Polarization images encode high resolution microstructural information even at low resolution. We propose a framework combining polarization imaging and traditional microscopy imaging, constructing a dualmodality machine learning framework that is not only accurate but also generalizable and interpretable. We demonstrate the viability of our proposed framework using the cervical intraepithelial neoplasia grading task, providing a polarimetry feature parameter to quantitatively characterize microstructural variations with lesion progression in hematoxylin-eosin-stained pathological sections of cervical precancerous tissues. By taking advantages of polarization imaging techniques and machine learning methods, the model enables interpretable and quantitative diagnosis of cervical precancerous lesion cases with improved sensitivity and accuracy in a lowresolution and wide-field system. The proposed framework applies routine image-analysis technology to identify the macro-structure and segment the target region in H&Estained pathological images, and then employs emerging polarization method to extract the micro-structure information of the target region, which intends to expand the boundary of the current image-heavy digital pathology, bringing new possibilities for quantitative medical diagnosis.

In this Letter, we report a study on the effects of spatial filtering for a transmission Mueller matrix imaging system. A spatial filter (SF) is placed on the back Fourier plane of the imaging lens in a dual-rotating-retarders Mueller matrix imaging system to select photons within a certain scattering angle. The system is then applied to three types of human cancerous tissues. When imaging with a small-aperture SF, some polarimetry basis parameters show sharp changes in contrast in the cancerous regions. Monte Carlo simulations using a simple sphere–cylinder scattering model also show that spatial filtering of the scattered photons provides extra information on the size and shape of the scattering particles. The results indicate that spatial filtering enhances the capability of polarization imaging as a powerful tool for biomedical diagnosis.

No abstract

Introduction: The incidence of cervical squamous cell carcinoma (CSCC) has expanded in recent years. However, the function of long non-coding RNA (lncRNA) MAGI2-AS3 in the occurrence and progression of CSCC remains unclear. Therefore, the role of lncRNA MAGI2-AS3 in cervical squamous cell carcinoma (CSCC) was investigated in our study. Methods: We used qRT-PCR analysis to identify the level of MAGI2-AS3 mRNA expression in CSCC clinical samples and cell lines. We investigated cell migration and invasion of CSCC cells transfected with MAGI2-AS3, miR-233 mimic, or EPB41L3 with transwell assays. Bioinformatics analysis and a luciferase reporter assay were employed to predict the interaction between MAGI2-AS3 and miR-233. Results: We found that MAGI2-AS3 and EPB41L3 were both downregulated in CSCC and the expression of this two was positively correlated. Bioinformatics analysis showed that MAGI2-AS3 might bind to miR-233, which could directly target EPB41L3. In CSCC cells, overexpression of MAGI2-AS3 led to upregulated, while overexpression of miRNA-233 led to downregulated expression of EPB41L3. However, MAGI2-AS3 and miR-233 did not affect the expression of each other. In addition, overexpression of MAGI2-AS3 and EPB41L3 led to inhibited cancer cell invasion and migration, while overexpression of miR-233 played an opposite role and attenuated the effects of overexpressing MAGI2-AS3. Conclusion: MAGI2-AS3 may sponge miR-233 to upregulate EPB41L3, thereby inhibiting CSCC cell invasion and migration.

We investigated the expression levels of nephroblastoma overexpressed (NOV, or CCN3 [cellular communication network factor 3]) in the serum and placenta of pregnant women and of pregnant mice fed a high-fat diet (HFD), and its effect on placental glucose transporter 3 (GLUT3) expression, to examine its role in gestational diabetes mellitus (GDM). NOV/CCN3 expression was increased in the mouse serum during pregnancy. At gestational day 18, NOV/CCN3 protein expression was increased in the serum and placenta of the HFD mice compared to that of mice fed a normal diet. Compared to non-GDM patients, the GDM patients had significantly increased serum NOV/CCN3 protein expression and placental NOV/CCN3 mRNA expression. Therefore, we hypothesized that NOV/CCN3 signaling may be involved in the pathogenesis of GDM. We administered NOV/CCN3 recombinant protein via intraperitoneal injections to pregnant mice fed HFD or normal diet. NOV/CCN3 overexpression led to glucose intolerance. Combined with the HFD, NOV/CCN3 exacerbated glucose intolerance and caused insulin resistance. NOV/CCN3 upregulates GLUT3 expression and affects the mammalian target of rapamycin (mTOR) pathway in the GDM environment in vivo and in vitro. In summary, our results demonstrate for the first time the molecular mechanism of NOV/CCN3 signaling in maternal metabolism to regulate glucose balance during pregnancy. NOV/CCN3 may be a potential target for detecting and treating GDM.