Prediction Model of Ocular Metastases in Gastric Adenocarcinoma: Machine Learning-Based Development and Interpretation Study.

Background: Although gastric adenocarcinoma (GA) related ocular metastasis (OM) is rare, its occurrence indicates a more severe disease. We aimed to utilize machine learning (ML) to analyze the risk factors of GA-related OM and predict its risks. Methods: This is a retrospective cohort study. The clinical data of 3532 GA patients were collected and randomly classified into training and validation sets in a ratio of 7:3. Those with or without OM were classified into OM and non-OM (NOM) groups. Univariate and multivariate logistic regression analyses and least absolute shrinkage and selection operator were conducted. We integrated the variables identified through feature importance ranking and further refined the selection process using forward sequential feature selection based on random forest (RF) algorithm before incorporating them into the ML model. We applied six ML algorithms to construct the predictive GA model. The area under the receiver operating characteristic (ROC) curve indicated the model's predictive ability. Also, we established a network risk calculator based on the best performance model. We used Shapley additive interpretation (SHAP) to identify risk factors and to confirm the interpretability of the black box model. We have de-identified all patient details. Results: The ML model, consisting of 13 variables, achieved an optimal predictive performance using the gradient boosting machine (GBM) model, with an impressive area under the curve (AUC) of 0.997 in the test set. Utilizing the SHAP method, we identified crucial factors for OM in GA patients, including LDL, CA724, CEA, AFP, CA125, Hb, CA153, and Ca2+. Additionally, we validated the model's reliability through an analysis of two patient cases and developed a functional online web prediction calculator based on the GBM model. Conclusion: We used the ML method to establish a risk prediction model for GA-related OM and showed that GBM performed best among the six ML models. The model may identify patients with GA-related OM to provide early and timely treatment.

Single-Channel Blind Image Separation Based on Transformer-Guided GAN.

Blind source separation (BSS) has been a great challenge in the field of signal processing due to the unknown distribution of the source signal and the mixing matrix. Traditional methods based on statistics and information theory use prior information such as source distribution independence, non-Gaussianity, sparsity, etc. to solve this problem. Generative adversarial networks (GANs) learn source distributions through games without being constrained by statistical properties. However, the current blind image separation methods based on GANs ignores the reconstruction of the structure and details of the separated image, resulting in residual interference source information in the generated results. This paper proposes a Transformer-guided GAN guided by an attention mechanism. Through the adversarial training of the generator and the discriminator, U-shaped Network (UNet) is used to fuse the convolutional layer features to reconstruct the structure of the separated image, and Transformer is used to calculate the position attention and guide the detailed information. We validate our method with quantitative experiments, showing that it outperforms previous blind image separation algorithms in terms of PSNR and SSIM.

MLN4924 Promotes Self-Renewal of Limbal Stem Cells and Ocular Surface Restoration.

OBJECTIVE: To study the role of MLN4924 in corneal stem cell maintenance and corneal injury repair. METHODS: In cell experiments, the Sprague-Dawley (SD) rat corneal epithelial cells were co-cultured with mitomycin C-inactivated mouse feeder cells in a supplemental hormonal epithelial medium (SHEM) with or without MLN4924. Cells were photographed using an optical microscope. Furthermore, we performed crystal violet, polymerase chain reaction (PCR), and immunofluorescence staining on limbal stem cells (LSCs). In animal experiments, we scraped the corneal epithelium with a central corneal diameter of 4 mm in SD rats. The area of the corneal epithelial defect was calculated by fluorescein sodium staining. RESULTS: LSCs in the MLN4924 group had significantly proliferated. The MLN4924 treatment evidently enhanced the clone formation rate and clone area of LSCs. The expression levels of Ki67, p63, ABCG2, Bmi1, and C/EBPδ increased in LSCs after MLN4924 treatment, whereas the expression of K12 decreased. At 12 and 24 h after scraping, the corneal epithelium recovery rate in the eyes of the MLN4924-treated rats was accelerated. CONCLUSIONS: MLN4924 can maintain stemness, reduce differentiation, promote the proliferative capacity of rat LSCs, and accelerate corneal epithelial wound healing in SD rats.

