Predicting the grade of hepatocellular carcinoma based on non-contrast-enhanced MRI radiomics signature

Wu, Minghui; Tan, Hongna; Gao, Fei; Hai, Jinjin; Ning, Peigang; Chen, Jian; Zhu, Shuchai; Wang, Meiyun; Dou, S. X.; Shi, Dapeng

doi:10.1007/s00330-018-5787-2

Cited by 151 publications

(113 citation statements)

References 31 publications

Supporting

Mentioning

109

Contrasting

Unclassified

Order By: Relevance

“…Recently, Wu et al applied the traditional machine learning with radiomics features for differentiation of high‐ and low‐grade HCCs and reported that the radiomics signatures based on T1 weighted MR and T2 weighted MR yielded AUC values of 72.48% and 73.6% for discriminating high‐grade and low‐grade HCCs, respectively . The performance of the conventional radiomics feature has also been assessed and compared with deep features with respect to multiple b‐value images, the log maps and the ADC map for HCC grading.…”

Section: Discussionmentioning

confidence: 99%

“…Recently, Wu et al applied the traditional machine learning with radiomics features for differentiation of high-and low-grade HCCs and reported that the radiomics signatures based on T1 weighted MR and T2 weighted MR yielded AUC values of 72.48% and 73.6% for discriminating highgrade and low-grade HCCs, respectively. 47 The performance of the conventional radiomics feature has also been assessed and compared with deep features with respect to multiple bvalue images, the log maps and the ADC map for HCC grading. Generally, it is difficult to conduct a fair comparison between deep features and the radiomics feature for HCC grading as the approaches of feature extraction and the mechanism of classifiers are remarkably different between the deep learning approach and the radiomics approach.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Grading of hepatocellular carcinoma based on diffusion weighted images with multiple b‐values using convolutional neural networks

et al. 2019

View full text Add to dashboard Cite

Purpose To effectively grade hepatocellular carcinoma (HCC) based on deep features derived from diffusion weighted images (DWI) with multiple b‐values using convolutional neural networks (CNN). Materials and Methods Ninety‐eight subjects with 100 pathologically confirmed HCC lesions from July 2012 to October 2018 were included in this retrospective study, including 47 low‐grade and 53 high‐grade HCCs. DWI was performed for each subject with a 3.0T MR scanner in a breath‐hold routine with three b‐values (0,100, and 600 s/mm2). First, logarithmic transformation was performed on original DWI images to generate log maps (logb0, logb100, and logb600). Then, a resampling method was performed to extract multiple 2D axial planes of HCCs from the log map to increase the dataset for training. Subsequently, 2D CNN was used to extract deep features of the log map for HCCs. Finally, fusion of deep features derived from three b‐value log maps was conducted for HCC malignancy classification. Specifically, a deeply supervised loss function was devised to further improve the performance of lesion characterization. The data set was split into two parts: the training and validation set (60 HCCs) and the fixed test set (40 HCCs). Four‐fold cross validation with 10 repetitions was performed to assess the performance of deep features extracted from single b‐value images for HCC grading using the training and validation set. Receiver operating characteristic curve (ROC) and area under the curve (AUC) values were used to assess the characterization performance of the proposed deep feature fusion method to differentiate low‐grade and high‐grade in the fixed test set. Results The proposed fusion of deep features derived from logb0, logb100, and logb600 with deeply supervised loss function generated the highest accuracy for HCC grading (80%), thus outperforming the method of deep feature derived from the ADC map directly (72.5%), the original b0 (65%), b100 (68%), and b600 (70%) images. Furthermore, AUC values of the deep features of the ADC map, the deep feature fusion with concatenation, and the proposed deep feature fusion with deeply supervised loss function were 0.73, 0.78, and 0.83, respectively. Conclusion The proposed fusion of deep features derived from the logarithm of the three b‐value images yields high performance for HCC grading, thus providing a promising approach for the assessment of DWI in lesion characterization.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Grading of hepatocellular carcinoma based on diffusion weighted images with multiple b‐values using convolutional neural networks

et al. 2019

View full text Add to dashboard Cite

show abstract

“…The diagnostic performance of non‐DWI‐related techniques differed among different studies. Wu et al found that a radiomics analysis based on MRI had the potential to discriminate HCCs with a high ES grade with an AUROC of 0.742 to 0.8. Chen et al found that the MRI perfusion parameter K trans had the best diagnostic sensitivity and specificity for distinguishing HCCs with a high grade (93.3% and 63.2%, respectively), with an AUROC of 0.78.…”

Section: Discussionmentioning

confidence: 99%

Diagnostic accuracy of quantitative diffusion parameters in the pathological grading of hepatocellular carcinoma: A meta‐analysis

