Dataset Information

Prediction of post transarterial chemoembolization MR images of hepatocellular carcinoma using spatio-temporal graph convolutional networks.

ABSTRACT: Magnetic resonance imaging (MRI) plays a critical role in the planning and monitoring of hepatocellular carcinomas (HCC) treated with locoregional therapies, in order to assess disease progression or recurrence. Dynamic contrast-enhanced (DCE)-MRI sequences offer temporal data on tumor enhancement characteristics which has strong prognostic value. Yet, predicting follow-up DCE-MR images from which tumor enhancement and viability can be measured, before treatment of HCC actually begins, remains an unsolved problem given the complexity of spatial and temporal information. We propose an approach to predict future DCE-MRI examinations following transarterial chemoembolization (TACE) by learning the spatio-temporal features related to HCC response from pre-TACE images. A novel Spatial-Temporal Discriminant Graph Neural Network (STDGNN) based on graph convolutional networks is presented. First, embeddings of viable, equivocal and non-viable HCCs are separated within a joint low-dimensional latent space, which is created using a discriminant neural network representing tumor-specific features. Spatial tumoral features from independent MRI volumes are then extracted with a structural branch, while dynamic features are extracted from the multi-phase sequence with a separate temporal branch. The model extracts spatio-temporal features by a joint minimization of the network branches. At testing, a pre-TACE diagnostic DCE-MRI is embedded on the discriminant spatio-temporal latent space, which is then translated to the follow-up domain space, thus allowing to predict the post-TACE DCE-MRI describing HCC treatment response. A dataset of 366 HCC's from liver cancer patients was used to train and test the model using DCE-MRI examinations with associated pathological outcomes, with the spatio-temporal framework yielding 93.5% classification accuracy in response identification, and generating follow-up images yielding insignificant differences in perfusion parameters compared to ground-truth post-TACE examinations.

SUBMITTER: Svecic A

PROVIDER: S-EPMC8651128 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:PurposeTo prospectively assess serial changes in contrast material-enhanced and diffusion-weighted (DW) magnetic resonance (MR) imaging values within 1 month after transarterial chemoembolization (TACE) in patients with unresectable hepatocellular carcinoma (HCC).Materials and methodsInstitutional review board approval was obtained for this prospective HIPAA-compliant study. MR imaging was performed before and within 24 hours after TACE in 24 patients with HCC (21 male, three female; mean age, 59 years and 62 years, respectively). Serial MR imaging was subsequently performed 1, 2, 3, and 4 weeks after therapy. The imaging protocol included fast spin-echo T2-weighted MR imaging, breath-hold DW echo-planar MR imaging, and breath-hold unenhanced and contrast-enhanced T1-weighted three-dimensional fat-suppressed gradient-recalled-echo MR imaging in the arterial and portal venous phases. Tumor size, enhancement, and apparent diffusion coefficient (ADC) values were recorded before and sequentially after treatment. Regression models for the correlated data were used to assess changes in these parameters over time after TACE.ResultsMean tumor size was 7.5 cm and was unchanged up to 4 weeks after therapy. Reduction in tumor enhancement in the arterial phase occurred immediately after TACE, with a consistent reduction occurring 1-3 weeks after therapy (P = .001). Reduction in tumor enhancement in the portal venous phase also occurred immediately after TACE, with a consistent reduction occurring 1-3 weeks after therapy (P = .0003). The increase in tumor ADC value was significant 1-2 weeks after therapy (P = .004), borderline significant 3 weeks after therapy, and insignificant 24 hours and 4 weeks after therapy.ConclusionSignificant reduction in tumor enhancement occurred within 24 hours after TACE and persisted up to 4 weeks after TACE. Lesser changes in the ADC value appeared 1 week after TACE, persisted through 2 weeks after TACE, and became less apparent 3 and 4 weeks after TACE. No change in tumor size was recorded during the follow-up period.

Project description:AimsThe purpose of the present study was to compare the efficacies of transarterial chemoembolization (TACE) combined with sorafenib versus TACE monotherapy for treating patients with advanced hepatocellular carcinoma (HCC).MethodsWe enrolled 321 patients and selected 280 with advanced HCC (Barcelona Clinic Liver Cancer stage C) who underwent TACE therapy between February 2009 and February 2013. TACE alone (monotherapy group) was administered to 198 patients (70.7%), and the remaining 82 (29.3%) underwent repeat combined TACE and sorafenib therapy (combined group). To minimize selection bias, these latter 82 patients were matched using propensity-score matching at a 1∶2 ratio with 164 patients who received TACE monotherapy. The primary endpoints were overall survival (OS) and related subgroup analysis. The secondary endpoints were time to progression (TTP) and treatment-related adverse events.ResultsOf the respective patients in the combined and monotherapy groups, 64.6% and 49.2% had vascular invasion, 87.8% and 91.1% had extrahepatic metastasis, and 54.3% and 47.1% had both. In the propensity-score-matched cohort, the OS survival of the combined group was significantly higher compared with the monotherapy group (7.0 months vs. 4.9 months, respectively, P = 0.003). The TTP was significantly longer in the combined group (2.6 months vs. 1.9 months, respectively, P = 0.001). Subgroup analysis showed that the outcomes of patients with advanced HCC without main portal vein invasion who were treated with combined therapy were significantly better compared with those who received monotherapy (P<0.05). Univariate and subsequent multivariate analyses revealed that the addition of sorafenib was an independent predictor of favorable OS and TTP (adjusted hazard ratios, 0.63 and 0.62, respectively; P<0.05 for both).ConclusionSorafenib plus TACE was more effective than TACE monotherapy for treating patients with advanced HCC without main portal vein invasion. Future trials with larger samples are required to validate these preliminary findings.

Project description:BACKGROUND:Optimal administration of transarterial chemoembolization (TACE), the standard approach for intermediate-stage hepatocellular carcinoma (HCC), requires clinical and technical expertise. We sought to evaluate whether TACE retains its effectiveness when administered across a broad range of health care settings. Furthermore, as the use of yttrium(90) (Y(90)) radioembolization has been increasing, we explored the comparative effectiveness of Y(90) as an alternative to TACE. METHODS:Patients with HCC diagnosed from 2004 through 2009 treated initially with TACE or Y(90) were identified from the SEER-Medicare linkage. Key covariates included prediagnosis ?-fetoprotein (AFP) screening, complications of cirrhosis, and tumor extent. Effect of treatment, patient, and health care system factors on overall survival (OS) was evaluated using multivariable Cox proportional hazards. Stratified OS estimates are provided. Propensity score (PS) weighting was used to compare effectiveness of Y(90) with TACE. RESULTS:Of 1528 patients who underwent intra-arterial embolization, 577 received concurrent chemotherapy (eg, TACE). Median OS was 21 months (95% CI, 18-23) following TACE and 9 months (95% CI, 1-41) following Y(90). Refined survival estimates stratified by stage, AFP screening, and liver comorbidity are presented. The 90-day mortality rate after TACE was 21% to 25% in patients with extrahepatic spread or vascular invasion. In the PS-weighted analysis, Y(90) was associated with inferior survival, with an adjusted hazard ratio of 1.39 (95% CI, 1.02-1.90). CONCLUSIONS:The effectiveness of TACE is generalizable to Medicare patients receiving care in a variety of treatment settings. However, early posttreatment mortality is high in patients with advanced disease. We found no evidence of improved outcomes with Y(90) compared with TACE. Survival estimates from this large cohort can be used to provide prognostic information to patients considering palliative TACE.

Dataset Information

Prediction of post transarterial chemoembolization MR images of hepatocellular carcinoma using spatio-temporal graph convolutional networks.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure