-VFA, RADTSE-CFA. A constrained objective function was used to optimize the refocusing flip angles. Phantom and/or in vivo data were used to assess relative contrast, T2 estimation, specific absorption rate (SAR), and focal liver lesion classification. STATISTICAL TESTS: t-Tests or Mann-Whitney Rank Sum tests were used. Phantom data did not show significant differences in mean relative contrast (P=0.10) and T2 accuracy (P=0.99) between RADTSE-VFA and RADTSE-CFA. Adding noise caused T2 overestimation predominantly for RADTSE-CFA and low T2 values. In vivo results did not show significant differences in mean spleen-to-liver (P=0.62) and kidney-to-liver (P=0.49) relative contrast between RADTSE-VFA and RADTSE-CFA. Mean T2 values were not significantly different between the two techniques for spleen (T2VFA =109.2 ± 12.3msec; T2CFA =110.7 ± 11.1msec; P=0.78) and kidney-medulla (T2VFA =113.0 ± 8.7msec; T2CFA =114.0 ± 8.6msec; P=0.79). Liver T2 was significantly higher for RADTSE-CFA (T2VFA =52.6 ± 6.6msec; T2CFA =60.4 ± 8.0msec) consistent with T2 overestimation in the phantom study. Focal liver lesion classification had comparable T2 distributions for RADTSE-VFA and RADTSE-CFA for malignancies (P=1.0) and benign lesions (P=0.39). RADTSE-VFA had significantly lower SAR than RADTSE-CFA increasing slice coverage by 1.5. RADTSE-VFA provided noise-robust T2 estimation compared to the constant flip angle counterpart while generating T2-weighted images with comparable contrast. The VFA scheme minimized SAR improving slice efficiency for breath-hold imaging. 2 TECHNICAL EFFICACY STAGE: 1.

Contrast-enhanced ultrasound is useful for the evaluation of focal liver lesions in children.

IntroductionContrast‐enhanced ultrasound (CEUS) is a widely used diagnostic method. In adults, it has been proven to be a useful alternative to CT and MRI for the characterisation of focal liver lesions (FLLs). However, since there is no official paediatric licensing for any ultrasound contrast agents in Europe, its use has been restricted.PurposeTo retrospectively outline our experience with CEUS as a tool for the characterisation of FLLs in paediatric patients.MethodsAn eleven‐year retrospective single‐centre study. During this period, we identified 287 CEUS examinations performed on children, of these 36 were relevant first‐time examinations with the aim of characterising a focal liver lesion. Clinical and radiological data were collected from the hospital chart.ResultsThe overall agreement between the CEUS diagnosis and the reference diagnosis for benign versus malignant differentiation was 75%. When analysing conclusive CEUS examinations only, the overall agreement was 96%. The specificity for correctly characterising a lesion as benign was 96%, and the negative predictive value was 100%. No side effects from CEUS were detected.ConclusionsOur study reinforces that CEUS can be useful in the medical workup for the identification and classification of focal liver lesions in children.

Automatic Detection and Classification of Focal Liver Lesions Based on Deep Convolutional Neural Networks: A Preliminary Study

With the increasing daily workload of physicians, computer-aided diagnosis (CAD) systems based on deep learning play an increasingly important role in pattern recognition of diagnostic medical images. In this paper, we propose a framework based on hierarchical convolutional neural networks (CNNs) for automatic detection and classification of focal liver lesions (FLLs) in multi-phasic computed tomography (CT). A total of 616 nodules, composed of three types of malignant lesions (hepatocellular carcinoma, intrahepatic cholangiocarcinoma, and metastasis) and benign lesions (hemangioma, focal nodular hyperplasia, and cyst), were randomly divided into training and test sets at an approximate ratio of 3:1. To evaluate the performance of our model, other commonly adopted CNN models and two physicians were included for comparison. Our model achieved the best results to detect FLLs, with an average test precision of 82.8%, recall of 93.4%, and F1-score of 87.8%. Our model initially classified FLLs into malignant and benign and then classified them into more detailed classes. For the binary and six-class classification, our model achieved average accuracy results of 82.5 and73.4%, respectively, which were better than the other three classification neural networks. Interestingly, the classification performance of the model was placed between a junior physician and a senior physician. Overall, this preliminary study demonstrates that our proposed multi-modality and multi-scale CNN structure can locate and classify FLLs accurately in a limited dataset, and would help inexperienced physicians to reach a diagnosis in clinical practice.

