Technology in Cancer Research & Treatment最新文献

{"title":"EML4-ALK in Non-small Cell Lung Cancer: Molecular Mechanisms and Targeted Therapies.","authors":"Siqi Li, Lingbo Bao, Daijun Zhou, Dong Li","doi":"10.1177/15330338261457490","DOIUrl":"10.1177/15330338261457490","url":null,"abstract":"<p><p>Lung cancer is a malignancy characterized by high global incidence and mortality rates, with approximately 85% of cases classified as non-small cell lung cancer (NSCLC). Advances in molecular biology and targeted therapeutic agents have ushered NSCLC into the era of precision medicine. The echinoderm microtubule-associated protein-like 4 (EML4) and anaplastic lymphoma kinase (ALK) fusion gene represents a pivotal target for personalized treatment in NSCLC. Although ALK inhibitors exhibit significant efficacy against EML4-ALK-positive tumors, addressing drug resistance remains a major challenge. Identifying novel therapeutic targets and implementing combination therapies are essential for optimizing subsequent treatment strategies and improving overall prognosis. In the management of patients with EML4-ALK gene fusion, continuous monitoring of tumor progression and timely adjustment of treatment regimens based on disease status are critical to achieving long-term clinical benefits. This review comprehensively summarizes recent advances in the clinicopathological features, targeted drug development, resistance mechanisms, and potential therapeutic targets associated with the EML4-ALK fusion gene.</p>","PeriodicalId":22203,"journal":{"name":"Technology in Cancer Research & Treatment","volume":"25 ","pages":"15330338261457490"},"PeriodicalIF":2.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13309637/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148340141","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Preoperative Ternary Classification of Pulmonary Ground-Glass Nodules (AIS/MIA/IAC): ResNet-10 Outperforms Radiomics and Clinicoradiographic Models in Multicenter Study.","authors":"Dan Long, Zhichao Zuo, Huchuan Zhou, Wanyin Qi, Sanhong Zhang, Jinqiu Deng, Ziqiang Yang","doi":"10.1177/15330338261423265","DOIUrl":"10.1177/15330338261423265","url":null,"abstract":"<p><p>IntroductionPreoperative differentiation among adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), and invasive adenocarcinoma (IAC) is crucial for guiding ground-glass nodule (GGN) management. This multicenter study evaluated the comparative utility of deep learning (DL), radiomics, and conventional machine learning (cML)-based clinicoradiographic models for this ternary classification.MethodsWe developed four DL models (DenseNet-121, ResNet-10, ResNet-18, and VGG-13) for the ternary classification of AIS, MIA, and IAC using multicenter CT datasets. For comparative analysis, we constructed two additional classification models: (1) a radiomics model employing feature engineering through analysis of variance, recursive feature elimination with cross-validation, and least absolute shrinkage and selection operator, and (2) the cML-based clinicoradiographic model utilizing 12 different classifiers. The performance of all models was evaluated using the macro area under the curve (Macro-AUC) metric.Results847 GGNs postoperatively confirmed as lung adenocarcinoma were included in this multicenter study, which were randomly split into a training set (70%, n=592) and a validation set (30%, n=255). The DL model ResNet-10 demonstrated superior performance, achieving a Macro-AUC of 0.8055 (95% CI: 0.7723-0.8387), an accuracy of 0.6300 (95% CI: 0.5541-0.6764), and an F1-score of 0.4206 (95% CI: 0.3821-0.4598). This performance surpassed that of the radiomics model, which had a Macro-AUC of 0.7801 (95% CI: 0.7432-0.8170), an accuracy of 0.6100 (95% CI: 0.5276-0.6204), and an F1-score of 0.5505 (95% CI: 0.4983-0.6017), and the cML-based clinicoradiographic model, which achieved a Macro-AUC of 0.7770 (95% CI: 0.708-0.846), an accuracy of 0.6000 (95% CI: 0.5376-0.6604), and an F1-score of 0.4438 (95% CI: 0.3925-0.4961).ConclusionThe ResNet-10 network established a novel ternary classification model for predicting the invasiveness of GGNs. This approach provides clinically actionable insights that support surgical planning and facilitate risk-adapted management.</p>","PeriodicalId":22203,"journal":{"name":"Technology in Cancer Research & Treatment","volume":"25 ","pages":"15330338261423265"},"PeriodicalIF":2.