Radiomic analysis reveals diverse prognostic and molecular insights into the response of breast cancer to neoadjuvant chemotherapy: a multicohort study.

BACKGROUND: Breast cancer patients exhibit various response patterns to neoadjuvant chemotherapy (NAC). However, it is uncertain whether diverse tumor response patterns to NAC in breast cancer patients can predict survival outcomes. We aimed to develop and validate radiomic signatures indicative of tumor shrinkage and therapeutic response for improved survival analysis. METHODS: This retrospective, multicohort study included three datasets. The development dataset, consisting of preoperative and early NAC DCE-MRI data from 255 patients, was used to create an imaging signature-based multitask model for predicting tumor shrinkage patterns and pathological complete response (pCR). Patients were categorized as pCR, nonpCR with concentric shrinkage (CS), or nonpCR with non-CS, with prediction performance measured by the area under the curve (AUC). The prognostic validation dataset (n = 174) was used to assess the prognostic value of the imaging signatures for overall survival (OS) and recurrence-free survival (RFS) using a multivariate Cox model. The gene expression data (genomic validation dataset, n = 112) were analyzed to determine the biological basis of the response patterns. RESULTS: The multitask learning model, utilizing 17 radiomic signatures, achieved AUCs of 0.886 for predicting tumor shrinkage and 0.760 for predicting pCR. Patients who achieved pCR had the best survival outcomes, while nonpCR patients with a CS pattern had better survival than non-CS patients did, with significant differences in OS and RFS (p = 0.00012 and p = 0.00063, respectively). Gene expression analysis highlighted the involvement of the IL-17 and estrogen signaling pathways in response variability. CONCLUSIONS: Radiomic signatures effectively predict NAC response patterns in breast cancer patients and are associated with specific survival outcomes. The CS pattern in nonpCR patients indicates better survival.

Differentiating between benign and malignant breast lesions using dual-energy CT-based model: development and validation.

OBJECTIVES: To develop and validate a dual-energy CT (DECT)-based model for noninvasively differentiating between benign and malignant breast lesions detected on DECT. MATERIALS AND METHODS: This study prospectively enrolled patients with suspected breast cancer who underwent dual-phase contrast-enhanced DECT from July 2022 to July 2023. Breast lesions were randomly divided into the training and test cohorts at a ratio of 7:3. Clinical characteristics, DECT-based morphological features, and DECT quantitative parameters were collected. Univariate analyses and multivariate logistic regression were performed to determine independent predictors of benign and malignant breast lesions. An individualized model was constructed. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the diagnostic ability of the model, whose calibration and clinical usefulness were assessed by calibration curve and decision curve analysis. RESULTS: This study included 200 patients (mean age, 49.9 ± 11.9 years; age range, 22-83 years) with 222 breast lesions. Age, lesion shape, and the effective atomic number (Zeff) in the venous phase were significant independent predictors of breast lesions (all p < 0.05). The discriminative power of the model incorporating these three factors was high, with AUCs of 0.844 (95%CI 0.764-0.925) and 0.791 (95% CI 0.647-0.935) in the training and test cohorts, respectively. The constructed model showed a preferable fitting (all p > 0.05 by the Hosmer-Lemeshow test) and provided enhanced net benefits than simple default strategies within a wide range of threshold probabilities in both cohorts. CONCLUSION: The DECT-based model showed a favorable diagnostic performance for noninvasive differentiation between benign and malignant breast lesions detected on DECT. CRITICAL RELEVANCE STATEMENT: The combination of clinical and morphological characteristics and DECT-derived parameter have the potential to identify benign and malignant breast lesions and it may be useful for incidental breast lesions on DECT to decide if further work-up is needed. KEY POINTS: It is important to characterize incidental breast lesions on DECT for patient management. DECT-based model can differentiate benign and malignant breast lesions with good performance. DECT-based model is a potential tool for distinguishing breast lesions detected on DECT.

Magnetic resonance imaging-based prognostic model for subsequent distant metastasis in patients with ipsilateral breast tumor recurrence following breast-conserving surgery.

