Academic Radiology最新文献

Rationale and objectives: Postoperative complications (15-76%) substantially affect patients with peripheral nerve sheath tumors (PNSTs), yet objective preoperative risk tools are lacking. Magnetic resonance imaging (MRI) radiomics has focused mainly on intratumoral regions, while the predictive value of the peritumoral microenvironment remains unclear.

Materials and methods: In this retrospective single-center study (Dec 2015-Jan 2024), 280 pathologically confirmed PNST patients with preoperative MRI were randomly split 8:2 into training (n=224) and test (n=56) cohorts. Intratumoral (Intra) and Peritumoral (Per; 2 mm expansion) regions were manually segmented and 1197 radiomic features extracted. Four models were constructed: Intra-model, Per-model, a fused Intra-model+Per-model region model (Imagefusion) and a concatenated Intra-model+Per-model feature model (intraPeri2mm). After reliability filtering, t-tests and least absolute shrinkage and selection operator selection, models were trained with machine-learning classifiers. Clinical predictors were assessed, and model performance evaluated using area under the receiver operating characteristic curve (AUC) and decision-curve analysis (DCA).

Results: Diabetes was the only independent clinical predictor, and the clinical model achieved a test AUC of 0.599. In the test cohort, both fusion models outperformed single-region models (intraPeri2mm AUC 0.899; Imagefusion AUC 0.895), with consistently greater net benefit on DCA. Incorporating diabetes yielded a small, nonsignificant gain for Imagefusion (Combined_1 AUC 0.917) and no further improvement for intraPeri2mm (Combined_2 AUC 0.889). All radiomics models significantly exceeded the clinical-only model (all p<0.001).

Conclusion: Integrating intra- and peritumoral radiomics enables effective preoperative prediction of PNST postoperative complications. An Imagefusion +clinical pathway offers robust clinical net benefit when clinical data are standardized, whereas an intraPeri2mm -only strategy may be preferable where clinical data are limited.

Rationale and objectives: To preliminarily develop and validate an integrated model based on clinical features and dual-layer detector spectral CT (DLCT) 3D volumetric parameters for the noninvasive prediction of a novel Grade of Malignancy (GOM) in pancreatic ductal adenocarcinoma (PDAC)-a composite index integrating histopathological differentiation and Ki-67 index.

Materials and methods: This retrospective study enrolled 183 patients with pathologically confirmed PDAC. Patients were randomly allocated into training (n=128) and validation (n=55) cohorts. From the portal venous phase scans, three quantitative 3D volumetric parameters were extracted from the tumor volume: iodine concentration (IC), the slope of the spectral attenuation curve, and effective atomic number. Independent predictors for the GOM were identified through univariate and multivariate logistic regression analysis. The discriminatory performance of the developed models was evaluated using receiver operating characteristic curve analysis, and clinical utility was assessed with decision curve analysis.

Results: The integrated model, which combined the DLCT parameter (3D volume of interest-IC) and CA125, demonstrated superior predictive performance compared to models using clinical or DLCT features alone. In the training cohort, the integrated model achieved an area under the curve (AUC) of 0.821 (95% CI: 0.743-0.899), which was robustly validated with an AUC of 0.806 (95% CI: 0.684-0.928) in the validation cohort. Decision curve analysis confirmed that this combined model provided the highest clinical net benefit across a wide range of threshold probabilities.

Conclusion: Our findings suggest that an integrated model incorporating the 3D volumetric parameter IC from DLCT and CA125 could be a useful and noninvasive adjunct for preoperative prediction of the comprehensive GOM in PDAC, though further validation is needed.

Rationale and objectives: The functional status of CD8⁺ T cells is a key factor influencing the prognosis in patients with non-small cell lung cancer (NSCLC). We aimed to develop a radiomics model predicting the functional state of tumor-infiltrating CD8⁺ T cells in NSCLC, explore semantic characteristics linking radiomic features to CD8⁺ T cell exhaustion, and establish a prognostic nomogram.