P.438 The capacity of schizophrenia and bipolar disorder individuals for autonomous treatment decision making

To compare the diagnostic efficacy of Gd-EOB-DTPA-enhanced magnetic resonance imaging (MRI) vs. multidetector computed tomography (MDCT) for the detection and classification of focal liver lesions, differentiated also for lesion entity and size; a separate analysis of pre- and postcontrast images as well as T2-weighted MRI sequences of focal and exclusively solid lesions was integrated.Twenty-nine patients with 130 focal liver lesions underwent MDCT (64-detector-row; contrast medium iopromide; native, arterial, portalvenous, venous phase) and MRI (1.5-T; dynamic and tissue-specific phase 20 min after application of Gd-EOB-DTPA). Hepatic lesions were verified against a standard of reference (SOR). CT and MR images were independently analysed by four blinded radiologists on an ordinal 6-point-scale, determining lesion classification and diagnostic confidence.Among 130 lesions, 68 were classified as malignant and 62 as benign by SOR. The detection of malignant and benign lesions differed significantly between combined and postcontrast MRI vs. MDCT; overall detection rate was 91.5% for combined MRI and 80.4% for combined MDCT (p < 0.05). Considering all four readers together, combined MDCT achieved sensitivity of 66.2%, specificity of 79.0%, and diagnostic accuracy of 72.3%; combined MRI reached superior diagnostic efficacy: sensitivity 86.8%, specificity 94.4%, accuracy 90.4% (p < 0.05). Differentiated for lesion size, in particular lesions <20 mm revealed diagnostic benefit by MRI. Postcontrast MRI also achieved higher overall sensitivity, specificity, and accuracy compared to postcontrast MDCT for focal and exclusively solid liver lesions (p < 0.05).Combined and postcontrast Gd-EOB-DTPA-enhanced MRI provided significantly higher overall detection rate and diagnostic accuracy, including low inter-observer variability, compared to MDCT in a single centre study.

Multiphase convolutional dense network for the classification of focal liver lesions on dynamic contrast-enhanced computed tomography.

BACKGROUNDThe accurate classification of focal liver lesions (FLLs) is essential to properly guide treatment options and predict prognosis. Dynamic contrast-enhanced computed tomography (DCE-CT) is still the cornerstone in the exact classification of FLLs due to its noninvasive nature, high scanning speed, and high-density resolution. Since their recent development, convolutional neural network-based deep learning techniques has been recognized to have high potential for image recognition tasks.AIMTo develop and evaluate an automated multiphase convolutional dense network (MP-CDN) to classify FLLs on multiphase CT.METHODSA total of 517 FLLs scanned on a 320-detector CT scanner using a four-phase DCE-CT imaging protocol (including precontrast phase, arterial phase, portal venous phase, and delayed phase) from 2012 to 2017 were retrospectively enrolled. FLLs were classified into four categories: Category A, hepatocellular carcinoma (HCC); category B, liver metastases; category C, benign non-inflammatory FLLs including hemangiomas, focal nodular hyperplasias and adenomas; and category D, hepatic abscesses. Each category was split into a training set and test set in an approximate 8:2 ratio. An MP-CDN classifier with a sequential input of the four-phase CT images was developed to automatically classify FLLs. The classification performance of the model was evaluated on the test set; the accuracy and specificity were calculated from the confusion matrix, and the area under the receiver operating characteristic curve (AUC) was calculated from the SoftMax probability outputted from the last layer of the MP-CDN.RESULTSA total of 410 FLLs were used for training and 107 FLLs were used for testing. The mean classification accuracy of the test set was 81.3% (87/107). The accuracy/specificity of distinguishing each category from the others were 0.916/0.964, 0.925/0.905, 0.860/0.918, and 0.925/0.963 for HCC, metastases, benign non-inflammatory FLLs, and abscesses on the test set, respectively. The AUC (95% confidence interval) for differentiating each category from the others was 0.92 (0.837-0.992), 0.99 (0.967-1.00), 0.88 (0.795-0.955) and 0.96 (0.914-0.996) for HCC, metastases, benign non-inflammatory FLLs, and abscesses on the test set, respectively.CONCLUSIONMP-CDN accurately classified FLLs detected on four-phase CT as HCC, metastases, benign non-inflammatory FLLs and hepatic abscesses and may assist radiologists in identifying the different types of FLLs.

