Development and validation of an interpretable delta radiomics-based model for predicting invasive ground-glass nodules in lung adenocarcinoma: a retrospective cohort study.

Background: Radiomics models based on computed tomography (CT) can be used to differentiate invasive ground-glass nodules (GGNs) in lung adenocarcinoma to help determine the optimal timing of GGN resection, improve the accuracy of prognostic prediction, and reduce unnecessary surgeries. However, general radiomics does not fully utilize follow-up data and often lacks model interpretation. Therefore, this study aimed to build an interpretable model based on delta radiomics to predict GGN invasiveness. Methods: A retrospective analysis was conducted on a set of 303 GGNs that were surgically resected and confirmed as lung adenocarcinoma in Shanghai Chest Hospital between September 2017 and August 2022. Delta radiomics and general radiomics features were extracted from preoperative follow-up CT scans and combined with clinical features for modeling. The performance of the delta radiomics-clinical model was compared to that of the radiomics-clinical model. Additionally, Shapley additive explanations (SHAP) was employed to interpret and visualize the model. Results: Two models were constructed using a combination of 34 radiomic features and 10 delta radiomic features, along with 14 clinical features. The radiomics-clinical model and the delta radiomics-clinical model exhibited area under the curve (AUC) of 0.986 [95% confidence interval (CI): 0.977-0.995] and 0.974 (95% CI: 0.959-0.987) in the training set, respectively, and 0.949 (95% CI: 0.908-0.978) and 0.927 (95% CI: 0.879-0.966) in the test set, respectively. The DeLong test of the two models showed no statistical significance (P=0.10) in the test set. SHAP was used to output a summary plot for global interpretation, which showed that preoperative mass, three-dimensional (3D) length, mean diameter, volume, mean CT value, and delta radiomics feature original_firstorder_RootMeanSquared were the relatively more important features in the model. Waterfall plots for local interpretation showed how each feature contributed to the prediction output of a given GGN. Conclusions: The delta radiomics-based model proved to be a helpful tool for predicting the invasiveness of GGNs in lung adenocarcinoma. This approach offers a precise, noninvasive alternative in informing clinical decision-making. Additionally, SHAP provided insightful and user-friendly interpretations and visualizations of the model, enhancing its clinical applicability.

A Radiological-Radiomics model for differentiation between minimally invasive adenocarcinoma and invasive adenocarcinoma less than or equal to 3 cm: A two-center retrospective study.

OBJECTIVE: To develop a Radiological-Radiomics (R-R) combined model for differentiation between minimal invasive adenocarcinoma (MIA) and invasive adenocarcinoma (IA) of lung adenocarcinoma (LUAD) and evaluate its predictive performance. METHODS: The clinical, pathological, and imaging data of a total of 509 patients (522 lesions) with LUAD diagnosed by surgical pathology from 2 medical centres were retrospectively collected, with 392 patients (402 lesions) from center 1 trained and validated using a five-fold cross-validation method, and 117 patients (120 lesions) from center 2 serving as an independent external test set. The least absolute shrinkage and selection operator (LASSO) method was utilized to filter features. Logistic regression was used to construct three models for predicting IA, namely, Radiological model, Radiomics model, and R-R model. Also, receiver operating curve curves (ROCs) were plotted, generating corresponding area under the curve (AUC), sensitivity, specificity, and accuracy. RESULTS: The R-R model for IA prediction achieved an AUC of 0.918 (95 % CI: 0.889-0.947), a sensitivity of 80.3 %, a specificity of 88.2 %, and an accuracy of 82.1 % in the training set. In the validation set, this model exhibited an AUC of 0.906 (95 % CI: 0.842-0.970), a sensitivity of 79.9 %, a specificity of 88.1 %, and an accuracy of 81.8 %. In the external test set, the AUC was 0.894 (95 % CI: 0.824-0.964), a sensitivity of 84.8 %, a specificity of 78.6 %, and an accuracy of 83.3 %. CONCLUSION: The R-R model showed excellent diagnostic performance in differentiating MIA and IA, which can provide a certain reference for clinical diagnosis and surgical treatment plans.

Global, regional, and national cancer burdens of respiratory and digestive tracts in 1990-2044: A cross-sectional and age-period-cohort forecast study.