Background: Ipsilateral breast tumor recurrence (IBTR) following breast-conserving surgery (BCS) has been considered a risk factor for distant metastasis (DM). Limited data are available regarding the subsequent outcomes after IBTR. Therefore, this study aimed to determine the clinical course after IBTR and develop a magnetic resonance imaging (MRI)-based predictive model for subsequent DM. Methods: We retrospectively extracted quantitative features from MRI to construct a radiomics cohort, with all eligible patients undergoing preoperative MRI at time of primary tumor and IBTR between 2010 and 2018. Multivariate Cox analysis was performed to identify factors associated with DM. Three models were constructed using different sets of clinicopathological, qualitative, and quantitative MRI features and compared. Additionally, Kaplan-Meier analysis was performed to assess the prognostic value of the optimal model. Results: Among the 183 patients who experienced IBTR, 47 who underwent MRI for both primary and recurrent tumors were enrolled. Multivariate analysis demonstrated that the independent prognostic factors were human epidermal growth factor receptor 2 (HER2) status [hazard ratio (HR) =5.40] and background parenchymal enhancement (BPE) (HR =7.94) (all P values <0.01). Furthermore, four quantitative MRI features of recurrent tumors were selected through the least absolute shrinkage and selection operator (LASSO) method. The combined model exhibited superior performance [concordance index (C-index) 0.77] compared to the clinicoradiological model (C-index 0.71; P=0.006) and radiomics model (C-index 0.70; and P=0.01). Furthermore, the combined model successfully categorized patients into low- and high-risk subgroups with distinct prognoses (P<0.001). Conclusions: The clinicopathological and MRI features of IBTR were associated with secondary events following surgery. Additionally, the MRI-based combined model exhibited the highest predictive efficacy. These findings could be helpful in risk stratification and tailoring follow-up strategies in patients with IBTR.

Predicting Regional Recurrence and Prognosis in Stereotactic Body Radiation Therapy-Treated Clinical Stage I Non-small Cell Lung Cancer Using a Radiomics Model Constructed With Surgical Data.

PURPOSE: Risk stratification of regional recurrence (RR) is clinically important in the design of adjuvant treatment and surveillance strategies in patients with clinical stage I non-small cell lung cancer (NSCLC) treated with stereotactic body radiation therapy (SBRT). This study aimed to develop a radiomics model predicting occult lymph node metastasis (OLNM) using surgical data and apply it to the prediction of RR in SBRT-treated early-stage NSCLC patients. METHODS AND MATERIALS: Patients with clinical stage I NSCLC who underwent curative surgery with systematic lymph node dissection from January 2013 to December 2018 (the training cohort) and from January 2019 to December 2020 (the validation cohort) were included. A preoperative computed tomography-based radiomics model, a clinical feature model, and a fusion model predicting OLNM were constructed. The performance of the 3 models was quantified and compared in the training and validation cohorts. Subsequently, the radiomics model was used to predict RR in a cohort of consecutive SBRT-treated early-stage NSCLC patients from 2 academic medical centers. RESULTS: A total of 769 patients were included. Eight computed tomography features were identified in the radiomics model, achieving areas under the curves of 0.85 (95% CI, 0.81-0.89) and 0.83 (95% CI, 0.80-0.88) in the training and validation cohorts, respectively. Nevertheless, adding clinical features did not improve the performance of the radiomics model. With a median follow-up of 40.0 (95% CI, 35.2-44.8) months, 32 of the 213 patients in the SBRT cohort developed RR and those in the high-risk group based on the radiomics model had a higher cumulative incidence of RR (P < .001) and shorter regional recurrence-free survival (P = .02), progression-free survival (P = .004) and overall survival (P = .006) than those in the low-risk group. CONCLUSIONS: The radiomics model based on pathologically confirmed data effectively identified patients with OLNM, which may be useful in the risk stratification among SBRT-treated patients with clinical stage I NSCLC.

Automatic prediction of non-iodine-avid status in lung metastases for radioactive I131 treatment in differentiated thyroid cancer patients.

Objectives: The growing incidence of differentiated thyroid cancer (DTC) have been linked to insulin resistance and metabolic syndrome. The imperative need for developing effective diagnostic imaging tools to predict the non-iodine-avid status of lung metastasis (LMs) in differentiated thyroid cancer (DTC) patients is underscored to prevent unnecessary radioactive iodine treatment (RAI). Methods: Primary cohort consisted 1962 pretreated LMs of 496 consecutive DTC patients with pretreated initially diagnosed LMs who underwent chest CT and subsequent post-treatment radioiodine SPECT. After automatic lesion segmentation by SE V-Net, SE Net deep learning was trained to predict non-iodine-avid status of LMs. External validation cohort contained 123 pretreated LMs of 24 consecutive patients from other two hospitals. Stepwise validation was further performed according to the nodule's largest diameter. Results: The SE-Net deep learning network yielded area under the receiver operating characteristic curve (AUC) values of 0.879 (95% confidence interval: 0.852-0.906) and 0.713 (95% confidence interval: 0.613-0.813) for internal and external validation. With the LM diameter decreasing from ≥10mm to ≤4mm, the AUCs remained relatively stable, for smallest nodules (≤4mm), the model yielded an AUC of 0.783. Decision curve analysis showed that most patients benefited using deep learning to decide radioactive I131 treatment. Conclusion: This study presents a noninvasive, less radioactive and fully automatic approach that can facilitate suitable DTC patient selection for RAI therapy of LMs. Further prospective multicenter studies with larger study cohorts and related metabolic factors should address the possibility of comprehensive clinical transformation.