Materials and methods: A retrospective cohort of 256 patients with NSCLC undergoing radical resection with CD8⁺ T cell functional status determined by multiplex immunofluorescence staining was randomly divided 7:3 into training and validation sets. Radiomic features from preoperative contrast-enhanced CT scans were used to develop predictive models for high density of tumor center pre-dysfunctional CD8⁺ T cells (high-T_pre) and high density of invasive margin dysfunctional CD8⁺ T cells (high-T_dys) through least absolute shrinkage and selection operator, followed by semantic analysis. A nomogram for predicting recurrence-free survival integrated radiomics models with clinical characteristics.

Results: Only the high-T_dys radiomics model was successfully established, yielding areas under the curve of 0.933 (training) and 0.792 (validation). Peritumoral imaging features on contrast-enhanced CT (fibrosis, inflammation, and atelectasis) were associated with CD8⁺ T cell exhaustion, evidenced by significantly higher high-T_dys proportions: 31.6% vs. 7.5% (P < 0.001), 35.6% vs. 8.1% (P < 0.001), and 40.0% vs. 11.8% (P = 0.028). The nomogram incorporating high-T_dys radiomics score, T stage, and N stage predicted 1- to 4-year predicting recurrence-free survival with areas under the curve of 0.733, 0.713, 0.637, and 0.600 (training), and 0.629, 0.669, 0.550, and 0.593 (validation).

Conclusion: Radiomics can predict the functional exhaustion of tumor-infiltrating CD8⁺ T cells in NSCLC, with specific imaging features associated with this process. Combining the radiomics model with clinical characteristics facilitates the assessment of patient prognosis.

Rationale and objectives: The non-invasive biomarkers for predicting progression-free survival (PFS) in patients with hepatocellular carcinoma (HCC) treated with immunotherapy and molecular targeted therapy combined with transarterial chemoembolization (IMT-MTT-TACE) are urgently needed to identify individuals who are likely to benefit from this treatment regimen. This study aims to develop a non-invasive imaging biomarker for predicting PFS in patients with HCC receiving IMT-MTT-TACE, leveraging the integration of deep learning, radiomics, and clinical factors.

Materials and methods: This study included 180 patients with HCC who were treated with IMT-MTT-TACE at two medical centers. Radiomic features were extracted from six sequences of multiparametric magnetic resonance imaging. Deep learning features were extracted based on the ResNet50 algorithm. A Cox regression combined model was developed by integrating significant clinical, radiomics, and deep learning features. Model performance was evaluated using the concordance index (C-index) and time-dependent receiver operating characteristic (ROC) analysis area under the curve (AUC).

Results: The C-reactive protein and alpha-fetoprotein in immunotherapy (CRAFITY) score was identified as an independent predictor of PFS (P < 0.05). In three cohorts, the C-index values for the deep learning model were 0.757, 0.751, and 0.744, respectively. The C-index values for the combined model were 0.803, 0.746, and 0.744, respectively. In the time-dependent ROC curve analysis predicting 1-year PFS, the AUC values for the combined model were 0.934 (95% confidence interval [CI]: 0.881-0.986), 0.842 (95% CI: 0.699-0.984), and 0.862 (95% CI: 0.725-0.998). The deep learning-based combined model demonstrated good predictive performance and exhibited strong robustness and generalizability.

Conclusion: The integration of CRAFITY score, radiomics, and deep learning features contributed to predicting PFS for patients with HCC undergoing IMT-MTT-TACE. This combined model holds promise for enabling precise pretreatment risk stratification and optimizing monitoring protocols, thereby guiding prognosis assessment and individualized clinical treatment decisions.

Rationale and objectives: This study aims to investigate the impact of iodine-125 intraluminal brachytherapy on stent patency time (SPT) and overall survival (OS) in patients with advanced pancreatic cancer complicated by obstructive jaundice, which may provide evidence for optimizing multimodal interventional strategies for patients with pancreatic cancer.