CasCRNN-GL-Net: cascaded convolutional and recurrent neural networks with global and local pathways for classification of focal liver lesions in multi-phase CT images

CasCRNN-GL-Net: cascaded convolutional and recurrent neural networks with global and local pathways for classification of focal liver lesions in multi-phase CT images

Automatic classification of focal liver lesions based on MRI and risk factors.

ObjectivesAccurate classification of focal liver lesions is an important part of liver disease diagnostics. In clinical practice, the lesion type is often determined from the abdominal MR examination, which includes T2-weighted and dynamic contrast enhanced (DCE) MR images. To date, only T2-weighted images are exploited for automatic classification of focal liver lesions. In this study additional MR sequences and risk factors are used for automatic classification to improve the results and to make a step forward to a clinically useful aid for radiologists.Materials and methodsClinical MRI data sets of 95 patients with in total 125 benign lesions (40 adenomas, 29 cysts and 56 hemangiomas) and 88 malignant lesions (30 hepatocellular carcinomas (HCC) and 58 metastases) were included in this study. Contrast curve, gray level histogram, and gray level co-occurrence matrix texture features were extracted from the DCE-MR and T2-weighted images. In addition, risk factors including the presence of steatosis, cirrhosis, and a known primary tumor were used as features. Fifty features with the highest ANOVA F-score were selected and fed to an extremely randomized trees classifier. The classifier evaluation was performed using the leave-one-out principle and receiver operating characteristic (ROC) curve analysis.ResultsThe overall accuracy for the classification of the five major focal liver lesion types is 0.77. The sensitivity/specificity is 0.80/0.78, 0.93/0.93, 0.84/0.82, 0.73/0.56, and 0.62/0.77 for adenoma, cyst, hemangioma, HCC, and metastasis, respectively.ConclusionThe proposed classification system using features derived from clinical DCE-MR and T2-weighted images, with additional risk factors is able to differentiate five common types of lesions and is a step forward to a clinically useful aid for focal liver lesion diagnosis.

Focal Liver Lesion Classification Based on Receiver Operating Characteristic Analysis

Background: Computer-Aided Diagnosis (CAD) on Focal Liver Lesion (FLL) has been widely researched. It aims at classifying liver images into malignant or benign, so as to help doctors to make corresponding diagnosis. In most existing CAD systems, the automatic decision strategies on challenging cases usually lead to risky diagnosis. Objective: In this paper, we adopted a ROC optimal abstention model for FLL classification to reduce the misclassification risk. Method: The workflow of ROC based FLL classification includes the stages of feature extraction, statistic for building ROC curve and ROC optimal abstaining classification. Through investigating the properties of ROC, we can automatically find two optimal thresholds for building the abstention model. A part of cases refrains from being classified to achieve the lowest misclassification cost. Results: The model classifies the FLL medical records into positive (malignant), negative (benign) and abstaining cases. The abstained challenging cases can be carefully examined by experts in order to reduce the misclassification risk. Conclusion: Abundant experiments indicate that the proposed method can achieve satisfied results and is effective for FLL diagnosis.

