Clinical analysis of 23 cases of epidermoid cyst of testis in children.

OBJECTIVE: This study aims to examine the clinical characteristics and surgical management of pediatric testicular epidermoid cysts, thereby contributing to the existing body of knowledge pertinent to the diagnosis and therapeutic intervention s for this condition. METHODS: A retrospective analysis was conducted on the clinical records of 23 pediatric patients diagnosed with testicular epidermoid cysts, who were admitted to our institution between April 2013 and February 2024. Concurrently, a comprehensive review and analysis of pertinent literature were undertaken to augment the findings. RESULTS: The mean age at which the onset of epidermoid cysts was observed was 6.0 years. All cases were singular and unilateral. B-ultrasound diagnosis categorized 6 cases as epidermoid cysts, 11 as teratomas, and 6 as indeterminate, yielding a diagnostic sensitivity of 26.1%. All patients underwent testicle-sparing mass resection, and nine patients underwent rapid intraoperative frozen section analysis, revealing eight cases of testicular epidermoid cysts and one teratoma, with a diagnostic sensitivity of 88.89%. Postoperative histopathological examination confirmed the diagnosis of testicular epidermoid cyst. CONCLUSIONS: Pediatric testicular epidermoid cysts are an uncommon occurrence, primarily presenting as a painless scrotal mass, which can mimic the clinical features of malignant testicular tumors. Imaging modalities and histopathological assessment are pivotal in the diagnostic process for pediatric testicular epidermoid cysts. For cases where B-ultrasound is inconclusive, rapid intraoperative pathological examination should be considered.

Development and validation of competing risk nomograms for predicting cancer­specific mortality in non-metastatic patients with non­muscle invasive urothelial bladder cancer.

We aimed to assess the cumulative incidences of cancer-specific mortality (CSM) in non-metastatic patients with non­muscle invasive urothelial bladder cancer (NMIUBC) and establish competing risk nomograms to predict CSM. Patient data was sourced from the Surveillance, Epidemiology, and End Results database, as well as the electronic medical record system in our institution to form the external validation cohort. Sub-distribution proportional hazards model was utilized to determine independent risk factors influencing CSM in non-metastatic NMIUBC patients. Competitive risk nomograms were constructed to predict 3-year, 5-year, and 8-year cancer-specific survival (CSS) in all patients group, TURBT group and cystectomy group, respectively. The discrimination and accuracy of the model were validated through the concordance index (C-index), the area under the receiver operating characteristic curve (AUC), and calibration curves. Decision curve analysis (DCA) and a risk stratification system was employed to evaluate the clinical utility of the model. Race, age, marital status, surgery in other sites, tumor size, histological type, histological grade, T stage and N stage were identified as independent risk factors to predict CSS in all patients group. The C-index for 3-year CSS was 0.771, 0.770 and 0.846 in the training, testing and external validation sets, respectively. The ROC curves showed well discrimination and the calibration plots were well fitted and consistent. Moreover, DCA demonstrated well clinical effectiveness. Altogether, the competing risk nomogram displayed excellent discrimination and accuracy for predicting CSS in non-metastatic NMIUBC patients, which can be applied in clinical practice to help tailor treatment plans and make clinical decisions.

Effects of Nanocomposite Derivatives of Ni-Fe, Ni-Co, Ni-Co-Fe Prussian Blue Analogues on the Thermal Decomposition Performance of Nitrocellulose.

The incorporation of nanomaterials generated from Prussian blue (PB) derivatives has emerged as a promising strategy to significantly improve the properties of energetic materials. In this study, we comprehensively investigated the influence of nanomaterials derived from PB on the thermal decomposition characteristics of energetic materials. To achieve this goal, we prepared nanomaterials using coprecipitation and heat treatment methods with PB derivatives as catalysts. Advanced techniques such as X-ray diffraction, scanning electron microscopy, transmission electron microscopy, Brunauer-Emmett-Teller (BET) analysis for specific surface area and pore size, and X-ray photoelectron spectroscopy were employed to thoroughly characterize these nanomaterials. Differential scanning calorimetry was used to assess the thermal behavior of nitrocellulose (NC), and the relevant kinetic parameters were determined through thermal decomposition kinetics calculations and analysis. This work revealed the influence of catalysts on the NC decomposition process and provided comprehensive insights into the effect of integrating nanomaterials derived from PB derivatives on the thermal decomposition performance of NC. The results of this work demonstrated the possibility of using nanomaterials generated from PB derivatives as effective catalysts to enhance the thermal decomposition characteristics of NC, offering interesting opportunities for their application in the field of high-energy materials.