BACKGROUND: Understanding the current status and future trends of cancer burdens by systems provides important information for specialists, policymakers, and specific risk populations. METHODS: The aim of this study was to compare the current and future cancer burdens of the gastrointestinal (GI) and respiratory tracts in terms of their magnitude and distribution. Data from a total of eight cancers of the digestive and respiratory tracts in the Global Burden of Disease (GBD) database were collected. The age-standardized incidence/death rates (ASIR/ASDRs), disability-adjusted life years (DALYs), and estimated annual percentage changes (EAPCs) were analyzed. Future trends were predicted with Bayesian age-period-cohort (BAPC) and NORDPRED models. RESULTS: In 2019, there was a significant increase in DALY for both digestive and respiratory tract cancers compared to 1990. Meanwhile, ASIR increased slightly and ASDR decreased notably. In 2019, the global cancer burdens of respiratory and digestive tracts were 38568363.53 and 66912328.72 in DALY, 34.28 and 55.32 in ASIR, and 656.82 and 808.22 in ASDR per 100,000 population with changes of +54.63% and +43.93%, +2.92% and +5.65%, and -17.39% and -26.83% compared to those in 1990, respectively. Significant cross-regional differences in the cancer burdens were observed among the regions. Compared to four representative chronic diseases, the burden of cancers showed less remission and greater global inequalities. The burdens of both digestive and respiratory tract cancers were higher in males than in females in terms of the ASIR, ASDR, and DALY. The incidence and mortality rates of respiratory tract cancers were up to 3-4 times higher in males than in females, whereas the difference between male and female rates of digestive tract cancers was relatively smaller. The main risk factor associated with all kinds of digestive and respiratory tract cancers is tobacco, leading to 18.5 in ASDR and 3.38×107 in DALY for respiratory tract cancers; 8.29 in ASDR and 1.60×107 in DALY for digestive tract cancers, in 2019. Additionally, alcohol use contributes to most digestive and respiratory tract cancers (1.23/1.03 in ASDR and 1.60×106/2.57×106 in DALY for respiratory tract cancers; 4.19/3.82 in ASDR and 4.49×106/8.06×106 in DALY for digestive tract cancers), except for stomach cancer and tracheal, bronchus, and lung cancer. The cancer burdens of respiratory and digestive tracts are likely to decrease substantially between 2020 and 2044. For most metrics, except for the ASIR and male-to-female ratios of ASDR and ASDALY in digestive tract cancers, the worldwide variances of burden metrics have been decreasing in the past decades and will possibly maintain stable trends in the future. CONCLUSIONS: The epidemiology of respiratory and GI tract cancers has common features and individual characteristics that are reflected in geography, age characteristics, and risk factors. Current epidemiological status, future trends, and the globalization of these disease burdens are important factors for making scientific planning of resources to minimize the cancer burden metrics and their cross-regional inequalities.

Combination of Contrast-enhanced FlAIR and Contrast-enhanced T1WI: A Quick and Efficient Method in Detecting Brain Metastases of Lung Cancers.

BACKGROUND: Some patients with suspected brain metastases (BM) could not tolerate longer scanning examinations according to the standardized MRI protocol. OBJECTIVE: The purpose of this study was to evaluate the clinical value of contrast-enhanced fast fluid-attenuated inversion recovery (CE FLAIR) imaging in combination with contrast-enhanced T1 weighted imaging (CE T1WI) in detecting BM of lung cancer and explore a quick and effective MRI protocol. MATERIAL AND METHODS: In 201 patients with lung cancers and suspected BM, T1WI and FLAIR were performed before and after administration of gadopentetate dimeglumine. Two radiologists reviewed pre- and post-contrast images to determine the presence of abnormal contrast enhancement or signal intensity and decided whether it was metastatic or not on CE T1WI (Group 1) and CE FLAIR (Group 2). The number, locations and features of abnormal findings in two groups were recorded. Receiver Operating Characteristic (ROC) analyses were conducted in three groups: Group 1, 2 and 3(combination of CE FLAIR and CE T1WI). RESULTS: A total of 714 abnormal findings were revealed, of which 672 were considered as BM and 42 nonmetastatic. Superficial and small metastases(≤10mm) in parenchyma and ependyma, leptomeningeal and non-expansive skull metastases were typically better seen on CE FLAIR. The areas under ROC in the three groups were 0.720,0.887 and 0.973, respectively. Group 3 was significantly better in diagnostic efficiency of BMs than Group 1 (p<0.0001) or Group 2 (p=0.0006). CONCLUSION: The combination of CE T1WI and CE FLAIR promotes diagnostic performance and results in better observation and characterization of BM in patients with lung cancers. It provides a quick and efficient way of detecting BM.