Tensor-based sparse representations of multi-phase medical images for classification of focal liver lesions

Medical images play an important role in clinics. In most clinic sites, the diagnosis of diseases and the comprehending of disease progression need firstly accurate interpretation of the available medical images, which would be time-consuming in manual interpretation of the accumulated large amount of medical images. Thus automatical analysis and understanding of the available medical images become an active research topic, and therein, feature extraction of medical images plays an important role for achieving diagnosis performance. Natural images hold two-dimensional structures and linear statistical methods, such as k-means, GMM, and sparse coding, are widely applied for extracting compact and inherent representations. In contrast, medical images themselves have three-dimensional structures and possibly consist of multi-phase extension. Directly applying linear methods on the available medical data would lead to high dimensional vectors by reshaping the multi-dimensional data and would destroy the correlated relation among different dimensions of the raw medical data domain. Therefore, this study proposes a multilinear extension of the linear sparse coding for extracting compact and effective intermediate representations of the multi-dimensional local structures in multi-phase CT images, and aggregating the intermediate representations in the Bag-of-Visual-Words (BoVW) manner for classification of focal liver lesions (FLLs). In the proposed approach, three-layer volumes from the corresponding slices of multi-phase CT images are formed and spatiotemporal local structures from the volumes are extracted as 3rd-order tensors. Regarding the high dimensional local structures as tensors, we propose a K-CP (CANDECOMP/PARAFAC) algorithm to learn a tensor dictionary in an iterative way and extract the sparse representation with a multilinear OMP method. The aggregation of the sparse representation is implemented in the BoVW manner, which has been proved to be an effective method for extracting features from natural images and medical images. The proposed strategy is evaluated in classification of focal liver lesions and achieved better results than conventional BoVW with linear statistical methods.

Clinical Applications of Contrast-Enhanced Ultrasound in the Pediatric Work-Up of Focal Liver Lesions and Blunt Abdominal Trauma: A Systematic Review.

In pediatrics ultrasound has long been viewed more favorably than imaging that exposes patients to radiation and iodinated contrast or requires sedation. It is child-friendly and diagnostic capabilities have been improved with the advent of contrast-enhanced ultrasound (CEUS). The application of CEUS is indeed promising. However, no ultrasound contrast agent manufactured today is registered for pediatric use in Europe. The contrast agent SonoVue® has recently been approved by the FDA under the name of Lumason® to be used in hepatic investigations in adults and children. This article reviews the literature with respect to 2 specific applications of CEUS in children: 1) identification of parenchymal injuries following blunt abdominal trauma, and 2) classification of focal liver lesions. Applications were chosen through the CEUS guidelines published by the European Federation of Societies for Ultrasound in Medicine and Biology and World Federation for Ultrasound in Medicine and Biology. Literature was obtained by searching Medline and Pubmed Central (using Pubmed), Scopus database and Embase. CEUS proved to be an effective investigation in the hemodynamically stable child for identifying parenchymal injuries and for the characterization of focal liver lesions. CEUS showed comparable performance to CT and MRI with a specificity of 98% for identifying benign lesions and a negative predictive value of 100%. For the applications reviewed here, CEUS holds promising perspectives and can help reduce radiation exposure and use of iodinated contrast agents in pediatrics, thereby potentially reducing complications in routine imaging.

Endoscopic ultrasonography - emerging applications in hepatology.

The inspection of the liver is a valuable part of the upper endoscopic ultrasonography (EUS) studies, regardless of the primary indication for the examination. The detailed images of the liver segments provided by EUS allows the use of this technique in the study of parenchymal liver disease and even in the diagnosis and classification of focal liver lesions. EUS has also emerged as an important tool in understanding the complex collateral circulation in patients with portal hypertension and their clinical and prognostic value. Recently, EUS-guided portal vein catheterization has been performed for direct portal pressure measurement as an alternative method to evaluate portal hemodynamics. In this review, the authors summarize the available evidence regarding the application of EUS to patients with liver diseases and how we can apply it in our current clinical practice.

Impact of Modified Anisotropic Diffusion-based Enhancement Method in Computer-Aided Classification of Focal Liver Lesions.

This work is presented with the objective to assess quantitatively the impact of modified anisotropic diffusion-based enhancement method of Mittal et al. in computer-aided classification of focal liver lesions. This assessment was made before and after enhancement of clinically acquired ultrasound images with the comparison of (a) discrimination capability of radiologically important texture contrast feature using box plot and p-value statistics and (b) test results of designed computer-aided classification schemes to detect/classify focal liver tissues using receiver operating characteristic curves. The results reveal that the application of enhancement method on clinically acquired ultrasound image may effectively improve the confidence of clinicians/radiologists in computer-aided diagnostic solutions to detect and classify focal liver lesions.