Tetrahedral DNA-linked aptamer-antibody-based sandwich-type electrochemical sensor with Ag@Au core-shell nanoparticles as a signal amplifier for highly sensitive detection of α-fetoprotein.

The conventional electrochemical detection strategy for alpha-fetoprotein (AFP) is limited by the antigen-antibody (Ag-Ab) reactions and suffers from low sensitivity and poor reproducibility due to the inconsistency of Ab-modified electrodes. Herein, we designed and explored a sandwich-type electrochemical sensor for highly sensitive detection of AFP based on aptamer (Apt)-AFP-Ab interaction mode with silver@gold (Ag@Au) core-shell nanoparticles (NPs) as a signal amplifier. AuNPs were electrodeposited onto MXene (Ti3C2TX)-modified glassy carbon electrode (GCE) to get AuNPs/MXene/GCE and further used as the signal amplification substrate. The tetrahedral DNA-linked AFP aptamers were immobilized onto AuNPs/MXene/GCE surface via Au-S bonds and used as the sensing and recognition platform for AFP capturing. Ag@AuNPs with core-shell structures were synthesized, characterized, and bound with Ab as detection elements by catalyzing H2O2 reduction. In the presence of AFP, a stable Apt-AFP-Ab sandwich structure was formed owing to the high affinities of aptamer and Ab toward the target AFP. The catalytic current produced by H2O2 reduction increased linearly with the logarithm of AFP concentration from 5 × 10-4 ng/mL to 1 × 105 ng/mL, accompanied by a low detection limit (1.6 × 10-4 ng/mL). Moreover, the novel sandwich-type electrochemical sensor shows high sensitivity, outstanding selectivity, and promising performance in the analysis of actual samples, displaying a broad application prospect in bioanalysis.

Pathogen resistance was negatively regulated by the NAC transcription factor ScATAF1 in sugarcane.

The NAC (NAM, ATAF, and CUC) is one of the largest transcription factor gene families in plants. In this study, 180, 141, and 131 NAC family members were identified from Saccharum complex, including S. officinarum, S. spontaneum, and Erianthus rufipilus. The Ka/Ks ratio of ATAF subfamily was all less than 1. Besides, 52 ATAF members from 12 representative plants were divided into three clades and there was only a significant expansion in maize. Surprisingly, ABA and JA cis-elements were abundant in hormonal response factor, followed by transcriptional regulator and abiotic stressor. The ATAF subfamily was differentially expressed in various tissues, under low temperature and smut pathogen treatments. Further, the ScATAF1 gene, with high expression in leaves, stem epidermis, and buds, was isolated. The encoded protein, lack of self-activation activity, was situated in the cell nucleus. Moreover, SA and JA stresses down-regulated the expression of this gene, while ABA, NaCl, and 4°C treatments led to its up-regulation. Interestingly, its expression in the smut susceptible sugarcane cultivars was much higher than the smut resistant ones. Notably, the colors presented slight brown in tobacco transiently overexpressing ScATAF1 at 1 d after DAB staining, while the symptoms were more obvious at 3 d after inoculation with Ralstonia solanacearum, with ROS, JA, and SA signaling pathway genes significantly up-regulated. We thus speculated ScATAF1 gene could negatively mediate hypersensitive reactions and produce ROS by JA and SA signaling pathways. These findings lay the groundwork for in-depth investigation on the biological roles of ATAF subfamily in sugarcane.

High-power electrically pumped terahertz topological laser based on a surface metallic Dirac-vortex cavity.

Topological lasers (TLs) have attracted widespread attention due to their mode robustness against perturbations or defects. Among them, electrically pumped TLs have gained extensive research interest due to their advantages of compact size and easy integration. Nevertheless, limited studies on electrically pumped TLs have been reported in the terahertz (THz) and telecom wavelength ranges with relatively low output powers, causing a wide gap between practical applications. Here, we introduce a surface metallic Dirac-vortex cavity (SMDC) design to solve the difficulty of increasing power for electrically pumped TLs in the THz spectral range. Due to the strong coupling between the SMDC and the active region, robust 2D topological defect lasing modes are obtained. More importantly, enough gain and large radiative efficiency provided by the SMDC bring in the increase of the output power to a maximum peak power of 150 mW which demonstrates the practical application potential of electrically pumped TLs.