Non-linear modifications enhance prediction of pathological response to pre-operative PD-1 blockade in lung cancer: A longitudinal hybrid radiological model.

Major pathologic remission (MPR, residual tumor <10%) is a promising clinical endpoint for prognosis analysis in patients with lung cancer receiving pre-operative PD-1 blockade therapy. Most of the current biomarkers for predicting MPR such as PD-L1 and tumor mutation burden (TMB) need to be obtained invasively. They cannot overcome the spatiotemporal heterogeneity or provide dynamic monitoring solutions. Radiomics and artificial intelligence (AI) models provide a practical tool enabling non-invasive follow-up observation of tumor structural information through high-throughput data analysis. Currently, AI-based models mainly focus on the single baseline scan or pipeline, namely sole radiomics or deep learning (DL). This work merged the delta-radiomics based on the slope of classic radiomics indexes within a time interval and the features extracted by deep networks from the subtraction between the baseline and follow-up images. The subtracted images describing the tumor changes were based on the transformation generated by registration. Stepwise optimization of components was performed by repeating experiments among various combinations of DL networks, registration methods, feature selection algorithms, and classifiers. The optimized model could predict MPR with a cross-validation AUC of 0.91 and an external validation AUC of 0.85. A core set of 27 features (eight classic radiomics, 15 delta-radiomics, one classic DL features, and three delta-DL features) was identified. The changes in delta-radiomics indexes during the treatment were fitted with mathematic models. The fitting results revealed that over half of the features were of non-linear dynamics. Therefore, non-linear modifications were made on eight features by replacing the original features with non-linear fitting parameters, and the modified model achieved an improved power. The dynamic hybrid model serves as a novel and promising tool to predict the response of lesions to PD-1 blockade, which implies the importance of introducing the non-linear dynamic effects and DL approaches to the original delta-radiomics in the future.

Improving the prediction of Spreading Through Air Spaces (STAS) in primary lung cancer with a dynamic dual-delta hybrid machine learning model: a multicenter cohort study.

BACKGROUND: Reliable pre-surgical prediction of spreading through air spaces (STAS) in primary lung cancer is essential for precision treatment and surgical decision-making. We aimed to develop and validate a dual-delta deep-learning and radiomics model based on pretreatment computed tomography (CT) image series to predict the STAS in patients with lung cancer. METHOD: Six hundred seventy-four patients with pre-surgery CT follow-up scans (with a minimum interval of two weeks) and primary lung cancer diagnosed by surgery were retrospectively recruited from three Chinese hospitals. The training cohort and internal validation cohort, comprising 509 and 76 patients respectively, were selected from Shanghai Chest Hospital; the external validation cohorts comprised 36 and 53 patients from two other centers, respectively. Four imaging signatures (classic radiomics features and deep learning [DL] features, delta-radiomics and delta-DL features) reflecting the STAS status were constructed from the pretreatment CT images by comprehensive methods including handcrafting, 3D views extraction, image registration and subtraction. A stepwise optimized three-step procedure, including feature extraction (by DL and time-base radiomics slope), feature selection (by reproducibility check and 45 selection algorithms), and classification (32 classifiers considered), was applied for signature building and methodology optimization. The interpretability of the proposed model was further assessed with Grad-CAM for DL-features and feature ranking for radiomics features. RESULTS: The dual-delta model showed satisfactory discrimination between STAS and non-STAS and yielded the areas under the receiver operating curve (AUCs) of 0.94 (95% CI, 0.92-0.96), 0.84 (95% CI, 0.82-0.86), and 0.84 (95% CI, 0.83-0.85) in the internal and two external validation cohorts, respectively, with interpretable core feature sets and feature maps. CONCLUSION: The coupling of delta-DL model with delta-radiomics features enriches information such as anisotropy of tumor growth and heterogeneous changes within the tumor during the radiological follow-up, which could provide valuable information for STAS prediction in primary lung cancer.

Establishment and visualization of a model based on high-resolution CT qualitative and quantitative features for prediction of micropapillary or solid components in invasive lung adenocarcinoma.

