Radiology. Imaging cancer最新文献

Purpose To develop deep learning (DL) radiopathomics models based on contrast-enhanced MRI and pathologic imaging to predict vessels encapsulating tumor clusters (VETC) and survival in hepatocellular carcinoma (HCC). Materials and Methods In this retrospective, multicenter study, 578 patients with HCC (mean age [±SD], 59 years ± 10; 442 male, 136 female) were divided into the training (n = 317), internal (n = 137), and external (n = 124) test sets. DL radiomics and pathomics models were developed to predict VETC using gadoxetic acid-enhanced MR and pathologic images. Deep radiomics score (DRS) and handcrafted and deep pathomics scores were compared between the group with VETC pattern in HCC (VETC+) and group without VETC pattern in HCC (VETC-). Multivariable Cox regression analyses were performed to identify independent prognostic factors, and the radiopathomics nomogram models were developed for early recurrence and progression-free survival (PFS). The prognostic power was evaluated using the concordance index (C index) and time-dependent receiver operating characteristic (ROC) curves. Results In the external test set, the Swin Transformer showed good performance for predicting VETC in both DL radiomics (area under the ROC curve [AUC], 0.77-0.79) and pathomics (AUC, 0.79) models. Patients with VETC+ HCC had significantly higher DRS and handcrafted and deep pathomics scores compared with patients with VETC- HCC in all datasets (all P < .001). The radiopathomics nomogram model incorporating DRS in the arterial phase and the handcrafted and deep pathomics scores achieved C indexes of 0.69, 0.60, and 0.67 for early recurrence and time-dependent AUCs of 0.83 (95% CI: 0.76, 0.91), 0.81 (95% CI: 0.68, 0.94), and 0.78 (95% CI: 0.67, 0.88) for 3-year PFS in the training, internal, and external test sets, respectively. Early recurrence and PFS rates statistically significantly differed between the high- and low-risk patients stratified by the radiopathomics nomogram model (all P < .05). Conclusion DL radiopathomics models effectively helped to predict VETC in HCC and assess the risk for early recurrence and PFS. Keywords: Hepatocellular Carcinoma, Deep Learning, MRI, Radiopathomics, Survival Supplemental material is available for this article. © RSNA, 2025.

Purpose To develop three nomograms integrating apparent diffusion coefficients (ADCs) derived from diffusion-weighted imaging to predict the status of pretreatment axillary lymph nodes (ALNs) (task 1), nonsentinel lymph nodes (task 2), and ALNs after neoadjuvant chemotherapy treatment (task 3) in patients with breast cancer. Materials and Methods Pretreatment MRI scans, including diffusion-weighted images, were retrospectively acquired from patients with breast cancer at multiple centers from May 2019 to May 2023. ADC values and clinicopathologic features were measured. Uni- and multivariable logistic regression analyses were performed to identify independent predictors of ALN metastasis. These predictors were incorporated into nomogram models for each of the three tasks. Model performance was assessed with area under the receiver operating characteristic curve (AUC) analysis in training and two external testing datasets. Results The study included 961 female patients (mean age ± SD, 50 years ± 10) with breast cancer from three hospitals. In the three tasks, the ADC values of the ALN metastasis groups were lower than those of the nonmetastasis groups (all P < .05). The nomogram models combining ADC values and clinicopathologic features demonstrated high predictive performance for each task in the training cohort (task 1: AUC, 0.90; task 2: AUC, 0.74; task 3: AUC, 0.75), external testing cohort 1 (task 1: AUC, 0.86; task 3: AUC, 0.82), and external testing cohort 2 (task 1: AUC, 0.90; task 3: AUC, 0.84). Conclusion Nomograms incorporating ADCs and clinicopathologic features demonstrated good performance in predicting ALN metastasis in patients with breast cancer. Keywords: Breast, MR-Functional Imaging, MR-Diffusion Weighted Imaging, Apparent Diffusion Coefficient, Axillary Lymph Node Metastasis, Nonsentinel Lymph Node Metastasis, Neoadjuvant Chemotherapy, Nonogram Supplemental material is available for this article. © RSNA, 2025.

Purpose To evaluate the diagnostic performance of intraoperative supine MRI (isMRI) in identifying residual tumor burden immediately after breast-conserving therapy (BCT). Materials and Methods This single-institution prospective study (April 2012-December 2022) included 43 consecutive participants with stage 0-II breast cancer. Three participants with multicentric disease were excluded from the final analysis. Preoperative supine MRI was performed after standard preoperative prone MRI to compare tumor sizes and distances to the nipple, chest wall, and skin. After lumpectomy, the saline-filled surgical cavity was assessed for residual tumor at 3-T isMRI in the operating suite. Diagnostic accuracy of isMRI findings in identifying residual tumor at resection margins was assessed using histopathology of shave margin specimens as the reference standard. Performance metrics of isMRI and re-excision rates were analyzed at per-participant and per-margin levels. Results Forty participants (median age, 58.5 years; range, 40-76 years) with 44 breast cancers (36 unifocal and four multifocal) underwent BCT, all with a single lumpectomy site. Margin assessment using isMRI yielded accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of 80%, 50%, 93%, 75%, and 81% per participant, respectively; and 93%, 52%, 97%, 65%, and 96% per margin. Second re-excision was avoided in two of six (33%) participants with true-positive isMRI readings, decreasing the final re-excision rate from 18% to 13%. Histopathology of six false-negative isMRI cases revealed residual invasive carcinomas, all smaller than 0.3 cm, or intermediate-to-high grade ductal carcinoma in situ. Conclusion Intraoperative assessment for residual tumor after BCT using isMRI demonstrated promising accuracy to guide targeted margin clearance within the same operation. Keywords: Breast, MR-Imaging, MR-Dynamic Contrast Enhanced, Oncology Supplemental material is available for this article. © RSNA, 2025.

Purpose To evaluate the safety and efficacy of US-guided thermal ablation in the treatment of Bethesda III thyroid nodules with negative eight-gene panel testing results. Materials and Methods This retrospective single-center study included patients with thyroid nodules diagnosed as Bethesda category III (atypia of undetermined significance) at fine-needle aspiration biopsy and with negative eight-gene testing results who were treated with US-guided microwave ablation (MWA) between July 2020 and September 2023. Incidence of complications, technical success rate (TSR), volume reduction rate (VRR), nodule recurrence, and thyroid function were evaluated over a follow-up period of 2 years. Data before and after MWA were compared using variance analysis and the Cochran-Mantel-Haenszel χ² test. Results A total of 101 Bethesda III nodules were detected in 95 patients (mean ± SD age, 47.08 years ±14.63; 79 female patients, 16 male patients), all of which were completely ablated (100% TSR). Two patients experienced mild neck swelling and pressure sensation after the minimally invasive operation, and the incidence of postoperative complications was 2% (two of 95). None of the patients experienced tumor recurrence or progression. At 2-year follow-up, the mean VRR of the ablated area was 90.88% ± 13.59 in 15 patients; 87% (13 of 15) of these patients had a 100% VRR. There was no evidence of a difference in thyroid function before and after MWA from 1 to 24 months (P = .15-.99). Conclusion US-guided MWA was safe and effective for the treatment of Bethesda III thyroid nodules with negative eight-gene panel testing results. Keywords: Ablation Techniques, Radiation Therapy/Oncology, Head/Neck, Thyroid, Safety, Observer Performance Published under a CC BY 4.0 license.