Visual detection and discrimination of nerve and blood agents using a dual-site fluorescent probe in living cells and mice.

Of all chemical warfare agents (CWAs), only nerve and blood agents cause massive mortality at low concentrations. To better detect and discriminate nerve and blood agents, a reliable detection method is desirable. We report a series of fluorescent probes for nerve and blood agent detection. Among the tested probes, SR-Pip detected nerve and blood agents quickly (within 10 s for nerve agents and 1 min for blood agents). SR-Pip coupled with nerve agent produced a weak orange fluorescence with good sensitivity [limit of detection (LOD)= 5.5 µM]. Upon reaction with blood agent, the fluorescence of SR-Pip changed from orange fluorescence to blue fluorescence with detection limits as low as 9.6 nM. This probe effectively visualised different concentrations of nerve agents in living cells and mice. A portable test kit using SR-Pip instantly detected nerve and blood agents. To the best of our knowledge, SR-Pip is the first fluorescent probe for nerve and blood agent detection.

Chemoenzymatic Dynamic Kinetic Resolution Protocol with an Immobilized Oxovanadium as a Racemization Catalyst.

An excellent compatible and cost-effective dynamic kinetic resolution (DKR) protocol has been developed by combining a novel immobilized oxovanadium racemization catalyst onto cheap diatomite (V-D) with an immobilized lipase LA resolution catalyst onto a macroporous resin (LA-MR). V-D was prepared via grinding immobilization, which may become a promising alternative for the immobilization of metals, especially precious metals due to its low cost, high efficiency, easy separation, and large reaction interface. The DKR afforded high yield (96.1%), e.e. (98.67%), and Sel (98.28%) under optimal conditions established using response surface methodology as follows: the amount of V-D 10.83 mg, reaction time 51.2 h, and temperature 48.1 °C, respectively, indicating that all the reactions in the DKR were coordinated very well. The DKR protocol was also found to have high stability up to six reuses. V-D exhibited excellent compatibility with LA-MR because the lipase immobilized onto MR did not physically contact with the vanadium species immobilized onto diatomite, thus avoiding inactivation. Considering that lipase, oxovanadium, diatomite, and MR used are relatively inexpensive, and the adsorption or grinding immobilization is simple, the LA-V-MD DKR by coupling LA-MR with V-D is a cost-effective and promising protocol for chiral secondary alcohols.

Prognostic Significance of the Highest Mediastinal Lymph Node Involvement in Patients with Stage III-N2 Non-small Cell Lung Cancer.

BACKGROUND: Highest mediastinal lymph node (HMLN) involvement is a category of uncertain resection, yet the prognostic significance of HMLN involvement remains controversial. METHODS: A total of 486 patients with pathological stage III-N2 disease who underwent radical resection were enrolled from January 2015 to December 2018. Patients were allocated into two groups-HMLN involvement (219 cases) and HMLN-negative (249 cases) groups. Kaplan-Meier analysis and Cox proportional hazard regression models were used to evaluate the impact of HMLN involvement on 5-year recurrence-free survival (RFS) and overall survival (OS). RESULTS: The proportion of patients with multiple N2 diseases (72.1% vs. 23.7%; p < 0.001) and stage IIIA (87.2% vs. 77.5%; p < 0.009) were greater in the HMLN-involvement group than in the HMLN-negative group, and the survival rates of the HMLN-involvement group were significantly lower than those of the HMLN-negative group (RFS: 27.2% vs. 49.8%, p < 0.001; OS: 42.1% vs. 59.2%, p = 0.001). HMLN status was an independent factor for OS only (RFS: adjusted hazard ratio [aHR] 1.26, 95% confidence interval CI 0.94-1.68; OS: aHR 1.45, 95% CI 1.07-1.99) in the entire stage III cohort. After stratification of patients according to stage, the involvement of HMLN decreased both RFS and OS in the stage IIIA group (RFS: aHR 1.46, 95% CI 1.06-2.02; OS: aHR 1.70, 95% CI 1.19-2.42); however, no such difference was observed within the stage IIIB group. CONCLUSIONS: HMLN involvement is a prognostic factor of deteriorating survival in highly advanced N2 disease only in patients with stage IIIA.