OBJECTIVE: To predict the existence of micropapillary or solid components in invasive adenocarcinoma, a model was constructed using qualitative and quantitative features in high-resolution computed tomography (HRCT). METHODS: Through pathological examinations, 176 lesions were divided into two groups depending on the presence or absence of micropapillary and/or solid components (MP/S): MP/S- group (n = 128) and MP/S + group (n = 48). Multivariate logistic regression analyses were used to identify independent predictors of the MP/S. Artificial intelligence (AI)-assisted diagnostic software was used to automatically identify the lesions and extract corresponding quantitative parameters on CT images. The qualitative, quantitative, and combined models were constructed according to the results of multivariate logistic regression analysis. The receiver operating characteristic (ROC) analysis was conducted to evaluate the discrimination capacity of the models with the area under the curve (AUC), sensitivity, and specificity calculated. The calibration and clinical utility of the three models were determined using the calibration curve and decision curve analysis (DCA), respectively. The combined model was visualized in a nomogram. RESULTS: The multivariate logistic regression analysis using both qualitative and quantitative features indicated that tumor shape (P = 0.029 OR = 4.89; 95% CI 1.175-20.379), pleural indentation (P = 0.039 OR = 1.91; 95% CI 0.791-4.631), and consolidation tumor ratios (CTR) (P < 0.001; OR = 1.05; 95% CI 1.036-1.070) were independent predictors for MP/S + . The areas under the curve (AUC) of the qualitative, quantitative, and combined models in predicting MP/S + were 0.844 (95% CI 0.778-0.909), 0.863 (95% CI 0.803-0.923), and 0.880 (95% CI 0.824-0.937). The combined model of AUC was the most superior and statistically better than qualitative model. CONCLUSION: The combined model could assist doctors to evaluate patient's prognoses and devise personalized diagnostic and treatment protocols for patients.

Prediction of high-grade patterns of stage IA lung invasive adenocarcinoma based on high-resolution CT features: a bicentric study.

OBJECTIVES: This study aims to predict the high-grade pattern (HGP) of stage IA lung invasive adenocarcinoma (IAC) based on the high-resolution CT (HRCT) features. METHODS: The clinical, pathological, and HRCT imaging data of 457 patients (from bicentric) with pathologically confirmed stage IA IAC (459 lesions in total) were retrospectively analyzed. The 459 lesions were classified into high-grade pattern (HGP) (n = 101) and non-high-grade pattern (n-HGP) (n = 358) groups depending on the presence of HGP (micropapillary and solid) in pathological results. The clinical and pathological data contained age, gender, smoking history, tumor stage, pathological type, and presence or absence of tumor spread through air spaces (STAS). CT features consisted of lesion location, size, density, shape, spiculation, lobulation, vacuole, air bronchogram, and pleural indentation. The independent predictors for HGP were screened by univariable and multivariable logistic regression analyses. The clinical, CT, and clinical-CT models were constructed according to the multivariable analysis results. RESULTS: The multivariate analysis suggested the independent predictors of HGP, encompassing tumor size (p = 0.001; OR = 1.090, 95% CI 1.035-1.148), density (p < 0.001; OR = 9.454, 95% CI 4.911-18.199), and lobulation (p = 0.002; OR = 2.722, 95% CI 1.438-5.154). The AUC values of clinical, CT, and clinical-CT models for predicting HGP were 0.641 (95% CI 0.583-0.699) (sensitivity = 69.3%, specificity = 79.2%), 0.851 (95% CI 0.806-0.896) (sensitivity = 79.2%, specificity = 79.6%), and 0.852 (95% CI 0.808-0.896) (sensitivity = 74.3%, specificity = 85.8%). CONCLUSION: The logistic regression model based on HRCT features has a good diagnostic performance for the high-grade pattern of stage IA IAC. KEY POINTS: • The AUC values of clinical, CT, and clinical-CT models for predicting high-grade patterns were 0.641 (95% CI 0.583-0.699), 0.851 (95% CI 0.806-0.896), and 0.852 (95% CI 0.808-0.896). • Tumor size, density, and lobulation were independent predictive markers for high-grade patterns. • The logistic regression model based on HRCT features has a good diagnostic performance for the high-grade patterns of invasive adenocarcinoma.

Sublethal heat treatment enhances lactic acid uptake in macrophages via MCT1, leading to reduced paraspeckle formation and a subsequent decrease in macrophage pyroptosis.