Predicting Prognosis of Phyllodes Tumors Using a Mammography- and Magnetic Resonance Imaging-Based Radiomics Model: A Preliminary Study.

PURPOSE: To investigate whether a radiomics model based on mammography (MG) and magnetic resonance imaging (MRI) can be used to predict disease-free survival (DFS) after phyllodes tumor (PT) surgery. METHOD: About 131 PT patients who underwent MG and MRI before surgery between January 2010 and December 2020 were retrospectively enrolled, including 15 patients with recurrence and metastasis and 116 without recurrence. 884 and 3138 radiomic features were extracted from MG and MR images, respectively. Then, multiple radiomics models were established to predict the recurrence risk of the patients by applying a support vector machine classifier. The area under the ROC curve (AUC) was calculated to evaluate model performance. After dividing the patients into high- and low-risk groups based on the predicted radiomics scores, survival analysis was conducted to compare differences between the groups. RESULTS: In total, 3 MG-related and 5 MRI-related radiomic models were established; the prediction performance of the T1WI feature fusion model was the best, with an AUC value of 0.93. After combining the features of MG and MRI, the AUC increased to 0.95. Furthermore, the MG, MRI and all-image radiomic models had statistically significant differences in survival between the high- and low-risk groups (P < .001). All-image radiomics model showed higher survival performance than the MG and MRI radiomics models alone. CONCLUSIONS: Radiomics features based on preoperative MG and MR images can predict DFS after PT surgery, and the prediction score of the image radiomics model can be used as a potential indicator of recurrence risk.

A cross-sectional study on the correlation of image quality, effective dose, and body composition with thyroid, chest, and abdominal computed tomography scans.

Background: The rapid increase in the use of radiodiagnostic examinations in China, especially computed tomography (CT) scans, has led to these examinations being the largest artificial source of per capita effective dose (ED). This study conducted a retrospective analysis of the correlation between image quality, ED, and body composition in 540 cases that underwent thyroid, chest, or abdominal CT scans. The aim of this analysis was to evaluate the correlation between the parameters of CT scans and body composition in common positions of CT examination (thyroid, chest, and abdomen) and ultimately inform potential measures for reducing radiation exposure. Methods: This study included 540 patients admitted to Fudan University Shanghai Cancer Center from January 2015 to December 2019 who underwent both thyroid or chest or abdominal CT scan and body composition examination. Average CT values and standard deviation (SD) values were collected for the homogeneous areas of the thyroid, chest, or abdomen, and the average CT values and SD values of adjacent subcutaneous fat tissue were measured in the same region of interest (ROI). All data were measured three times, and the average was taken to calculate the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) for each area. The dose-length product (DLP) was recorded, and the ED was calculated with the following: formula ED = k × DLP. Dual-energy X-ray was used to determine body composition and obtain indicators such as percentage of spinal and thigh muscle. Pearson correlation coefficient was used to analyze the correlations between body composition indicators, height, weight, body mass index (BMI), and ED. Results: The correlation coefficients between the SNR of abdominal CT scan and weight, BMI, and body surface area (BSA) were -0.470 (P=0.001), -0.485 (P=0.001), and -0.437 (P=0.002), representing a moderate correlation strength with statistically significant differences. The correlation coefficients between the ED of chest CT scans and weight, BMI, spinal fat percentage, and BSA were 0.488 (P=0.001), 0.473 (P=0.002), 0.422 (P=0.001), and 0.461 (P=0.003), respectively, indicating a moderate correlation strength with statistical differences. There was a weak statistically significant correlation between the SNR, CNR, and ED of the other scans with each physical and body composition index (P=0.023). Conclusions: There were varying degrees of correlation between CT image quality and ED and physical and body composition indices, which may inform novel solutions for reducing radiation exposure.