Establishment and validation of nomograms to predict the overall survival and cancer-specific survival for non-metastatic bladder cancer patients: A large population-based cohort study and external validation.

This study aimed to develop nomograms to accurately predict the overall survival (OS) and cancer-specific survival (CSS) of non-metastatic bladder cancer (BC) patients. Clinicopathological information of 260,412 non-metastatic BC patients was downloaded from the Surveillance, Epidemiology, and End Results (SEER) database from 2000 to 2020. LASSO method and Cox proportional hazard regression analysis were utilized to discover the independent risk factors, which were used to develop nomograms. The accuracy and discrimination of models were tested by the consistency index (C-index), the area under the subject operating characteristic curve (AUC) and the calibration curve. Decision curve analysis (DCA) was used to test the clinical value of nomograms compared with the TNM staging system. Nomograms predicting OS and CSS were constructed after identifying independent prognostic factors. The C-index of the training, internal validation and external validation cohort for OS was 0.722 (95%CI: 0.720-0.724), 0.723 (95%CI: 0.721-0.725) and 0.744 (95%CI: 0.677-0.811). The C-index of the training, internal validation and external validation cohort for CSS was 0.794 (95%CI: 0.792-0.796), 0.793 (95%CI: 0.789-0.797) and 0.879 (95%CI: 0.814-0.944). The AUC and the calibration curves showed good accuracy and discriminability. The DCA showed favorable clinical potential value of nomograms. Kaplan-Meier curve and log-rank test uncovered statistically significance survival difference between high- and low-risk groups. We developed nomograms to predict OS and CSS for non-metastatic BC patients. The models have been internally and externally validated with accuracy and discrimination and can assist clinicians to make better clinical decisions.

Machine Learning Supported the Modified Gustafson's Criteria for Dental Age Estimation in Southwest China.

Adult age estimation is one of the most challenging problems in forensic science and physical anthropology. In this study, we aimed to develop and evaluate machine learning (ML) methods based on the modified Gustafson's criteria for dental age estimation. In this retrospective study, a total of 851 orthopantomograms were collected from patients aged 15 to 40 years old. The secondary dentin formation (SE), periodontal recession (PE), and attrition (AT) of four mandibular premolars were analyzed according to the modified Gustafson's criteria. Ten ML models were generated and compared for age estimation. The partial least squares regressor outperformed other models in males with a mean absolute error (MAE) of 4.151 years. The support vector regressor (MAE = 3.806 years) showed good performance in females. The accuracy of ML models is better than the single-tooth model provided in the previous studies (MAE = 4.747 years in males and MAE = 4.957 years in females). The Shapley additive explanations method was used to reveal the importance of the 12 features in ML models and found that AT and PE are the most influential in age estimation. The findings suggest that the modified Gustafson method can be effectively employed for adult age estimation in the southwest Chinese population. Furthermore, this study highlights the potential of machine learning models to assist experts in achieving accurate and interpretable age estimation.

The role of IL-36 and 37 in hepatocellular carcinoma.

Hepatocellular carcinoma (HCC) has garnered considerable attention due to its morbidity and mortality. Although the precise mechanisms underlying HCC tumorigenesis remain to be elucidated, evidence suggests that host immunity plays a pivotal role in its development. IL-36 and IL-37 are important immunoregulatory cytokines classified as pro-inflammatory and anti-inflammatory respectively. In the context of HCC, the downregulation of intrahepatic IL-36 is inversely correlated with cirrhosis, but positively correlated with 5-year survival rates, suggesting that IL-36 offers protection during HCC development. However, IL-36 may lose its hepatoprotective effects as the disease progresses to HCC in the context of dysregulated immunity in cirrhotic patients. Substantially increased circulating IL-36 in HCC patients is likely a systemic response to HCC stimulation, but is insufficient to suppress progression towards HCC. Intrahepatic IL-37 is suppressed in HCC patients, consistent with the inverse correlation between intrahepatic IL-37 and the level of AFP in HCC patients, suggesting IL-37 exerts hepatoprotection. There is no significant difference in IL-37 among differentiations of HCC or with respect to clinical BCLC stages or cirrhosis status in HCC patients. However, IL-37 protection is demonstrated in an IL-37 transfected HCC animal model, showing significantly reduced tumour size. IL-36/37 may inhibit HCC by enhancing M1 tumour-associated macrophages while not affecting M2 macrophages. The interplay between IL-36 (pro-inflammatory) and IL-37 (anti-inflammatory) is emerging as a crucial factor in host protection against the development of HCC. Further research is needed to investigate the complex mechanisms involved and the therapeutic potential of targeting these cytokines in HCC management.