Diffusion Weighted Magnetic Resonance Imaging for the Classification of Focal Liver Lesions as Benign or Malignant.

To assess the role of diffusion weighted imaging sequence (DWI), routinely used in hepatic magnetic resonance imaging (MRI) for the differentiation of focal liver lesions (FLLs) as benign or malignant. 99 FLLs assessed by liver MRI in 80 patients were included in the present study. All lesions were retrospectively analyzed by two experienced radiologists, independent from each other, who were not aware of the previous results obtained by using different imaging techniques. All included FLLs had a final histological diagnosis or a final diagnosis based on consensus reading by two experienced radiologists and follow-up at 6 months. The FLLs signal was qualitatively appreciated on the b-800 sequences and on the apparent diffusion coefficient (ADC) map. The ADC value of each FLL was measured and the ADC ratio between the ADC value of the assessed FLL and that of the surrounding liver parenchyma were calculated. The mean ADC value for benign FLLs as assessed by the two independent readers was 1.78 x 10¯³ and 1.72 x 10¯³, respectively. The mean ADC value for malignant FLLs was 0.92 x 10¯³ for the first reader and 0.95 x 10¯³ for the second reader. The mean ADC ratio for benign FLLs was 1.91 and 1.85 for the two readers and for malignant FLLs was 0.91 and 0.94, respectively. Using an ADC value lower than 1.024 x 10¯³ offers a specificity of 100% and a sensitivity of 62.5% for the diagnosis of malignant FLLs. The ADC value is an indicator which is less prone to interobserver variability (correlation of 0.919→1). The ADC ratio has, as the analysis of the ROC curve shows, the best predictive value for differentiation between benign and malignant FLLs. Analysis of the signal intensity on the DWI b-800 image alone is of no significance in differentiating benign from malignant FLLs (p>0.05). The ADC value and the ADC ratio assessed on liver DWI are useful diagnostic tools in the differential diagnosis of benign vs. malignant FLLs. Quantitative methods such as calculating the ADC value or ADC ratio have better diagnostic value than the qualitative techniques.

Classification of focal liver lesions on ultrasound images by extracting hybrid textural features and using an artificial neural network.

This paper focuses on the improvement of the diagnostic accuracy of focal liver lesions by quantifying the key features of cysts, hemangiomas, and malignant lesions on ultrasound images. The focal liver lesions were divided into 29 cysts, 37 hemangiomas, and 33 malignancies. A total of 42 hybrid textural features that composed of 5 first order statistics, 18 gray level co-occurrence matrices, 18 Law's, and echogenicity were extracted. A total of 29 key features that were selected by principal component analysis were used as a set of inputs for a feed-forward neural network. For each lesion, the performance of the diagnosis was evaluated by using the positive predictive value, negative predictive value, sensitivity, specificity, and accuracy. The results of the experiment indicate that the proposed method exhibits great performance, a high diagnosis accuracy of over 96% among all focal liver lesion groups (cyst vs. hemangioma, cyst vs. malignant, and hemangioma vs. malignant) on ultrasound images. The accuracy was slightly increased when echogenicity was included in the optimal feature set. These results indicate that it is possible for the proposed method to be applied clinically.

Development of a Support Vector Machine - Based Image Analysis System for Focal Liver Lesions Classification in Magnetic Resonance Images