Development and validation of peritumoral vascular and intratumoral radiomics to predict pathologic complete responses to neoadjuvant chemotherapy in patients with triple-negative breast cancer.

BACKGROUND: To develop and validate a peritumoral vascular and intratumoral radiomics model to improve pretreatment predictions for pathologic complete responses (pCRs) to neoadjuvant chemoradiotherapy (NAC) in patients with triple-negative breast cancer (TNBC). METHODS: A total of 282 TNBC patients (93 in the primary cohort, 113 in the validation cohort, and 76 in The Cancer Imaging Archive [TCIA] cohort) were retrospectively included. The peritumoral vasculature on the maximum intensity projection (MIP) from pretreatment DCE-MRI was segmented by a Hessian matrix-based filter and then edited by a radiologist. Radiomics features were extracted from the tumor and peritumoral vasculature of the MIP images. The LASSO method was used for feature selection, and the k-nearest neighbor (k-NN) classifier was trained and validated to build a predictive model. The diagnostic performance was assessed using the ROC analysis. RESULTS: One hundred of the 282 patient (35.5%) with TNBC achieved pCRs after NAC. In predicting pCRs, the combined peritumoral vascular and intratumoral model (fusion model) yields a maximum AUC of 0.82 (95% confidence interval [CI]: 0.75, 0.88) in the primary cohort, a maximum AUC of 0.67 (95% CI: 0.57, 0.76) in the internal validation cohort, and a maximum AUC of 0.65 (95% CI: 0.52, 0.78) in TCIA cohort. The fusion model showed improved performance over the intratumoral model and the peritumoral vascular model, but not significantly (p > 0.05). CONCLUSION: This study suggested that combined peritumoral vascular and intratumoral radiomics model could provide a non-invasive tool to enable prediction of pCR in TNBC patients treated with NAC.

Artificial intelligence-based classification of breast lesion from contrast enhanced mammography: a multicenter study.

PURPOSE: The authors aimed to establish an artificial intelligence (AI)-based method for preoperative diagnosis of breast lesions from contrast enhanced mammography (CEM) and to explore its biological mechanism. MATERIALS AND METHODS: This retrospective study includes 1430 eligible patients who underwent CEM examination from June 2017 to July 2022 and were divided into a construction set (n=1101), an internal test set (n=196), and a pooled external test set (n=133). The AI model adopted RefineNet as a backbone network, and an attention sub-network, named convolutional block attention module (CBAM), was built upon the backbone for adaptive feature refinement. An XGBoost classifier was used to integrate the refined deep learning features with clinical characteristics to differentiate benign and malignant breast lesions. The authors further retrained the AI model to distinguish in situ and invasive carcinoma among breast cancer candidates. RNA-sequencing data from 12 patients were used to explore the underlying biological basis of the AI prediction. RESULTS: The AI model achieved an area under the curve of 0.932 in diagnosing benign and malignant breast lesions in the pooled external test set, better than the best-performing deep learning model, radiomics model, and radiologists. Moreover, the AI model has also achieved satisfactory results (an area under the curve from 0.788 to 0.824) for the diagnosis of in situ and invasive carcinoma in the test sets. Further, the biological basis exploration revealed that the high-risk group was associated with the pathways such as extracellular matrix organization. CONCLUSIONS: The AI model based on CEM and clinical characteristics had good predictive performance in the diagnosis of breast lesions.

Quantitative analysis from ultrafast dynamic contrast-enhanced breast MRI using population-based versus individual arterial input functions, and comparison with semi-quantitative analysis.

PURPOSE: To evaluate the value of inline quantitative analysis of ultrafast dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) using a population-based arterial input function (P-AIF) compared with offline quantitative analysis with an individual AIF (I-AIF) and semi-quantitative analysis for diagnosing breast cancer. METHODS: This prospective study included 99 consecutive patients with 109 lesions (85 malignant and 24 benign). Model-based parameters (Ktrans, kep, and ve) and model-free parameters (washin and washout) were derived from CAIPIRINHA-Dixon-TWIST-VIBE (CDTV) DCE-MRI. Univariate analysis and multivariate logistic regression analysis with forward stepwise covariate selection were performed to identify significant variables. The AUC and F1 score were assessed for semi-quantitative and two quantitative analyses. RESULTS: kep from inline quantitative analysis with P-AIF for diagnosing breast cancer provided an AUC similar to kep from offline quantitative analysis with I-AIF (0.782 vs 0.779, p = 0.954), higher compared to washin from semi-quantitative analysis (0.782 vs 0.630, p = 0.034). Furthermore, the inline quantitative analysis with P-AIF achieved the larger F1 score (0.920) compared with offline quantitative analysis with I-AIF (0.780) and semi-quantitative analysis (0.480). There were no statistically significant differences for kep values between the two quantitative analysis schemes (p = 0.944). CONCLUSION: The inline quantitative analysis with P-AIF from CDTV in characterizing breast lesions could offer similar diagnostic accuracy to offline quantitative analysis with I-AIF, and higher diagnostic accuracy to semi-quantitative analysis.