She

Wang

Yang

et al. 2019

Magnetic Resonance Imaging

View full text Add to dashboard Cite

Background: Accurate preoperative assessment of the pathological grade of hepatocellular carcinoma (HCC) could greatly benefit prognostic predictions. Purpose: To assess and compare the diagnostic accuracy of the apparent diffusion coefficient (ADC) and tissue diffusivity (D) for the noninvasive pathological grading of HCC. Study Type: Meta-analysis. Data Sources: PubMed/Medline, EMBASE, the Web of Science, and the Cochrane Library were searched to find related original articles published up to May 30, 2019. Field Strength/Sequence: Diffusion-weighted imaging (DWI) and/or intravoxel incoherent motion (IVIM) were performed with a 1.5T or 3.0T scanner. Assessment: The Quality Assessment of Diagnostic Accuracy Studies 2 tool was used to assess the methodologic quality. Statistical Tests: The bivariate random-effects model was used to obtain the pooled sensitivity and specificity, and the area under summary receiver operating characteristic curve (AUROC) was obtained. Subgroup analyses were performed. Results: A total of 16 original articles (1428 HCCs) were included. Most studies had a low to unclear risk of bias and minimal concerns regarding applicability. For the discrimination of well-differentiated HCCs, the pooled sensitivity and specificity of the ADC value were 85% and 92%, respectively. For the discrimination of poorly differentiated HCCs, the pooled sensitivity and specificity of the ADC value and D were 84% and 80%, and 92% and 77%, respectively. The summary AUROC of D (0.94) was significantly higher than that of ADC (0.89) (z = -2.718, P = 0.007). The subgroup analyses identified three covariates including size, number of included lesions in the studies, and blindness to the reference standard as possible sources of heterogeneity. Data Conclusion: This meta-analysis showed that the ADC and D values had a high to excellent accuracy for the noninvasive pathological grading of HCCs and that the D value was superior to the ADC value for discriminating poorly differentiated HCCs. Level of Evidence: 3 Technical Efficacy Stage: 2 J. MAGN. RESON. IMAGING 2020;51:1581-1593 H EPATOCELLULAR CARCINOMA (HCC) is the most common type of primary malignant liver neoplasm and the second cause of malignancy-related death worldwide. 1Although treatments have progressed in recent years, the prognosis of HCC patients remains poor because of the high recurrence rate. 2 Studies have confirmed that the pathological View this article online at wileyonlinelibrary.com.

show abstract

“…This research approach utilises high-throughput extraction of feature data from radiographic images, 12 and can potentially develop models to predict lesion phenotypes and prognosis in a non-invasive manner. 13,14 To our knowledge, there are relatively limited radiomics analysis data about prognosis estimation in HCC, with most of the radiomics models established on the basis of computed tomography (CT) images, [15][16][17][18][19] and only a few studies investigating the role of magnetic resonance imaging (MRI), [20][21][22][23] especially contrast-enhanced MRI. Moreover, previous MRI-based radiomics analyses of patients with HCC were based on a few MRI sequences from small-sample studies, and no MRI-based radiomics model for long-term survival prediction in HCC is currently available.…”

Section: Introductionmentioning

confidence: 99%

MRI-based radiomics model for preoperative prediction of 5-year survival in patients with hepatocellular carcinoma

et al. 2020

View full text Add to dashboard Cite

BACKGROUND: Recurrence is the major cause of mortality in patients with resected HCC. However, without a standard approach to evaluate prognosis, it is difficult to select candidates for additional therapy. METHODS: A total of 201 patients with HCC who were followed up for at least 5 years after curative hepatectomy were enrolled in this retrospective, multicentre study. A total of 3144 radiomics features were extracted from preoperative MRI. The random forest method was used for radiomics signature building, and five-fold cross-validation was applied. A radiomics model incorporating the radiomics signature and clinical risk factors was developed. RESULTS: Patients were divided into survivor (n = 97) and non-survivor (n = 104) groups based on the 5-year survival after surgery. The 30 most survival-related radiomics features were selected for the radiomics signature. Preoperative AFP and AST were integrated into the model as independent clinical risk factors. The model demonstrated good calibration and satisfactory discrimination, with a mean AUC of 0.9804 and 0.7578 in the training and validation sets, respectively. CONCLUSIONS: This radiomics model is a valid method to predict 5-year survival in patients with HCC and may be used to identify patients for clinical trials of perioperative therapies and for additional surveillance.

show abstract

Predicting the grade of hepatocellular carcinoma based on non-contrast-enhanced MRI radiomics signature

Cited by 151 publications

References 31 publications

Grading of hepatocellular carcinoma based on diffusion weighted images with multiple b‐values using convolutional neural networks

Grading of hepatocellular carcinoma based on diffusion weighted images with multiple b‐values using convolutional neural networks

Diagnostic accuracy of quantitative diffusion parameters in the pathological grading of hepatocellular carcinoma: A meta‐analysis

MRI-based radiomics model for preoperative prediction of 5-year survival in patients with hepatocellular carcinoma

Contact Info

Product

Resources

About