Introduction: Heat ablation is one of the key modalities in treating liver cancer, yet the residual cancer tissues suffering sublethal heat treatment possess a potential for increased malignancy. This study conducts a comprehensive analysis of cellular dynamics, metabolic shifts, and macrophage polarization within the tumor microenvironment following sublethal heat treatment. Methods: We observed significant acidification in tumor cell supernatants, attributed to increased lactic acid production. The study focused on how this pH shift, crucial in tumor progression and resistance, influences macrophage polarization, especially towards the M2 phenotype known for tumor-promoting functions. We also examined the upregulation of MCT1 expression post sublethal heat treatment and its primary role in lactic acid transport. Results: Notably, the study found minimal disparity in MCT1 expression between hepatocellular carcinoma patients and healthy liver tissues, highlighting the complexity of cancer biology. The research further revealed an intricate relationship between lactic acid, MCT1, and the inhibition of macrophage pyroptosis, offering significant insights for therapeutic strategies targeting the tumor immune environment. Post sublethal heat treatment, a reduction in paraspeckle under lactic acid exposure was observed, indicating diverse cellular impacts. Additionally, PKM2 was identified as a key molecule in this context, with decreased levels after sublethal heat treatment in the presence of lactic acid. Discussion: Collectively, these findings illuminate the intertwined mechanisms of sublethal heat treatments, metabolic alterations, and immune modulation in the tumor milieu, providing a deeper understanding of the complex interplay in cancer biology and treatment.

Gastric duplication cyst with ectopic pancreas in a teenager successfully resected by endoscopic submucosal dissection.

BACKGROUND: Gastric duplication cyst associated with ectopic pancreas is rare and we aimed to alert clinician to this congenital anomaly. CASE PRESENTATION: A 15-year-old girl presented with intermittent vomiting. Gastroscopy showed a submucosal tumor with an approximate diameter of 40 mm in the anterior wall of the gastric antrum. The lesion had a central umbilication and was diagnosed preliminarily as gastric ectopic pancreas with pseudocyst formation on the basis of its appearance. However, computed tomographic scan showed a thick-walled cystic lesion with an enhanced outline of the cystic wall in the antrum of stomach, suggestive of duplication cyst. Serum amylase was normal. Endoscopic ultrasonography revealed a solid-cystic lesion; the solid portion were inhomogeneously mixed with echoes, and had indistinct border to muscularis propria; the cystic portion had echogenic internal mucosal layer and distinct border to muscularis propria. Endoscopic submucosal dissection (ESD) was suggested for the patient to relieve symptoms and diagnose the lesion definitely. The operation procedure was uneventful and the solid-cystic lesion was resected completely. Histopathologic examination revealed that the solid portion was ectopic pancreas, and the cystic portion was gastric duplication cyst. After resection, the patient discharged successfully and neither symptoms nor tumors recurred during the 9 months follow-up period. CONCLUSIONS: This is the first case of a solid-cystic lesion with central umbilication in the stomach diagnosed as gastric duplication cyst associated with ectopic pancreas. ESD could be an optional treatment to provide a definitive diagnosis.

A method for distinguishing benign and malignant pulmonary nodules based on 3D dual path network aided by K-means clustering analysis.

In the USA, there were about 1 |806 |590 new cancer cases in 2020, and 606 520 cancer deaths are expected to have occurred in 2021. Lung cancer has become the leading cause of death from cancer in both men and women (Siegel et al., 2020). Clinical studies show that the five-year survival rate of lung cancer patients after early diagnosis and treatment intervention can reach 80%, compared with that of patients having advanced lung cancer. Thus, the early diagnosis of lung cancer is a key factor to reduce mortality.

Establishment and validation of a radiological-radiomics model for predicting high-grade patterns of lung adenocarcinoma less than or equal to 3 cm.

Objective: We aimed to develop a Radiological-Radiomics (R-R) based model for predicting the high-grade pattern (HGP) of lung adenocarcinoma and evaluate its predictive performance. Methods: The clinical, pathological, and imaging data of 374 patients pathologically confirmed with lung adenocarcinoma (374 lesions in total) were retrospectively analyzed. The 374 lesions were assigned to HGP (n = 81) and non-high-grade pattern (n-HGP, n = 293) groups depending on the presence or absence of high-grade components in pathological findings. The least absolute shrinkage and selection operator (LASSO) method was utilized to screen features on the United Imaging artificial intelligence scientific research platform, and logistic regression models for predicting HGP were constructed, namely, Radiological model, Radiomics model, and R-R model. Also, receiver operating curve (ROC) curves were plotted on the platform, generating corresponding area under the curve (AUC), sensitivity, specificity, and accuracy. Using the platform, nomograms for R-R models were also provided, and calibration curves and decision curves were drawn to evaluate the performance and clinical utility of the model. The statistical differences in the performance of the models were compared by the DeLong test. Results: The R-R model for HGP prediction achieved an AUC value of 0.923 (95% CI: 0.891-0.948), a sensitivity of 87.0%, a specificity of 83.4%, and an accuracy of 84.2% in the training set. In the validation set, this model exhibited an AUC value of 0.920 (95% CI: 0.887-0.945), a sensitivity of 87.5%, a specificity of 83.3%, and an accuracy of 84.2%. The DeLong test demonstrated optimal performance of the R-R model among the three models, and decision curves validated the clinical utility of the R-R model. Conclusion: In this study, we developed a fusion model using radiomic features combined with radiological features to predict the high-grade pattern of lung adenocarcinoma, and this model shows excellent diagnostic performance. The R-R model can provide certain guidance for clinical diagnosis and surgical treatment plans, contributing to improving the prognosis of patients.