A modified diffusion-weighted magnetic resonance imaging-based model from the radiologist's perspective: improved performance in determining the surgical resectability of advanced high-grade serous ovarian cancer.

BACKGROUND: Complete resection of all visible lesions during primary debulking surgery is associated with the most favorable prognosis in patients with advanced high-grade serous ovarian cancer. An accurate preoperative assessment of resectability is pivotal for tailored management. OBJECTIVE: This study aimed to assess the potential value of a modified model that integrates the original 8 radiologic criteria of the Memorial Sloan Kettering Cancer Center model with imaging features of the subcapsular or diaphragm and mesenteric lesions depicted on diffusion-weighted magnetic resonance imaging and growth patterns of all lesions for predicting the resectability of advanced high-grade serous ovarian cancer. STUDY DESIGN: This study included 184 patients with high-grade serous ovarian cancer who underwent preoperative diffusion-weighted magnetic resonance imaging between December 2018 and May 2023 at 2 medical centers. The patient cohort was divided into 3 subsets, namely a study cohort (n=100), an internal validation cohort (n=46), and an external validation cohort (n=38). Preoperative radiologic evaluations were independently conducted by 2 radiologists using both the Memorial Sloan Kettering Cancer Center model and the modified diffusion-weighted magnetic resonance imaging-based model. The morphologic characteristics of the ovarian tumors depicted on magnetic resonance imaging were assessed as either mass-like or infiltrative, and transcriptomic analysis of the primary tumor samples was performed. Univariate and multivariate statistical analyses were performed. RESULTS: In the study cohort, both the scores derived using the Memorial Sloan Kettering Cancer Center (intraclass correlation coefficients of 0.980 and 0.959, respectively; both P<.001) and modified diffusion-weighted magnetic resonance imaging-based models (intraclass correlation coefficients of 0.962 and 0.940, respectively; both P<.001) demonstrated excellent intra- and interobserver agreement. The Memorial Sloan Kettering Cancer Center model (odds ratio, 1.825; 95% confidence interval, 1.390-2.395; P<.001) and the modified diffusion-weighted magnetic resonance imaging-based model (odds ratio, 1.776; 95% confidence interval, 1.410-2.238; P<.001) independently predicted surgical resectability. The modified diffusion-weighted magnetic resonance imaging-based model demonstrated improved predictive performance with an area under the curve of 0.867 in the study cohort and 0.806 and 0.913 in the internal and external validation cohorts, respectively. Using the modified diffusion-weighted magnetic resonance imaging-based model, patients with scores of 0 to 2, 3 to 4, 5 to 6, 7 to 10, and ≥11 achieved complete tumor debulking rates of 90.3%, 66.7%, 53.3%, 11.8%, and 0%, respectively. Most patients with incomplete tumor debulking had infiltrative tumors, and both the Memorial Sloan Kettering Cancer Center and the modified diffusion-weighted magnetic resonance imaging-based models yielded higher scores. The molecular differences between the 2 morphologic subtypes were identified. CONCLUSION: When compared with the Memorial Sloan Kettering Cancer Center model, the modified diffusion-weighted magnetic resonance imaging-based model demonstrated enhanced accuracy in the preoperative prediction of resectability for advanced high-grade serous ovarian cancer. Patients with scores of 0 to 6 were eligible for primary debulking surgery.

Multi-task reconstruction network for synthetic diffusion kurtosis imaging: Predicting neoadjuvant chemoradiotherapy response in locally advanced rectal cancer.