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12909756/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146207796","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Metabolomic Characteristics of Patients With Non-Small Cell Lung Cancer in Plasma, Tissue and Exhaled Breath Condensate: A Case-Control Study.","authors":"Dewei Li, Shuting Yang, Ling Zhang, Lingling Dong, Yueeryeti Sailai, Limin Yao, Chengyu Jin, Xuemei Wei","doi":"10.1177/15330338261470635","DOIUrl":"10.1177/15330338261470635","url":null,"abstract":"<p><p>IntroductionLung cancer is a malignant tumor with high global incidence and mortality. Early screening and diagnosis can substantially extend patient survival, yet achieving this remains a challenging objective.MethodsWe collected tissue and plasma samples from patients with benign pulmonary nodules and non-small cell lung cancer (NSCLC). Exhaled breath condensate (EBC) samples were obtained from these patients using a breath condenser. Differential metabolites were identified in samples from 22 NSCLC patients and 20 benign controls by means of ultra-performance liquid chromatography-high resolution mass spectrometry-based untargeted metabolomics.ResultsSeveral differential metabolites overlapped between plasma and tissue samples. Further analysis indicated that the differential metabolites from all three sample types participated in central carbon metabolism in cancer, protein digestion and absorption, and aminoacyl-tRNA biosynthesis. ROC curve analysis of plasma and tissue metabolites showed that each of the top 10 upregulated and downregulated metabolites yielded AUC values greater than 0.8. In EBC, only 29 differential metabolites were detected. Among these, lysine, acetildenafil, and 1-(cyclohexylmethyl) proline may be involved in the progression of NSCLC.ConclusionThis study identified distinctive metabolic markers in plasma, tissue, and EBC from NSCLC patients. However, the purpose of this observational study was merely exploratory, rather than to use these differential metabolites as a diagnostic test for non-small cell lung cancer.</p>","PeriodicalId":22203,"journal":{"name":"Technology in Cancer Research & Treatment","volume":"25 ","pages":"15330338261470635"},"PeriodicalIF":2.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13428115/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148631327","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Neoadjuvant Chemoimmunotherapy versus Neoadjuvant Chemoradiotherapy versus Neoadjuvant Chemotherapy for Esophageal Squamous Cell Carcinoma: A Real-World, Three-Arm, Retrospective Cohort Study from the Chinese National Cancer Center.","authors":"Long Zhang, Zehang Zhang, Guochao Zhang, Chao Zheng, Yuzhuo Zhang, Zhichao Zuo, Shuangping Zhang, Yong Li, Yin Li, Qi Xue","doi":"10.1177/15330338261459197","DOIUrl":"10.1177/15330338261459197","url":null,"abstract":"<p><p>IntroductionThe optimal neoadjuvant strategy for locally-advanced esophageal squamous cell carcinoma (ESCC) remains controversial. This study aimed to compare the efficacy and safety of neoadjuvant chemoimmunotherapy (NCIT), chemoradiotherapy (NCRT), and chemotherapy alone (NCT).MethodsWe retrospectively analyzed perioperative clinical data from ESCC patients treated with NCIT, NCRT, or NCT at the National Cancer Center in China from 2009 to 2023. The primary outcomes were pathological response, survival, recurrence pattern and surgical complications.ResultsClinicopathological data from 462 patients (NCIT=122, NCRT=195, NCT=145) were analyzed. The pathological complete response (pCR) rate in the NCIT group (29.5%) was comparable to the NCRT group (28.7%) but significantly higher than the NCT group (3.4%). The overall postoperative complication rate in the NCIT group was higher than in the NCT group (33.6% vs. 14.5%, p<0.001) but comparable to the NCRT group (33.6% vs. 28.2%, p=0.309), with a manageable level. The overall recurrence rates were lower in NCIT (27.0%) than NCRT (48.2%) and NCT (44.1%). Notably, the distant metastasis rate in the NCIT group was significantly lower than that in the NCRT group (11.5% vs. 22.1%, p=0.017) and the NCT group (11.5% vs. 22.8%, p=0.016). As for long-term survival, the 5-year overall survival rate for NCIT was 72.5%, significantly higher than both NCRT (55.7%, p<0.001) and NCT (60.8%, p=0.004). Similarly, the 5-year recurrence-free survival rate for NCIT was 70.7%, significantly surpassing both NCRT (49.8%, p<0.001) and NCT (46.2%, p<0.001). No significant survival differences were observed between the NCRT and NCT groups.ConclusionNCIT provides a significant survival benefit over both NCRT and NCT in locally advanced ESCC in real-world clinical practice, achieving a comparable pCR rate to NCRT and decreased recurrence rate. It's a favorable neoadjuvant option, with the potential to be a novel treatment strategy. Prospective randomized trials to validate our findings are warranted.</p>","PeriodicalId":22203,"journal":{"name":"Technology in Cancer Research & Treatment","volume":"25 ","pages":"15330338261459197"},"PeriodicalIF":2.