This study aimed to explore the added value of viscoelasticity measured by magnetic resonance elastography (MRE) in the prediction of Ki-67 expression in hepatocellular carcinoma (HCC) using a deep learning combined radiomics (DLCR) model. This retrospective study included 108 histopathology-proven HCC patients (93 males; age, 59.6 ± 11.0 years) who underwent preoperative MRI and MR elastography. They were divided into training (n = 87; 61.0 ± 9.8 years) and testing (n = 21; 60.6 ± 10.1 years) cohorts. An independent validation cohort including 43 patients (60.1 ± 11.3 years) was included for testing. A DLCR model was proposed to predict the expression of Ki-67 with cMRI, including T2W, DW, and dynamic contrast enhancement (DCE) images as inputs. The images of the shear wave speed (c-map) and phase angle (φ-map) derived from MRE were also fed into the DLCR model. The Ki-67 expression was classified into low and high groups with a threshold of 20%. Both c and φ values were ranked within the top six features for Ki-67 prediction with random forest selection, which revealed the value of MRE-based viscosity for the assessment of tumor proliferation status in HCC. When comparing the six CNN models, Xception showed the best performance for classifying the Ki-67 expression, with an AUC of 0.80 ± 0.03 (CI: 0.79–0.81) and accuracy of 0.77 ± 0.04 (CI: 0.76–0.78) when cMRI were fed into the model. The model with all modalities (MRE, AFP, and cMRI) as inputs achieved the highest AUC of 0.90 ± 0.03 (CI: 0.89–0.91) in the validation cohort. The same finding was observed in the independent testing cohort, with an AUC of 0.83 ± 0.03 (CI: 0.82–0.84). The shear wave speed and phase angle improved the performance of the DLCR model significantly for Ki-67 prediction, suggesting that MRE-based c and φ-maps can serve as important parameters to assess the tumor proliferation status in HCC.
Background Estimating liver function reserve is essential for preoperative surgical planning and predicting post-hepatectomy complications in patients with hepatocellular carcinoma (HCC). We investigated hepatic viscoelasticity quantified by tomoelastography, a multifrequency magnetic resonance elastography technique, to predict liver function reserve. Methods One hundred fifty-six patients with suspected HCC (mean age, 60 ± 1 years; 131 men) underwent preoperative tomoelastography examination between July 2020 and August 2021. Sixty-nine were included in the final analysis, and their 15-min indocyanine green retention rates (ICG-R15s) were obtained to determine liver function reserve. Tomoelastography quantified the shear wave speed (c, m/s), which represents stiffness, and loss angle (φ, rad), which represents fluidity. Both were correlated with the ICG-R15. A prediction model based on logistic regression for major hepatectomy tolerance (ICG-R15 ≥ 14%) was established. Results Patients were assigned to either the ICG-R15 < 14% (n = 50) or ICG-R15 ≥ 14% (n = 19) group. Liver c (r = 0.617) and φ (r = 0.517) were positively correlated with the ICG-R15 (both p < 0.001). At fibrosis stages F1–2, φ was positively correlated with the ICG-R15 (r = 0.528; p = 0.017), but c was not (p = 0.104). At stages F3–4, c (r = 0.642; p < 0.001) and φ (r = 0.377; p = 0.008) were both positively correlated with the ICG-R15. The optimal cutoffs of c and φ for predicting ICG-R15 ≥ 14% were 2.04 m/s and 0.79 rad, respectively. The area under the receiver operating characteristic curve was higher for c (0.892) than for φ (0.779; p = 0.045). Conclusions Liver stiffness and fluidity, quantified by tomoelastography, were correlated with liver function and may be used clinically to noninvasively assess liver function reserve and stratify treatments.
BackgroundGlypican-3 (GPC3) expression is investigated as a promising target for tumor-specific immunotherapy of hepatocellular carcinoma (HCC). This study aims to determine whether GPC3 alters the viscoelastic properties of HCC and whether tomoelastography, a multifrequency magnetic resonance elastography (MRE) technique, is sensitive to it.MethodsNinety-five participants (mean age, 58 ± 1 years; 78 men and 17 women) with 100 pathologically confirmed HCC lesions were enrolled in this prospective study from July 2020 to August 2021. All patients underwent preoperative multiparametric MRI and tomoelastography. Tomoelastography provided shear wave speed (c, m/s) representing tissue stiffness and loss angle (φ, rad) relating to viscosity. Clinical, laboratory, and imaging parameters were compared between GPC3-positive and -negative groups. Univariable and multivariable logistic regression were performed to determine factors associated with GPC3-positive HCC. The diagnostic performance of combined biomarkers was established using logistic regression analysis. Area-under-the-curve (AUC) analysis was done to assess diagnostic performance in detecting GPC3-positive HCC.FindingsGPC3-positive HCCs (n=72) had reduced stiffness compared with GPC3-negative HCCs (n=23) while viscosity was not different (c: 2.34 ± 0.62 versus 2.72 ± 0.62 m/s, P=0.010, φ: 1.11 ± 0.21 vs 1.18 ± 0.27 rad, P=0.21). Logistic regression showed c and elevated serum alpha-fetoprotein (AFP) level above 20 ng/mL were independent factors for GPC3-positive HCC. Stiffness with a cutoff of c = 2.8 m/s in conjunction with an elevated AFP yielded a sensitivity of 80.3%, specificity of 70.8%, and AUC of 0.80.InterpretationReduced stiffness quantified by tomoelastography may be a mechanical signature of GPC3-positive HCC. Combining reduced tumor stiffness and elevated AFP level may provide potentially valuable biomarker for GPC3-targeted immunotherapy.
Glypican-3 (GPC3) expression in hepatocellular carcinoma (HCC) is often associated with a poor prognosis. GPC3 is a promising target for tumor-specific immunotherapy in HCC. We investigated the diagnostic performance of viscoelastic properties quantified by tomoelastography, a multifrequency MR-elastography (MRE) technique, for the detection of GPC3-positive HCC. Preliminary results showed that reduced stiffness quantified by tomoelastography is a mechanical signature of positive GPC3 expression in HCC. Combining stiffness and serum alpha-fetoprotein (AFP) level could be considered as a viable biomarker for detecting GPC3-positive HCC as well as for predicting the outcome of GPC3-targeted immunotherapy.
Hepatic venous pressure gradient (HVPG) measurement is the standard technique of assessing portal pressure. Portal hypertension is mainly caused by increased intrahepatic resistance to portal blood flow related to structural changes, including fibrosis and vascular distortion. Here, we used dynamic contrast-enhancement (DCE) MRI to quantify segmented liver contours and vascular morphology in evaluating portal pressure. Our study showed the MRI measurements correlated well with HVPG measurement. A non-invasive prediction model combining liver vascular morphology and liver surface nodularity (LSN) can potentially be used to assess clinically significant portal hypertension (CSPH).
In this study, a combined deep learning and radiomics (DLR) approach using six different network architectures was tested and compared for the prediction of high Ki-67 expressions in patients with hepatocellular carcinoma (HCC). The model was based primarily on data from MRI and tomoelastography, a multifrequency MR elastography technique. Xception delivered the best performance and recognized seven prominent features among which four were obtained from tomoelastography. Our findings demonstrated that biomechanical properties, especially viscosity and the fluid behavior of the tumor, are crucial imaging features that are important for imaging-based cancer diagnostics.
Tumor-liver biomechanical interaction investigated by multifrequency MR elastography in patients with HCC
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
334 Leonard St
Brooklyn, NY 11211
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.