PURPOSE: To assess the feasibility and clinical value of synthetic diffusion kurtosis imaging (DKI) generated from diffusion weighted imaging (DWI) through multi-task reconstruction network (MTR-Net) for tumor response prediction in patients with locally advanced rectal cancer (LARC). METHODS: In this retrospective study, 120 eligible patients with LARC were enrolled and randomly divided into training and testing datasets with a 7:3 ratio. The MTR-Net was developed for reconstructing Dapp and Kapp images from apparent diffusion coefficient (ADC) images. Tumor regions were manually segmented on both true and synthetic DKI images. The synthetic image quality and manual segmentation agreement were quantitatively assessed. The support vector machine (SVM) classifier was used to construct radiomics models based on the true and synthetic DKI images for pathological complete response (pCR) prediction. The prediction performance for the models was evaluated by the receiver operating characteristic (ROC) curve analysis. RESULTS: The mean squared error (MSE), peak signal-to-noise ratio (PSNR), and structural similarity index measure (SSIM) for tumor regions were 0.212, 24.278, and 0.853, respectively, for the synthetic Dapp images and 0.516, 24.883, and 0.804, respectively, for the synthetic Kapp images. The Dice similarity coefficient (DSC), positive predictive value (PPV), sensitivity (SEN), and Hausdorff distance (HD) for the manually segmented tumor regions were 0.786, 0.844, 0.755, and 0.582, respectively. For predicting pCR, the true and synthetic DKI-based radiomics models achieved area under the curve (AUC) values of 0.825 and 0.807 in the testing datasets, respectively. CONCLUSIONS: Generating synthetic DKI images from DWI images using MTR-Net is feasible, and the efficiency of synthetic DKI images in predicting pCR is comparable to that of true DKI images.

Integrated multiomic profiling of breast cancer in the Chinese population reveals patient stratification and therapeutic vulnerabilities.

Molecular profiling guides precision treatment of breast cancer; however, Asian patients are underrepresented in publicly available large-scale studies. We established a comprehensive multiomics cohort of 773 Chinese patients with breast cancer and systematically analyzed their genomic, transcriptomic, proteomic, metabolomic, radiomic and digital pathology characteristics. Here we show that compared to breast cancers in white individuals, Asian individuals had more targetable AKT1 mutations. Integrated analysis revealed a higher proportion of HER2-enriched subtype and correspondingly more frequent ERBB2 amplification and higher HER2 protein abundance in the Chinese HR+HER2+ cohort, stressing anti-HER2 therapy for these individuals. Furthermore, comprehensive metabolomic and proteomic analyses revealed ferroptosis as a potential therapeutic target for basal-like tumors. The integration of clinical, transcriptomic, metabolomic, radiomic and pathological features allowed for efficient stratification of patients into groups with varying recurrence risks. Our study provides a public resource and new insights into the biology and ancestry specificity of breast cancer in the Asian population, offering potential for further precision treatment approaches.

Enhancing brain metastasis prediction in non-small cell lung cancer: a deep learning-based segmentation and CT radiomics-based ensemble learning model.

BACKGROUND: Brain metastasis (BM) is most common in non-small cell lung cancer (NSCLC) patients. This study aims to enhance BM risk prediction within three years for advanced NSCLC patients by using a deep learning-based segmentation and computed tomography (CT) radiomics-based ensemble learning model. METHODS: This retrospective study included 602 stage IIIA-IVB NSCLC patients, 309 BM patients and 293 non-BM patients, from two centers. Patients were divided into a training cohort (N = 376), an internal validation cohort (N = 161) and an external validation cohort (N = 65). Lung tumors were first segmented by using a three-dimensional (3D) deep residual U-Net network. Then, a total of 1106 radiomics features were computed by using pretreatment lung CT images to decode the imaging phenotypes of primary lung cancer. To reduce the dimensionality of the radiomics features, recursive feature elimination configured with the least absolute shrinkage and selection operator (LASSO) regularization method was applied to select the optimal image features after removing the low-variance features. An ensemble learning algorithm of the extreme gradient boosting (XGBoost) classifier was used to train and build a prediction model by fusing radiomics features and clinical features. Finally, Kaplan‒Meier (KM) survival analysis was used to evaluate the prognostic value of the prediction score generated by the radiomics-clinical model. RESULTS: The fused model achieved area under the receiver operating characteristic curve values of 0.91 ± 0.01, 0.89 ± 0.02 and 0.85 ± 0.05 on the training and two validation cohorts, respectively. Through KM survival analysis, the risk score generated by our model achieved a significant prognostic value for BM-free survival (BMFS) and overall survival (OS) in the two cohorts (P < 0.05). CONCLUSIONS: Our results demonstrated that (1) the fusion of radiomics and clinical features can improve the prediction performance in predicting BM risk, (2) the radiomics model generates higher performance than the clinical model, and (3) the radiomics-clinical fusion model has prognostic value in predicting the BMFS and OS of NSCLC patients.