TIPE2 Inhibits MGD Inflammation by Regulating Macrophage Polarization.

BACKGROUND: The aim of this study was to decide the role of the polarization of macrophages regulated by tumor necrosis factor-α (TNF-α)-induced protein 8-like 2 (TIPE2) in meibomian gland dysfunction (MGD). METHODS: Firstly, the secretory function of the meibomian gland (MG) in apolipoprotein E knockout (ApoE-/-) MGD mice and normal mice was detected by oil red staining. Then, the expression levels of markers of M1 and M2 macrophages were detected by immunofluorescence staining in MGD, normal mice, and mild and severe MGD corpses to decide the role of M1 and M2 macrophages in MGD inflammation. Meanwhile, the expression levels of TIPE2 in MGD mice and MGD patients were detected by immunofluorescence staining, and the correlations among TIPE2, M1 and M2 macrophages were analyzed by immunofluorescence double staining in MGD mice and MGD patients. Furthermore, lipopolysaccharide (LPS) and interleulkin-4 (IL-4) were used to induce M1 and M2 polarization of macrophages, and the mRNA level of TIPE2 was detected in M1 and M2 macrophages. RESULTS: Oil red staining showed that eyelid fat congestion was more severe in (ApoE-/-) MGD mice than in normal mice, and the M1 macrophage was the primary inflammatory cell infiltrated in (ApoE-/-) MGD mice (p < 0.05). The results of the immunofluorescence staining showed that the infiltration of macrophages in MGD mice was more obvious than that in the normal group, and M1 macrophage was the dominant group (p < 0.05). Similar to the results of the MGD mouse model, more macrophage infiltration was observed in MGD patients' MG tissues, and there were more M1 cells in the severe group than in the mild group (p < 0.05). Moreover, the expression of TIPE2 was positively correlated with the expression of M2 macrophages in MGD patients and mice MG tissues (p < 0.05). The expression of TIPE2 mRNA in LPS-induced M1 macrophages declined, while the expression of TIPE2 mRNA in IL-4-induced M2 macrophages increased (p < 0.05). CONCLUSION: M1 macrophage was the dominant group infiltrated in the MG tissue of MGD, and TIPE2 is a potential anti-inflammatory target for preventing the development of MGD by promoting the M2 polarization of macrophages.

Altered Intrinsic Brain Activity in Patients With Toothache Using the Percent Amplitude of a Fluctuation Method: A Resting-State fMRI Study.

Objective: The percent amplitude of fluctuation (PerAF) technique was utilized to evaluate the neural functions of specific cerebrum areas in patients with toothache (TA). Patients and Methods: An aggregation of 18 patients with TA (eight males and 10 females) were included in the study. We also recruited 18 healthy controls (HCs; eight men and 10 women) aligned for sex and age. Resting functional magnetic resonance imaging (rs-fMRI) scans were obtained. Then, we utilized the PerAF method and a support vector machine (SVM) to analyze the image data and measure neural abnormalities in related cerebrum areas. Receiver operating characteristic (ROC) curve analysis was utilized to appraise the two data sets. Results: The PerAF signals in the right dorsolateral superior frontal gyrus (RDSFG) and the right posterior central gyrus (RPCG) of TA sufferers were lower than HC signals. These results may reveal neural dysfunctions in relevant cerebrum regions. The AUC values of PerAF in the two areas were 0.979 in the RDSFG and 0.979 in the RPCG. The SVM results suggested that PerAF could be utilized to distinguish the TA group from HCs with a sensitivity of 75.00%, a specificity of 66.67%, and an accuracy of 70.83%. Conclusion: Patients with TA had marked differences in PerAF values in some regions of the cerebrum. Changes in PerAF values represented distinctions in blood oxygen level dependent semaphore intensity, which reflected the overactivity or inactivation of some cerebrum areas in those suffering from TA. At the same time, we analyzed the PerAF values of TAs with ROC curve, which can be helpful for the diagnosis of TA severity and subsequent treatment. Our results may help to elucidate the pathological mechanism of TA.