Methods: This retrospective study enrolled 262 patients with advanced pancreatic cancer complicated by obstructive jaundice who had underwent metallic biliary stent implantation. Patients were stratified into the ¹²⁵I group and the control group. A 1:1 propensity score matching (PSM) was performed using a caliper width of 0.2 standard deviations of the propensity score.

Results: After 1 month of biliary stent implantation, significant reductions of serum bilirubin, transaminases, and CA19-9 levels compared to predrainage were observed in this study. A median SPT of 7.63 months (95% CI 7.10-8.16), a stent restenosis rate of 36.2%, and a median OS of 9.40 months (95% CI 9.11-9.69) was demonstrated in the entire cohort demonstrated. After 1:1 PSM (62 patients per group), the ¹²⁵ I group showed a significantly longer median SPT (9.44 months vs. 6.21 months, P < 0.001) and lower stent restenosis rate (17.7% vs. 43.5%, P = 0.002) compared to the control group. However, no significant OS difference was observed between groups (10.59 months vs. 9.07 months, P = 0.248). Cox analysis suggested that intratumoral brachytherapy (HR 0.382, P < 0.001) was an independent protective factor for SPT; post-stent Eastern Cooperative Oncology Group (ECOG) 2 score (HR 1.572, P = 0.019) was an independent risk factor for SPT. The independent risk factors for OS included post-stent ECOG 2 score (HR 1.469, P = 0.042) and post-stent CA19-9≥500 ku/L (HR 1.322, P = 0.046).

Conclusion: Biliary stent combined with iodine-125 intraluminal brachytherapy significantly prolonged SPT compared to stent-alone in patients with advanced pancreatic cancer complicated by obstructive jaundice, although it did not significantly improve survival outcomes. Higher ECOG performance score and higher CA19-9 level after jaundice reduce were associated with poor prognosis.

Rationale and objectives: Preoperative meningioma grading is crucial for therapeutic planning and prognosis. This study aimed to prospectively evaluate microstructural parameters derived from time-dependent diffusion MRI (t_d-dMRI) for preoperative grading of low-grade meningiomas (LGMs) and high-grade meningiomas (HGMs), and to assess meningioma proliferative activity.

Materials and methods: In this prospective study, 214 meningioma patients underwent t_d-dMRI using pulsed and oscillating gradient diffusion sequences. Diffusion data were fitted using an optimized model to obtain intracellular volume fraction (f_in), cell diameter, cellularity, and extracellular diffusivity (D_ex). Patients were divided into two cohorts based on time: a derivation cohort (N = 159) and an independent validation cohort (N = 55). Receiver operating characteristic (ROC) analysis and logistic regression assessed diagnostic performance and defined optimal cut-offs in the derivation cohort. Pre-specified cut-offs were tested in the validation cohort. Spearman's rank correlations were calculated between t_d-dMRI-derived parameters, Ki-67 index, and histology-based measurements.

Results: HGMs showed significantly higher f_in and cellularity, and lower diameter and D_ex than LGMs in tumor solid regions (all P < 0.001). In the derivation cohort, f_in showed the highest AUC for grading (0.936). The combined model achieved a higher AUC (0.951). Both f_in and cellularity showed strong positive correlations with the Ki-67 index (r=0.711-0.721), whereas diameter and D_ex showed negative correlations (r=-0.663 to -0626). In the validation cohort, pre-specified cutoffs achieved similar AUCs for grading meningiomas. The t_d-dMRI-derived parameters correlated strongly with the histology-based measurements (r=0.641-0.773).

Conclusion: T_d-dMRI-derived microstructural parameters provide promising, non-invasive biomarkers for preoperative grading of LGMs and HGMs, with consistent diagnostic performance in both derivation and validation cohorts.