Extent of surgical resection for radiologically subsolid T1N0 invasive lung adenocarcinoma: When is a wedge resection acceptable?

OBJECTIVE: To evaluate whether wedge resection (WR) was appropriate for the patients with peripheral T1 N0 solitary subsolid invasive lung adenocarcinoma. METHODS: Patients with peripheral T1N0 solitary subsolid invasive lung adenocarcinoma who received sublobar resection were retrospectively reviewed. Clinicopathologic characteristics, 5-year recurrence-free survival, and 5-year lung cancer-specific overall survival were analyzed. Cox regression model was used to elucidate risk factors for recurrence. RESULTS: Two hundred fifty-eight patients receiving WR and 1245 patients receiving segmentectomy were included. The mean follow-up time was 36.87 ± 16.21 months. Five-year recurrence-free survival following WR was 96.89% for patients with ground-glass nodule (GGN) ≤2 cm and 0.25< consolidation-to-tumor ratio (CTR) ≤0.5, not statistically different from 100% for those with GGN≤2 cm and CTR ≤0.25 (P = .231). The 5-year recurrence-free survival was 90.12% for patients with GGN between 2 and 3 cm and CTR ≤0.5, significantly lower than that of patients with GGN ≤2 cm and CTR ≤0.25 (P = .046). For patients with GGN≤2 cm and 0.25

Contrast-Enhanced Mammography Radiomics Analysis for Preoperative Prediction of Breast Cancer Molecular Subtypes.

BACKGROUND: Predicting breast cancer molecular subtypes can help guide individualised clinical treatment of patients who need the rational preoperative treatment. This study aimed to investigate the efficacy of preoperative prediction of breast cancer molecular subtypes by contrast-enhanced mammography (CEM) radiomic features. METHODS: This retrospective two-centre study included women with breast cancer who underwent CEM preoperatively between August 2016 and May 2022. We included 356 patients with 386 lesions, which were grouped into training (n = 162), internal test (n = 160) and external test sets (n = 64). Radiomics features were extracted from low-energy (LE) images and recombined (RC) images and selected. Three dichotomous tasks were established according to postoperative immunohistochemical results: Luminal vs. non-Luminal, human epidermal growth factor receptor (HER2)-enriched vs. non-HER2-enriched, and triple-negative breast cancer (TNBC) vs. non-TNBC. For each dichotomous task, the LE, RC, and LE+RC radiomics models were built by the support vector machine classifier. The prediction performance of the models was assessed by the area under the receiver operating characteristic curve (AUC). Then, the accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were calculated for the models. DeLong's test was utilised to compare the AUCs. RESULTS: Radiomics models based on CEM are valuable for predicting breast cancer molecular subtypes. The LE+RC model achieved the best performance in the test set. The LE+RC model predicted Luminal, HER2-enriched, and TNBC subtypes with AUCs of 0.93, 0.89, and 0.87 in the internal test set and 0.82, 0.83, and 0.69 in the external test set, respectively. In addition, the LE model performed more satisfactorily than the RC model. CONCLUSION: CEM radiomics features can effectively predict breast cancer molecular subtypes preoperatively, and the LE+RC model has the best predictive performance.

Components of the tumor immune microenvironment based on m-IHC correlate with prognosis and subtype of triple-negative breast cancer.

BACKGROUND AND AIM: The spatial distribution and interactions of cells in the tumor immune microenvironment (TIME) might be related to the different responses of triple-negative breast cancer (TNBC) to immunomodulators. The potential of multiplex IHC (m-IHC) in evaluating the TIME has been reported, but the efficacy is insufficient. We aimed to research whether m-IHC results could be used to reflect the TIME, and thus to predict prognosis and complement the TNBC subtyping system. METHODS: The clinical, imaging, and prognosis data for 86 TNBC patients were retrospectively reviewed. CD3, CD4, CD8, Foxp3, PD-L1, and Pan-CK markers were stained by m-IHC. Particular cell spatial distributions and interactions in the TIME were evaluated with the HALO multispectral analysis platform. Then, we calculated the prognostic value of components of the TIME and their correlations with TNBC transcriptomic subtypes and MRI radiomic features reflecting TNBC subtypes. RESULTS: The components of the TIME score were established by m-IHC and demonstrated positive prognostic value for TNBC (p = 0.0047, 0.039, <0.0001 for DMFS, RFS, and OS). The score was calculated from several indicators, including Treg% in the tumor core (TC) or stromal area (SA), PD-L1+ cell% in the SA, CD3 + cell% in the TC, and PD-L1+ /CD8+ cells in the invasive margin and SA. According to the TNBC subtyping system, a few TIME indicators were significantly different in different subtypes and significantly correlated with MRI radiomic features reflecting TNBC subtypes. CONCLUSION: We demonstrated that the m-IHC-based quantitative score and indicators related to the spatial distribution and interactions of cells in the TIME can aid in the accurate diagnosis of TNBC in terms of prognosis and classification.