Integrative Serum Metabolic Fingerprints Based Multi-Modal Platforms for Lung Adenocarcinoma Early Detection and Pulmonary Nodule Classification.

Identification of novel non-invasive biomarkers is critical for the early diagnosis of lung adenocarcinoma (LUAD), especially for the accurate classification of pulmonary nodule. Here, a multiplexed assay is developed on an optimized nanoparticle-based laser desorption/ionization mass spectrometry platform for the sensitive and selective detection of serum metabolic fingerprints (SMFs). Integrative SMFs based multi-modal platforms are constructed for the early detection of LUAD and the classification of pulmonary nodule. The dual modal model, metabolic fingerprints with protein tumor marker neural network (MP-NN), integrating SMFs with protein tumor marker carcinoembryonic antigen (CEA) via deep learning, shows superior performance compared with the single modal model Met-NN (p < 0.001). Based on MP-NN, the tri modal model MPI-RF integrating SMFs, tumor marker CEA, and image features via random forest demonstrates significantly higher performance than the clinical models (Mayo Clinic and Veterans Affairs) and the image artificial intelligence in pulmonary nodule classification (p < 0.001). The developed platforms would be promising tools for LUAD screening and pulmonary nodule management, paving the conceptual and practical foundation for the clinical application of omics tools.

Terahertz meta-biosensor based on high-Q electrical resonance enhanced by the interference of toroidal dipole.

Electrical dipole resonances typically have low Q factor and broad resonant linewidth caused by strong free-space coupling with high radiative loss. Here, we present a strategy for enhancing the Q factor of the electrical resonance via the interference of a toroidal dipole. To validate such a strategy, a metasurface consisting of two resonators is designed that responsible to the electric and toroidal dipoles. According to constructive and destructive hybridizations of the two dipole modes, enhanced and decreased Q factors are found respectively for the two hybrid modes, compared to the one for the conventional electric dipole resonance. As a practical application of such high Q resonance, we further experimentally investigate the sensing performance of the metasurface biosensor by detecting the cell concentration of lung cancer cells (type A549). Moreover, through monitoring both resonance frequency and amplitude variation of the metasurface biosensor, the dielectric permittivity of the lung cancer cells is delicately estimated by the conjoint analysis of both simulated and measured results. Our proposed metasurface paves a promising way for the study of multipole interference in the field of nanophotonics and validates its effectiveness in biomedical sensing.

[Segmentation of ground glass pulmonary nodules using full convolution residual network based on atrous spatial pyramid pooling structure and attention mechanism].

Accurate segmentation of ground glass nodule (GGN) is important in clinical. But it is a tough work to segment the GGN, as the GGN in the computed tomography images show blur boundary, irregular shape, and uneven intensity. This paper aims to segment GGN by proposing a fully convolutional residual network, i.e., residual network based on atrous spatial pyramid pooling structure and attention mechanism (ResAANet). The network uses atrous spatial pyramid pooling (ASPP) structure to expand the feature map receptive field and extract more sufficient features, and utilizes attention mechanism, residual connection, long skip connection to fully retain sensitive features, which is extracted by the convolutional layer. First, we employ 565 GGN provided by Shanghai Chest Hospital to train and validate ResAANet, so as to obtain a stable model. Then, two groups of data selected from clinical examinations (84 GGN) and lung image database consortium (LIDC) dataset (145 GGN) were employed to validate and evaluate the performance of the proposed method. Finally, we apply the best threshold method to remove false positive regions and obtain optimized results. The average dice similarity coefficient (DSC) of the proposed algorithm on the clinical dataset and LIDC dataset reached 83.46%, 83.26% respectively, the average Jaccard index (IoU) reached 72.39%, 71.56% respectively, and the speed of segmentation reached 0.1 seconds per image. Comparing with other reported methods, our new method could segment GGN accurately, quickly and robustly. It could provide doctors with important information such as nodule size or density, which assist doctors in subsequent diagnosis and treatment.