Purpose: The design and implementation of a computer-based image analysis system employing the support vector machine (SVM) classifier system for the classification of Focal Liver Lesions (FLLs) on routine non-enhanced, T2-weighted Magnetic Resonance (MR) images. Materials and Methods: The study comprised 92 patients; each one of them has undergone MRI performed on a Magnetom Concerto (Siemens). Typical signs on dynamic contrast-enhanced MRI and biopsies were employed towards a three class categorization of the 92 cases: 40-benign FLLs, 25-Hepatocellular Carcinomas (HCC) within Cirrhotic liver parenchyma and 27-liver metastases from Non-Cirrhotic liver. Prior to FLLs classification an automated lesion segmentation algorithm based on Marcov Random Fields was employed in order to acquire each FLL Region of Interest. 42 texture features derived from the gray-level histogram, co-occurrence and run-length matrices and 12 morphological features were obtained from each lesion. Stepwise multi-linear regression analysis was utilized to avoid feature redundancy leading to a feature subset that fed the multiclass SVM classifier designed for lesion classification. SVM System evaluation was performed by means of leave-one-out method and ROC analysis. Results: Maximum accuracy for all three classes (90.0%) was obtained by means of the Radial Basis Kernel Function and three textural features (Inverse- Different-Moment, Sum-Variance and Long-Run-Emphasis) that describe lesion's contrast, variability and shape complexity. Sensitivity values for the three classes were 92.5%, 81.5% and 96.2% respectively, whereas specificity values were 94.2%, 95.3% and 95.5%. The AUC value achieved for the selected subset was 0.89 with 0.81 - 0.94 confidence interval. Conclusion: The proposed SVM system exhibit promising results that could be utilized as a second opinion tool to the radiologist in order to decrease the time/cost of diagnosis and the need for patients to undergo invasive examination.

Sa1000 Focal Liver Lesions Classification by Artificial Neural Networks and Support Vector Machines Employing Dynamic Imaging Data

RHH by pooling all available evidence in a systematic review and meta-analysis. Methods We conducted a comprehensive literature search using Ovid on MEDLINE (1946-2013), EMBASE (1988-2013), PubMed (1946-2013), Cochrane Library, and the Web of Science to identify studies on TIPS for RHH. Two reviewers independently screened article titles and abstracts for eligibility. Among information abstracted from each study was number of participants, outcomes of interest (mortality/survival, symptom response, HE and change in hepatic venous pressure gradient (HVPG)). Complete response was defined by resolution of HH without further need for thoracentesis, while a partial response was defined as improvement in symptoms and/or a decrease need for thoracentesis. New onset or worsening of baseline HE within 30 days of TIPS was considered a complication of the procedure. Data were pooled using a fixed-effects meta-analysis. Outcomes are expressed as proportions. Heterogeneity was assessed using the I2 and Cochrane Q test. Results Seven studies, involving 203 patients were included in the analyses. The mean age of the patients was 56.3 years, with 50.7% being male. The majority of patients were Child-Pugh Class C (56.9%), while 40.7% and 0.8% were Child-Pugh Class B and A, respectively. The mean follow up period was 11.9 ± 3.6 months. The mean preand post-TIPS HVPG was 20.14 mmHg (range: 17.4-26) and 7.37 mmHg (range: 5.7-10), respectively. Response to TIPS was complete in 55.8% (95% CI 44.7-66.9) and partial in 17.6% (95% CI 10.9-24.2). There was no response in 21.2% (95% CI 14.2-28.3) of patients. The incidence of post-TIPS HE was 11.7% (95% CI 6.3-17.2). The mortality within 30days of TIPS placement was 16.6% (95% CI 10.222.9), while the overall mortality was 49.1% (95% CI 38.8-59.5). The mean 1-year survival (reported by 2 studies) was 56%. Predictors of mortality included older age, severity of liver disease and non-response to TIPS. Conclusion TIPS is associated with symptomatic relief in approximately 75% of patients with RHH. Although TIPS increases the incidence of HE, response to treatment may be associated with improved survival. TIPS should be used cautiously in older patients with more severe underlying liver disease.

Deep learning based classification of focal liver lesions with contrast-enhanced ultrasound

Classification of liver masses is important to early diagnosis of patients. In this paper, a diagnostic system of liver disease classification based on contrast enhanced ultrasound (CEUS) imaging is proposed. In the proposed system, the dynamic CEUS videos of hepatic perfusion are firstly retrieved. Secondly, time intensity curves (TICs) are extracted from the dynamic CEUS videos using sparse non-negative matrix factorizations. Finally, deep learning is employed to classify benign and malignant focal liver lesions based on these TICs. Quantitative comparisons demonstrate that the proposed method outperforms the compared classification methods in accuracy, sensitivity and specificity.