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13247292/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148206125","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Lightweight Skeletal Muscle Intelligent Segmentation Network Based on Planning CT for Cervical Cancer Radiotherapy.","authors":"Liming Lu, Xiwei Chen, Jing Liu, Gongfu Chen, Junyue Shi, Zhe Wu, Gaokui He","doi":"10.1177/15330338261446535","DOIUrl":"10.1177/15330338261446535","url":null,"abstract":"<p><p>PurposeA lightweight deep learning network SMA-Net was proposed to intelligently segment the skeletal muscle of the third lumbar (L3) level in patients with cervical cancer radiotherapy, and the segmentation performance of the network was evaluated.Methods and MaterialsA total of 160 eligible patients with cervical cancer admitted to the oncology department of our hospital from September 2021 to June 2024 were randomly divided into training set (N = 112), validation set (N = 16) and test set (N = 32) according to 7 : 1 : 2. The lightweight Mamba architecture is introduced into the UNet network, and the SAB and CAB attention mechanisms are introduced on the skip connection. The attention mechanism is used to suppress the irrelevant information in the image and highlight the important local features. The trained network is geometrically evaluated on the test set for segmentation performance, comparison of manual segmentation and predicted skeletal muscle area (SMA). Compare the parameters and computations of SMA-Net with existing networks.ResultsThe dice similarity coefficient of SMA-Net network for skeletal muscle segmentation was 89.16%, the sensitivity SEN was 88.21%, the positive predictive value PPV was 90.13% and the 95% Hausdorff distance was 5.30mm. Manual segmentation is basically close to the predicted SMA. Our proposed network for cervical cancer patients predicted sarcopenia with 87.5% accuracy, 92.31% precision, 80% recall, 85.72% F1-Score, and 0.871 AUC. The calculation amount of SMA-UNet network is 1.50 GFLOPS, and the parameter amount is 1.24 M. The radiologist's scores show that minor and no revision accounted for 93.75% on manual revision of skeletal muscle.ConclusionThe lightweight SMA-Net proposed in this study can accurately segment L3 skeletal muscle and quickly calculate its area, which basically meets the clinical application and is convenient for clinical deployment. It is helpful for clinicians to quickly diagnose sarcopenia in patients with cervical cancer, save medical resources, reduce the workload of physicians, and improve diagnostic efficiency.</p>","PeriodicalId":22203,"journal":{"name":"Technology in Cancer Research & Treatment","volume":"25 ","pages":"15330338261446535"},"PeriodicalIF":2.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13133491/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147781794","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Research on Predicting the Efficacy of Immunotherapy for Non-small Cell Lung Cancer Based on CT Radiomics and Peripheral Blood Inflammatory Markers.","authors":"Xinyuan Wen, Yihan Li, Yindi Hu, Aiqi Chen, Kai Wang, Wentao Zou, Bo Xie, Yichuan Ma","doi":"10.1177/15330338261465744","DOIUrl":"10.1177/15330338261465744","url":null,"abstract":"<p><p>IntroductionWe developed and validated predictive models based on CT radiomics and peripheral blood inflammatory markers to predict immunotherapy response in non-small cell lung cancer (NSCLC).MethodsThis retrospective study included 128 NSCLC patients receiving anti-PD-1/PD-L1 monotherapy as second-line therapy between January 2020 and January 2024. All eligible patients who met the inclusion criteria during the study period were consecutively included to minimize selection bias. Baseline non-contrast and contrast-enhanced chest CT acquired within 15 days before immunotherapy initiation was used for radiomics. Patients were randomly split into training and testing cohorts (7:3). Radiomics features were extracted and selected using LASSO with cross-validation followed by MRMR. A radiomics model, a clinical model (inflammatory markers), and a combined nomogram model were developed. Model performance was assessed by AUC and calibration; clinical utility was evaluated by decision curve analysis. Reporting followed TRIPOD+AI guidelines.ResultsThe combined model achieved the best performance (AUC 0.8802, 95% CI 0.8106-0.9497 in the training cohort; AUC 0.7329, 95% CI 0.5679-0.8979 in the testing cohort), outperforming the clinical model (AUC 0.6481 and 0.5325) and the radiomics model (AUC 0.7878 and 0.5621). DeLong tests for pairwise AUC comparisons yielded P values of 0.243 and 0.478, respectively.ConclusionsThe combined clinical-radiomics model may help identify NSCLC patients more likely to respond to immunotherapy. External validation is required before clinical implementation.</p>","PeriodicalId":22203,"journal":{"name":"Technology in Cancer Research & Treatment","volume":"25 ","pages":"15330338261465744"},"PeriodicalIF":2.