Purpose This study aims to develop and evaluate multi‐parametric MRI‐based radiomics for preoperative identification of epidermal growth factor receptor (EGFR) mutation, which is important in treatment planning for patients with thoracic spinal metastases from primary lung adenocarcinoma. Methods A total of 110 patients were enrolled between January 2016 and March 2019 as a primary cohort. A time‐independent validation cohort was conducted containing 52 patients consecutively enrolled from July 2019 to April 2021. The patients were pathologically diagnosed with thoracic spinal metastases from primary lung adenocarcinoma; all underwent T1‐weighted (T1W), T2‐weighted (T2W), and T2‐weighted fat‐suppressed (T2FS) MRI scans of the thoracic spinal. Handcrafted and deep learning‐based features were extracted and selected from each MRI modality, and used to build the radiomics signature. Various machine learning classifiers were developed and compared. A clinical‐radiomics nomogram integrating the combined rad signature and the most important clinical factor was constructed with receiver operating characteristic (ROC), calibration, and decision curves analysis (DCA) to evaluate the prediction performance. Results The combined radiomics signature derived from the joint of three modalities can effectively classify EGFR mutation and EGFR wild‐type patients, with an area under the ROC curve (AUC) of 0.886 (95% confidence interval [CI]: 0.826–0.947, SEN =0.935, SPE =0.688) in the training group and 0.803 (95% CI: 0.682–0.924, SEN = 0.700, SPE = 0.818) in the time‐independent validation group. The nomogram incorporating the combined radiomics signature and smoking status achieved the best prediction performance in the training (AUC = 0.888, 95% CI: 0.849–0.958, SEN = 0.839, SPE = 0.792) and time‐independent validation (AUC = 0.821, 95% CI: 0.692–0.929, SEN = 0.667, SPE = 0.909) cohorts. The DCA confirmed potential clinical usefulness of our nomogram. Conclusion Our study demonstrated the potential of multi‐parametric MRI‐based radiomics on preoperatively predicting the EGFR mutation. The proposed nomogram model can be considered as a new biomarker to guide the selection of individual treatment strategies for patients with thoracic spinal metastases from primary lung adenocarcinoma.

Background Preoperative assessment of the acquired resistance T790M mutation in patients with metastatic non‐small cell lung cancer (NSCLC) based on brain metastasis (BM) is important for early treatment decisions. Purpose To investigate preoperative magnetic resonance imaging (MRI)‐based radiomics for assessing T790M resistance mutation after epidermal growth factor receptor (EGFR)‐tyrosine kinase inhibitor (TKI) treatment in NSCLC patients with BM. Study Type Retrospective. Population One hundred and ten primary NSCLC patients with pathologically confirmed BM and T790M mutation status assessment from two centers divided into primary training (N = 53), internal validation (N = 27), and external validation (N = 30) sets. Field Strength/Sequence Contrast‐enhanced T1‐weighted (T1CE) and T2‐weighted (T2W) fast spin echo sequences at 3.0 T. Assessment Forty‐five (40.9%) patients were T790M‐positive and 65 (59.1%) patients were T790M‐negative. The tumor active area (TAA) and peritumoral edema area (POA) of BM were delineated on pre‐treatment T1CE and T2W images. Radiomics signatures were built based on features selected from TAA (RS‐TAA), POA (RS‐POA), and their combination (RS‐Com) to assess the T790M resistance mutation after EGFR‐TKI treatment. Statistical Tests Receiver operating characteristic (ROC) curves were used to assess the capabilities of the developed RSs. The area under the ROC curves (AUC), sensitivity, and specificity were generated as comparison metrics. Results We identified two features (from TAA) and three features (from POA) that are highly associated with the T790M mutation status. The developed RS‐TAA, RS‐POA, and RS‐Com showed good performance, with AUCs of 0.807, 0.807, and 0.864 in the internal validation, and 0.783, 0.814, and 0.860 in the external validation sets, respectively. Data Conclusion Pretreatment brain MRI of NSCLC patients with BM might effectively detect the T790M resistance mutation, with both TAA and POA having important values. The multi‐region combined radiomics signature may have potential to be a new biomarker for assessing T790M mutation. Level of Evidence 3 Technical Efficacy Stage 2

Background Axillary lymph node dissection (ALND) can be safely avoided in women with T1 or T2 primary invasive breast cancer (BC) and one to two metastatic sentinel lymph nodes (SLNs). However, cancellation of ALND based solely on SLN biopsy (SLNB) may lead to adverse outcomes. Therefore, preoperative assessment of LN tumor burden becomes a new focus for ALN status. Objective This study aimed to develop and validate a nomogram incorporating the radiomics score (rad-score) based on automated breast ultrasound system (ABUS) and other clinicopathological features for evaluating the ALN status in patients with early-stage BC preoperatively. Methods Totally 354 and 163 patients constituted the training and validation cohorts. They were divided into ALN low burden (<3 metastatic LNs) and high burden (≥3 metastatic LNs) based on the histopathological diagnosis. The radiomics features of the segmented breast tumor in ABUS images were extracted and selected to generate the rad-score of each patient. These rad-scores, along with the ALN burden predictors identified from the clinicopathologic characteristics, were included in the multivariate analysis to establish a nomogram. It was further evaluated in the training and validation cohorts. Results High ALN burdens accounted for 11.2% and 10.8% in the training and validation cohorts. The rad-score for each patient was developed based on 7 radiomics features extracted from the ABUS images. The radiomics nomogram was built with the rad-score, tumor size, US-reported LN status, and ABUS retraction phenomenon. It achieved better predictive efficacy than the nomogram without the rad-score and exhibited favorable discrimination, calibration and clinical utility in both cohorts. Conclusion We developed an ABUS-based radiomics nomogram for the preoperative prediction of ALN burden in BC patients. It would be utilized for the identification of patients with low ALN burden if further validated, which contributed to appropriate axillary treatment and might avoid unnecessary ALND.

No abstract