Altered Intrinsic Regional Spontaneous Brain Activity in Patients With Severe Obesity and Meibomian Gland Dysfunction: A Resting-State Functional Magnetic Resonance Imaging Study.

Purpose: To evaluate potential regional homogeneity (ReHo) cerebrum function lesions in people with severe obesity and meibomian gland dysfunction (SM) and probe the connection between aberrant cerebrum activity and clinical manifestations. Patients and Methods: An aggregation of 12 patients with SM, and 12 healthy controls (HCs) closely matched in age and gender were enrolled. We applied corneal confocal microscopy and fundus angiography to compare imaging distinctions between the two groups. SMs were required to carefully fill out the Hospital Anxiety Depression Scale (HADS) forms, and a correlation analysis was performed. ReHo was also utilized to appraise partial differences in spontaneous cerebrum function. Receiver operating characteristic (ROC) curves were created to partition ReHo values between patients with SM and the HCs. Results: ReHo values for the left cerebellum (LC), right fusiform gyrus (RFG), left inferior temporal gyrus (LITG), left rectus gyrus (LRG), right thalamus (RT), right caudate (RC), left insula (LI), and left thalamus (LT) of subjects with SM were notably higher than those of the HCs (P < 0.05). ReHo values of the right middle frontal gyrus (RMFG) in subjects with SM were decreased notably compared to the HCs (P < 0.05). ReHo values for the RMFG showed a negative correlation with the anxiety scores (ASs; r = -0.961, P < 0.001) and ReHo values for the RFG showed a positive correlation with the depression scores (DSs; r = 0.676, P = 0.016). The areas under the ROC curve were 1.000 (P < 0.001) for the RMFG, LC, LITG, LRG, RC, LI, and LT and 0.993 (P < 0.001) for the RFG and RT. The results from the ROC curve analysis indicated that changes in the ReHo values of some brain regions may help diagnose SM. Conclusion: Our research emphasized that patients with SM had lesions in synchronized neural activity in many encephalic areas. Our discoveries may provide beneficial information for exploring the neuromechanics of SM.

Decreased gray matter volume and increased white matter volume in patients with neovascular age-related macular degeneration: a voxel-based morphometry study.

OBJECTIVE: To measure white and gray matter volume (WMV, GMV) in patients with neovascular age-related macular degeneration (nAMD) using voxel-based morphometry (VBM). MATERIAL: Eighteen patients (9 men, 9 women) with nAMD and 18 (9 men, 9 women) healthy controls (HCs) aligned were recruited. Functional magnetic resonance imaging (fMRI) and VBM of three-dimensional T1 brain images were analyzed. And we also apply t-tests to look for GMV and WMV differences between groups. Correlation analysis was utilized to probe the connection between observational GMV and WMV values of diverse brain areas and the severity of HADS (hospital anxiety and depression scale). Also, distinctions between nAMD and HCs in GMV can be presented with the help of a ROC (receiver operating characteristic) curve. RESULTS: Compared with HCs, GMV values were significantly lower in patients with neovascular age-related macular degeneration in the right inferior frontal gyrus, temporal pole of left superior temporal gyrus, left superior temporal gyrus, left middle frontal gyrus, left anterior cingulate and para cingulate gyrus. While WMV was slightly increased in these areas. HADS (hospital anxiety and depression scale) scores portrayed a non-linear correlation with the GMV value of the right inferior frontal gyrus, left middle frontal gyrus, left anterior cingulate and paracingulate gyrus of the nAMD group (r=-0.6629, P=0.0027)(r=-0.9451, P<0.0001)(r=-0.6183, P=0.0062). Moreover, the ROC curve analysis of the mean VBM values for altered brain regions indicated high diagnostic accuracy. CONCLUSION: These results indicated that patients with nAMD have abnormal GMV and WMV and formed a basis for future research on pathological mechanisms in this disease. Moreover, decreased gray matter volume in particular brain regions might be associated with choroidal neovascularization and abnormal HADS score. It might help to explain the pathological mechanism of anxiety and depression in patients with nAMD.