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13332304/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148377042","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Chimeric Antigen Receptor T-Cell Therapy for the Treatment of Melanoma: A Systematic Review of Phase One Clinical Trials.","authors":"Pavan Shet, Prerana Ghosh, Karan Varshney, Sameen Sawdagar","doi":"10.1177/15330338261452884","DOIUrl":"10.1177/15330338261452884","url":null,"abstract":"<p><p>Introduction<i>Melanoma is a growing health concern,</i> with global incidence increasing annually. However, treatment for refractory disease is currently limited. Chimeric antigen receptor (CAR) T-cell therapy has shown promising results in haematological malignancies. Despite this, its utility in solid cancers, such as melanoma, is not yet established. We performed a systematic review of phase one trials to help determine and describe foundational understandings of the safety and efficacy of CAR T-cell therapy for melanoma.MethodsThis systematic review was conducted in accordance with the Cochrane Collaboration Handbook for Systematic Review of Interventions and the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) Statement Guidelines. Databases were searched to identify articles before November 18, 2025, which described the use of CAR T-cell therapy in melanoma patients with published results. Study, patient, and treatment characteristics were summarised and analysed.ResultsIn total, 2726 articles were screened. Ultimately 5 studies comprising 15 patients with melanoma treated with CAR T-cell therapy were analysed and discussed. T-cell therapies were all autologous and the targets for the CAR T-cell therapies included GD2, PD-1 and cMET antigens across a wide variety of cell doses. Across pooled data, clinical responses varied, with some patients displaying either a partial or complete clinical response. Treatment related effects were of grade one or two severity, with no serious adverse effects reported, including neurotoxicity side effects, or treatment limited related toxicity/mortality events. Peak expansion typically occurred within day 7 to 28, but persistence was limited across studies.ConclusionCAR T-cell therapy for melanoma is a novel treatment approach in its infancy. Results are preliminary, and remain largely descriptive, as findings are limited by heterogeneity and small sample sizes. Further investigation through additional phase one trials, and subsequent phase two/three trials, are required for better establishing direct clinical viability.</p>","PeriodicalId":22203,"journal":{"name":"Technology in Cancer Research & Treatment","volume":"25 ","pages":"15330338261452884"},"PeriodicalIF":2.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13291571/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148272937","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Proteomics in Acute Myeloid Leukemia: Current Applications in Precision Medicine and Targeted Immunotherapy.","authors":"Eva De Backer, Florian Van Oers, Zwi Berneman, Sébastien Anguille","doi":"10.1177/15330338261464315","DOIUrl":"10.1177/15330338261464315","url":null,"abstract":"<p><p>Acute myeloid leukemia (AML) is a heterogeneous and aggressive malignancy with limited therapeutic options and high relapse rates. Despite advances in genomic profiling, many genetic aberrations remain untargetable, and current risk stratification models often fail to predict treatment responses. Proteomics offers a complementary approach by directly measuring protein abundance, post-translational modifications, and protein-protein interactions, providing mechanistic insights into drug resistance, disease progression, and therapeutic vulnerabilities. In this review, we explore the emerging role of proteomics in AML, focusing on its application in biomarker discovery, prediction of drug responses, and identification of novel therapeutic targets. Special attention is given to antigen discovery for immunotherapy, where surface and immunopeptidomics enable the identification of AML-specific antigens and neoepitopes. These insights are critical for the development of antigen-targeted therapies, including chimeric antigen receptor (CAR) and T cell receptor (TCR)-based immunotherapies. Integrating proteomics into a multiomics framework could provide actionable insights for guiding precision medicine and improving AML outcomes.</p>","PeriodicalId":22203,"journal":{"name":"Technology in Cancer Research & Treatment","volume":"25 ","pages":"15330338261464315"},"PeriodicalIF":2.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13319607/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148346128","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Development and Preliminary Dosimetric Evaluation of a 3D-Printed Custom Tissue Compensator for Head and Neck Cancer Radiotherapy: A Proof-of-Concept Case Series.","authors":"Chuxiang Jian, Xiang Xia, Jiazhou Wang, Chenlu Lian, Xuexue Cao, Ping Zhou, Dayu Xu, Qian Wu, Chengrun Du","doi":"10.1177/15330338261478186","DOIUrl":"https://doi.org/10.1177/15330338261478186","url":null,"abstract":"<p><p>IntroductionAchieving an optimal therapeutic ratio in head and neck cancer (HNC) radiotherapy requires balancing dose conformity with organ-at-risk (OAR) sparing. Traditional auxiliary devices often lack the customized geometry needed for modern