Silk fibroin/methacrylated gelatine/hydroxyapatite biomimetic nanofibrous membranes for guided bone regeneration.

By mimicking in vivo bionic microenvironment and promoting osteogenic differentiation, the hybrid organic-inorganic nanofibrous membranes provide promising potential for guided bone regeneration (GBR) in the treatment of clinical bone defects. To develop a degradable and osteogenic membrane for GBR by combining the natural biomacromolecule silk fibroin (SF) and gelatine with the bioactive nano hydroxyapatite (nHA), the anhydride-modified gelatine-nano hydroxyapatite (GelMA-nHA) composites were synthesized in situ and introduced into silk fibroin to prepare nanofibrous membranes with different ratios using electrospinning and photocrosslinking. The nanofibrous membranes, particularly those with a mass ratio of 7:2:1, were found to exhibit satisfactory elongation at break up to 110 %, maintain the nanofibrous structure for up to 28 days, and rapidly form bone-like apatite within 3 days, thus offering advantages when it comes to guided bone regeneration. In vitro cell results showed that the SF/GelMA/nHA membranes had excellent biocompatibility and enhanced osteogenic differentiation of hBMSCs. In vivo studies revealed that the hybrid composite membranes can improve bone regeneration of critical-sized calvarial defects in rat model. Therefore, the novel hybrid nanofibrous membrane is proposed to be a alternative candidate for creating a bionic microenvironment that promotes bone regeneration, indicating their potential application to bone injury treatment.

Adult age estimation from the sternum using maximum intensity projection images of CT and data mining in a Chinese population.

This study aimed to explore and develop data mining models for adult age estimation based on CT reconstruction images from the sternum. Maximum intensity projection (MIP) images of chest CT were retrospectively collected from a modern Chinese population, and data from 2700 patients (1349 males and 1351 females) aged 20 to 70 years were obtained. A staging technique within four indicators was applied. Several data mining models were established, and mean absolute error (MAE) was the primary comparison parameter. The intraobserver and interobserver agreement levels were good. Within internal validation, the optimal data mining model obtained the lowest MAE of 9.08 in males and 10.41 in females. For the external validation (N = 200), MAEs were 7.09 in males and 7.15 in females. In conclusion, the accuracy of our model for adult age estimation was among similar studies. MIP images of the sternum could be a potential age indicator. However, it should be combined with other indicators since the accuracy level is still unsatisfactory.

NR_103776.1 as a novel diagnostic biomarker for systemic lupus erythematosus.

BACKGROUND: With the development of sequencing technologies, there is increasing evidence that long noncoding RNAs (lncRNAs) are involved in systemic lupus erythematosus (SLE). The level of NR_103776.1 expression in SLE and its clinical associations are still not well defined. OBJECTIVE: To identify differentially expressed lncRNAs and explore their functional roles in SLE. METHODS: Transcriptome sequencing was used to screen differentially expressed lncRNAs and mRNAs. Expression validation of clinical samples was performed by QRT-PCR. Bioinformatics was used to analyze its prognostic value and potential function. RESULTS: Of the 231 significantly differentially expressed lncRNAs, NR_103776.1 could be used to distinguish not only SLE patients and rheumatoid arthritis patients but also active SLE patients, stable SLE patients, and healthy controls. NR_103776.1 was significantly and negatively correlated with inflammatory indexes (CRP and ESR). NR_103776.1 dysregulation might contribute to the metabolism of RNA and proteins in SLE patients. CONCLUSIONS: This study not only provided a transcriptome profile of lncRNAs aberrantly expressed in individual nucleated cells of SLE patients but also suggested NR_103776.1 as a novel potential diagnostic biomarker.

Machine learning and deep learning enabled age estimation on medial clavicle CT images.