Prediction of future imagery of lung nodule as growth modeling with follow-up computed tomography scans using deep learning: a retrospective cohort study.

Background: Risk prediction models of lung nodules have been built to alleviate the heavy interpretative burden on clinicians. However, the malignancy scores output by those models can be difficult to interpret in a clinically meaningful manner. In contrast, the modeling of lung nodule growth may be more readily useful. This study developed a CT-based visual forecasting system that can visualize and quantify a nodule in three dimensions (3D) in any future time point using follow-up CT scans. Methods: We retrospectively included 246 patients with 313 lung nodules with at least 1 follow-up CT scan. For the manually segmented nodules, we calculated geometric properties including CT value, diameter, volume, and mass, as well as growth properties including volume doubling time (VDT), and consolidation-to-tumor ratio (CTR) at follow-ups. These nodules were divided into growth and non-growth groups by thresholding their VDTs. We then developed a convolutional neural network (CNN) to model the imagery change of the nodules from baseline CT image (combined with the nodule mask) to follow-up CT image with a particular time interval. The model was evaluated on the geometric and radiological properties using either logistic regression or receiver operating characteristic (ROC) curve. Results: The lung nodules consisted of 115 ground glass nodules (GGN) and 198 solid nodules and were followed up for an average of 354 days with 2 to 11 scans. The 2 groups differed significantly in most properties. The prediction of our forecasting system was highly correlated with the ground truth with small relative errors regarding the four geometric properties. The prediction-derived VDTs had an area under the curve (AUC) of 0.857 and 0.843 in differentiating growth and non-growth nodules for GGN and solid nodules, respectively. The prediction-derived CTRs had an AUC of 0.892 in classifying high- and low-risk nodules. Conclusions: This proof-of-concept study demonstrated that the deep learning-based model can accurately forecast the imagery of a nodule in a given future for both GGNs and solid nodules and is worthy of further investigation. With a larger dataset and more validation, such a system has the potential to become a prognostication tool for assessing lung nodules.

Computed tomography versus frozen sections for distinguishing lung adenocarcinoma: A cohort study of concordance rate.

BACKGROUND: Computed tomography (CT) imaging can help to predict the pathological invasiveness of early-stage lung adenocarcinoma and guide surgical resection. This retrospective study investigated whether CT imaging could distinguish pre-invasive lung adenocarcinoma from IAC. It also compared final pathology prediction accuracy between CT imaging and intraoperative frozen section analysis. METHODS: This study included 2093 patients with early-stage peripheral lung adenocarcinoma who underwent CT imaging and intraoperative frozen section analysis between March 2013 and November 2014. Nodules were classified as ground-glass (GGNs), part-solid (PSNs), and solid nodules according to CT findings; they were classified as pre-IAC and IAC according to final pathology. Univariate, multivariate, and receiver operating characteristic (ROC) curve analyses were performed to evaluate whether CT imaging could distinguish pre-IAC from IAC. The concordance rates of CT imaging and intraoperative frozen section analyses with final pathology were also compared to determine their accuracies. RESULTS: Multivariate analysis identified tumor size as an independent distinguishing factor. ROC curve analyses showed that the optimal cut-off sizes for distinguishing pre-IAC from IAC for GGNs, PSNs, and solid nodules were 10.79, 11.48, and 11.45 mm, respectively. The concordance rate of CT imaging with final pathology was significantly greater than the concordance rate of intraoperative frozen section analysis with final pathology (P = 0.041). CONCLUSION: CT imaging could distinguish pre-IAC from IAC in patients with early-stage lung adenocarcinoma. Because of its accuracy in predicting final pathology, CT imaging could contribute to decisions associated with surgical extent. Multicenter standardized trials are needed to confirm the findings in this study.

Perfluorooctyl bromide nanoemulsions holding MnO2 nanoparticles with dual-modality imaging and glutathione depletion enhanced HIFU-eliciting tumor immunogenic cell death.