techniques. This study developed and evaluated a 3D-printed integrated tissue compensator that serves as both a positioning stent for OAR displacement and a conformal bolus for dose build-up.MethodsIn this case series, five HNC patients (including floor of mouth, tongue, and nasal cavity and sinus cancers) were prospectively and consecutively enrolled, and their computed tomography (CT) data were used to fabricate the compensators. Dosimetric comparisons were conducted between the 3D-printed compensator, a simulated air cavity, and a conventional bite block.ResultsThe integrated device demonstrated adequate geometric fidelity, maintaining random translational setup errors (σ) at ≤ 1.34 mm, comparable to traditional immobilization. Compared to the conventional group, the device demonstrated a numerical reduction in the maximum dose to the hard palate and the mean dose to the external surface, while ensuring effective dose build-up in the target region. Compared with the air cavity group, the observed numerical reductions in dose for key OARs (oral cavity, mandible, and parotid gland) approached but did not reach conventional statistical significance (p = 0.063). Furthermore, the device maintained comparable conformity (CI) and homogeneity (HI) indices, showing an observed numerical improvement in homogeneity for high-dose targets.ConclusionThis feasibility study demonstrates that a 3D-printed integrated tissue compensator offers potential dosimetric advantages over conventional approaches. By maintaining setup reproducibility and OAR sparing, this device represents a practical exploratory tool for HNC radiotherapy.</p>","PeriodicalId":22203,"journal":{"name":"Technology in Cancer Research & Treatment","volume":"25 ","pages":"15330338261478186"},"PeriodicalIF":2.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13473770/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148762150","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Transformer-Based Deep Learning Model Using MRI-Derived Microvascular Atlas for Predicting Lymphovascular Invasion in Breast Cancer Patients.","authors":"Hui Zhang, Qiaomei Zhao, Qian Wang, Yan Zhu, Yating Wang, Wenting Guan, Bo Zhu, Genji Bai","doi":"10.1177/15330338261426280","DOIUrl":"10.1177/15330338261426280","url":null,"abstract":"<p><p>IntroductionLymphovascular invasion (LVI), an aggressive pathological manifestation of breast cancer, is closely associated with increased risk of distant metastasis and poor prognosis. This study proposes a novel modeling strategy that integrates MRI-derived microvascular atlas parameters with the TwinsSVT deep learning architecture to enable noninvasive prediction of LVI status in breast cancer patients and to explore its biological interpretability.Materials and MethodsA total of 436 breast cancer patients from two medical centers, all pathologically confirmed postoperatively, were retrospectively enrolled. All patients underwent high-resolution multi-b-value diffusion-weighted imaging (DWI) prior to surgery. From the MRI data, four types of microvascular simulation parameter maps were reconstructed within tumor regions: apparent diffusion coefficient (ADC), mean flow velocity (v_m), velocity dispersion (v_s), and angiographic branching index (ANB), aiming to characterize intratumoral microcirculation and vascular structural complexity. These functional parametric maps were individually input into separate encoder branches of the TwinsSVT model to extract multi-scale spatial features. A multi-layer Transformer fusion module was then employed to capture structural interactions across modalities, thereby constructing a multi-parametric fusion model. Model performance was evaluated using metrics including area under the curve (AUC) and F1 score.ResultsCompared with single-parameter models, the multi-parametric fusion model demonstrated significantly improved predictive performance, with AUCs of 0.881 (95% CI: 0.781-0.982) and 0.859 (95% CI: 0.764-0.953) in internal and external validation cohorts, respectively. Grad-CAM visualizations revealed that the model predominantly focused on tumor margins and regions of high vascular density, suggesting a strong correlation between the model's attention and actual pathological structures.ConclusionThe deep learning model constructed based on MRI-derived microvascular simulation atlases enables noninvasive preoperative prediction of LVI status in breast cancer patients. By effectively capturing structural information and offering biological interpretability, the model holds promise as a robust imaging-based tool for precision subtyping and clinical decision support.</p>","PeriodicalId":22203,"journal":{"name":"Technology in Cancer Research & Treatment","volume":"25 ","pages":"15330338261426280"},"PeriodicalIF":2.7,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12954026/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147318306","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}