The medial clavicle epiphysis is a crucial indicator for bone age estimation (BAE) after hand maturation. This study aimed to develop machine learning (ML) and deep learning (DL) models for BAE based on medial clavicle CT images and evaluate the performance on normal and variant clavicles. This study retrospectively collected 1049 patients (mean± SD: 22.50±4.34 years) and split them into normal training and test sets, and variant training and test sets. An additional 53 variant clavicles were incorporated into the variant test set. The development stages of normal MCE were used to build a linear model and support vector machine (SVM) for BAE. The CT slices of MCE were automatically segmented and used to train DL models for automated BAE. Comparisons were performed by linear versus ML versus DL, and normal versus variant clavicles. Mean absolute error (MAE) and classification accuracy was the primary parameter of comparison. For BAE, the SVM had the best MAE of 1.73 years, followed by the commonly-used CNNs (1.77-1.93 years), the linear model (1.94 years), and the hybrid neural network CoAt Net (2.01 years). In DL models, SE Net 18 was the best-performing DL model with similar results to SVM in the normal test set and achieved an MAE of 2.08 years in the external variant test. For age classification, all the models exhibit superior performance in the classification of 18-, 20-, 21-, and 22-year thresholds with limited value in the 16-year threshold. Both ML and DL models produce desirable performance in BAE based on medial clavicle CT.

Automated bone age assessment from knee joint by integrating deep learning and MRI-based radiomics.

Bone age assessment (BAA) is a crucial task in clinical, forensic, and athletic fields. Since traditional age estimation methods are suffered from potential radiation damage, this study aimed to develop and evaluate a deep learning radiomics method based on multiparametric knee MRI for noninvasive and automatic BAA. This retrospective study enrolled 598 patients (age range,10.00-29.99 years) who underwent MR examinations of the knee joint (T1/T2*/PD-weighted imaging). Three-dimensional convolutional neural networks (3D CNNs) were trained to extract and fuse multimodal and multiscale MRI radiomic features for age estimation and compared to traditional machine learning models based on hand-crafted features. The age estimation error was greater in individuals aged 25-30 years; thus, this method may not be suitable for individuals over 25 years old. In the test set aged 10-25 years (n = 95), the 3D CNN (a fusion of T1WI, T2*WI, and PDWI) demonstrated the lowest mean absolute error of 1.32 ± 1.01 years, which is higher than that of other MRI modalities and the hand-crafted models. In the classification for 12-, 14-, 16-, and 18- year thresholds, accuracies and the areas under the ROC curves were all over 0.91 and 0.96, which is similar to the manual methods. Visualization of important features showed that 3D CNN estimated age by focusing on the epiphyseal plates. The deep learning radiomics method enables non-invasive and automated BAA from multimodal knee MR images. The use of 3D CNN and MRI-based radiomics has the potential to assist radiologists or medicolegists in age estimation.

Predicting micropapillary or solid pattern of lung adenocarcinoma with CT-based radiomics, conventional radiographic and clinical features.

BACKGROUND: To build prediction models with radiomics features, clinical/conventional radiographic signs and combined scores for the discrimination of micropapillary or solid subtypes (high-risk subtypes) of lung adenocarcinoma. METHODS: This retrospective study enrolled 351 patients with and without high-risk subtypes. Least Absolute Shrinkage and Selection Operator (LASSO) regression with cross-validation was performed to determine the optimal features of radiomics model. Missing clinical data were imputed by Multiple Imputation with Chain Equations (MICE). Clinical model with radiographic signs was built and scores of both models were integrated to establish combined model. Receiver operating characteristics (ROC) curves, area under ROC curves and decision curve analysis (DCA) were plotted to evaluate the model performance and clinical application. RESULTS: Stratified splitting allocated 246 patients into training set. MICE for missing values obtained complete and unbiased data for the following analysis. Ninety radiomic features and four clinical/conventional radiographic signs were used to predict the high-risk subtypes. The radiomic model, clinical model and combined model achieved AUCs of 0.863 (95%CI: 0.817-0.909), 0.771 (95%CI: 0.713-0.713) and 0.872 (95%CI: 0.829-0.916) in the training set, and 0.849 (95%CI: 0.774-0.924), 0.778 (95%CI: 0.687-0.868) and 0.853 (95%CI: 0.782-0.925) in the test set. Decision curve showed that the radiomic and combined models were more clinically useful when the threshold reached 37.5%. CONCLUSIONS: Radiomics features could facilitate the prediction of subtypes of lung adenocarcinoma. A simple combination of radiomics and clinical scores generated a robust model with high performance for the discrimination of micropapillary or solid subtype of lung adenocarcinoma.