Tumor-targeted immunotherapy is a remarkable breakthrough, offering the inimitable advantage of specific tumoricidal effects with reduced immune-associated cytotoxicity. However, existing platforms suffer from low efficacy, inability to induce strong immunogenic cell death (ICD), and restrained capacity of transforming immune-deserted tumors into immune-cultivated ones. Here, an innovative platform, perfluorooctyl bromide (PFOB) nanoemulsions holding MnO2 nanoparticles (MBP), was developed to orchestrate cancer immunotherapy, serving as a theranostic nanoagent for MRI/CT dual-modality imaging and advanced ICD. By simultaneously depleting the GSH and eliciting the ICD effect via high-intensity focused ultrasound (HIFU) therapy, the MBP nanomedicine can regulate the tumor immune microenvironment by inducing maturation of dendritic cells (DCs) and facilitating the activation of CD8+ and CD4+ T cells. The synergistic GSH depletion and HIFU ablation also amplify the inhibition of tumor growth and lung metastasis. Together, these findings inaugurate a new strategy of tumor-targeted immunotherapy, realizing a novel therapeutics paradigm with great clinical significance.

Comparison of prophylactic percutaneous endoscopic gastrostomy with reactive enteral nutrition in patients with head and neck cancer undergoing radiotherapy or chemoradiotherapy: A systematic review.

BACKGROUND & AIMS: Nutrition support is frequently indicated in patients with head and neck cancer (HNC). However, the optimal timing of enteral tube placement and feeding commencement is unknown. This review aims to compare the outcomes for patients with HNC undergoing curative intent radiotherapy (RT) or chemoradiotherapy (CRT) receiving either prophylactic percutaneous endoscopic gastrostomy (pPEG) tube placement/feeding or reactive enteral nutrition (rEN). METHODS: A literature search was conducted in March 2020 across PubMed, CINAHL, Embase, Web of Science, and Scopus. Randomized controlled trials (RCTs) of patients (≥18 years) with HNC who had received either pPEG or rEN were included. Outcomes examined were weight change, nutritional status, body mass index, treatment interruptions, quality of life (QoL), disease-free survival and overall survival. Study quality and certainty of evidence were assessed using the Cochrane Risk-of-bias Tool for Randomized Trials Version 2 and the Grading of Recommendations Assessment, Development and Evaluation system, respectively. RESULTS: Five studies (three RCTs) (n = 298) were included and definitions of pPEG and rEN were heterogenous. pPEG was associated with a clinically important reduction in short-term critical weight loss (>10% weight loss), and significantly improved short-term QoL in patients with HNC. The timing of nutrition support commencement had no effect on all other outcomes. The overall certainty of evidence was 'moderate' for: nutritional status; treatment interruptions; short-term QoL; disease-free survival; and 'low' for all other outcomes. CONCLUSIONS: Patients with HNC undergoing RT or CRT receiving pPEG tube feeding/placement were less likely to experience short-term critical weight loss and have improved short-term QoL compared to rEN. Further well-designed RCTs with consistent definitions of tube feeding protocols and the use of validated tools to evaluate nutritional status, will assist to increase the certainty of evidence and confirm the beneficial effects observed.

A deep learning- and CT image-based prognostic model for the prediction of survival in non-small cell lung cancer.

OBJECTIVE: To assist clinicians in arranging personalized treatment, planning follow-up programs and extending survival times for non-small cell lung cancer (NSCLC) patients, a method of deep learning combined with computed tomography (CT) imaging for survival prediction was designed. METHODS: Data were collected from 484 patients from four research centers. The data from 344 patients were utilized to build the A_CNN survival prognosis model to classify 2-year overall survival time ranges (730 days cut-off). Data from 140 patients, including independent internal and external test sets, were utilized for model testing. First, a series of preprocessing techniques were used to process the original CT images and generate training and test data sets from the axial, coronal, and sagittal planes. Second, the structure of the A_CNN model was designed based on asymmetric convolution, bottleneck blocks, the uniform cross-entropy (UC) loss function, and other advanced techniques. After that, the A_CNN model was trained, and numerous comparative experiments were designed to obtain the best prognostic survival model. Last, the model performance was evaluated, and the predicted survival curves were analyzed. RESULTS: The A_CNN survival prognosis model yielded a high patient-level accuracy of 88.8%, a patch-level accuracy of 82.9%, and an area under the receiver operating characteristic (ROC) curve (AUC) of 0.932. When tested on an external data set, the maximum patient-level accuracy was 80.0%. CONCLUSIONS: The results suggest that using a deep learning method can improve prognosis in patients with NSCLC and has important application value in establishing individualized prognostic models.