Development and validation of CT-based radiomics deep learning signatures to predict lymph node metastasis in non-functional pancreatic neuroendocrine tumors: a multicohort study.

Background: Lymph node status is an important factor for the patients with non-functional pancreatic neuroendocrine tumors (NF-PanNETs) with respect to the surgical methods, prognosis, recurrence. Our aim is to develop and validate a combination model based on contrast-enhanced CT images to predict the lymph node metastasis (LNM) in NF-PanNETs. Methods: Retrospective data were gathered for 320 patients with NF-PanNETs who underwent curative pancreatic resection and CT imaging at two institutions (Center 1, n = 236 and Center 2, n = 84) between January 2010 and March 2022. RDPs (Radiomics deep learning signature) were developed based on ten machine-learning techniques. These signatures were integrated with the clinicopathological factors into a nomogram for clinical applications. The evaluation of the model's performance was conducted through the metrics of the area under the curve (AUC). Findings: The RDPs showed excellent performance in both centers with a high AUC for predicting LNM and disease-free survival (DFS) in Center 1 (AUC, 0.88; 95% CI: 0.84-0.92; DFS, p < 0.05) and Center 2 (AUC, 0.91; 95% CI: 0.85-0.97; DFS, p < 0.05). The clinical factors of vascular invasion, perineural invasion, and tumor grade were associated with LNM (p < 0.05). The combination nomogram showed better prediction capability for LNM (AUC, 0.93; 95% CI: 0.89-0.96). Notably, our model maintained a satisfactory predictive ability for tumors at the 2-cm threshold, demonstrating its effectiveness across different tumor sizes in Center 1 (≤2 cm: AUC, 0.90 and >2 cm: AUC, 0.86) and Center 2 (≤2 cm: AUC, 0.93 and >2 cm: AUC, 0.91). Interpretation: Our RDPs may have the potential to preoperatively predict LNM in NF-PanNETs, address the insufficiency of clinical guidelines concerning the 2-cm threshold for tumor lymph node dissection, and provide precise therapeutic strategies. Funding: This work was supported by JSPS KAKENHI Grant Number JP22K20814; the Rare Tumor Research Special Project of the National Natural Science Foundation of China (82141104) and Clinical Research Special Project of Shanghai Municipal Health Commission (202340123).

Hybrid deep multi-task learning radiomics approach for predicting EGFR mutation status of non-small cell lung cancer in CT images.

Objective.Epidermal growth factor receptor (EGFR) mutation genotyping plays a pivotal role in targeted therapy for non-small cell lung cancer (NSCLC). We aimed to develop a computed tomography (CT) image-based hybrid deep radiomics model to predict EGFR mutation status in NSCLC and investigate the correlations between deep image and quantitative radiomics features.Approach.First, we retrospectively enrolled 818 patients from our centre and 131 patients from The Cancer Imaging Archive database to establish a training cohort (N= 654), an independent internal validation cohort (N= 164) and an external validation cohort (N= 131). Second, to predict EGFR mutation status, we developed three CT image-based models, namely, a multi-task deep neural network (DNN), a radiomics model and a feature fusion model. Third, we proposed a hybrid loss function to train the DNN model. Finally, to evaluate the model performance, we computed the areas under the receiver operating characteristic curves (AUCs) and decision curve analysis curves of the models.Main results.For the two validation cohorts, the feature fusion model achieved AUC values of 0.86 ± 0.03 and 0.80 ± 0.05, which were significantly higher than those of the single-task DNN and radiomics models (allP< 0.05). There was no significant difference between the feature fusion and the multi-task DNN models (P> 0.8). The binary prediction scores showed excellent prognostic value in predicting disease-free survival (P= 0.02) and overall survival (P< 0.005) for validation cohort 2.Significance.The results demonstrate that (1) the feature fusion and multi-task DNN models achieve significantly higher performance than that of the conventional radiomics and single-task DNN models, (2) the feature fusion model can decode the imaging phenotypes representing NSCLC heterogeneity related to both EGFR mutation and patient NSCLC prognosis, and (3) high correlations exist between some deep image and radiomics features.