Deep Learning Model for Static Ocular Torsion Detection Using Synthetically Generated Fundus Images.

Purpose: The objective of the study is to develop deep learning models using synthetic fundus images to assess the direction (intorsion versus extorsion) and amount (physiologic versus pathologic) of static ocular torsion. Static ocular torsion assessment is an important clinical tool for classifying vertical ocular misalignment; however, current methods are time-intensive with steep learning curves for frontline providers. Methods: We used a dataset (n = 276) of right eye fundus images. The disc-foveal angle was calculated using ImageJ to generate synthetic images via image rotation. Using synthetic datasets (n = 12,740 images per model) and transfer learning (the reuse of a pretrained deep learning model on a new task), we developed a binary classifier (intorsion versus extorsion) and a multiclass classifier (physiologic versus pathologic intorsion and extorsion). Model performance was evaluated on unseen synthetic and nonsynthetic data. Results: On the synthetic dataset, the binary classifier had an accuracy and area under the receiver operating characteristic curve (AUROC) of 0.92 and 0.98, respectively, whereas the multiclass classifier had an accuracy and AUROC of 0.77 and 0.94, respectively. The binary classifier generalized well on the nonsynthetic data (accuracy = 0.94; AUROC = 1.00). Conclusions: The direction of static ocular torsion can be detected from synthetic fundus images using deep learning methods, which is key to differentiate between vestibular misalignment (skew deviation) and ocular muscle misalignment (superior oblique palsies). Translational Relevance: Given the robust performance of our models on real fundus images, similar strategies can be adopted for deep learning research in rare neuro-ophthalmologic diseases with limited datasets.

Fenretinide combines perturbation of signaling kinases, cell-extracellular matrix interactions and matrix metalloproteinase activation to inhibit invasion in oral squamous cell carcinoma cells.

Basement membrane invasion defines malignant transformation of surface premalignancy. Treatment of oral squamous cell carcinoma (OSCC) cells with the synthetic vitamin A derivative, fenretinide (4HPR), induces numerous cancer-preventive effects including suppression of basement membrane invasion, elimination of anchorage-independent growth, disruption of actin cytoskeletal components and inhibition of the invasion-enabling focal adhesive kinase. The purpose of this study was to elucidate 4HPR's effects on additional invasion-relevant mechanisms including matrix metalloproteinase (MMP) activation and function, cell-extracellular matrix (ECM) attachments and interaction with a kinase that is essential for the epithelial-myoepithelial transformation i.e. c-Jun NH2-terminal kinase (JNK). Our data revealed that 4HPR binds with high affinity to the ATP-binding site of all three JNK isoforms with concurrent suppression of kinase function. Additional studies showed 4HPR treatment inhibited both OSCC cell-ECM adhesion and MMP activation and function. JNK downregulation and induced expression studies confirmed that the JNK3 isoform conveyed that largest impact on OSCC migration and invasion. Biodegradable polymeric implants formulated to preserve 4HPR's function and bioavailability were employed to assess 4HPR's chemopreventive impact on an OSCC tumor induction model. These studies revealed 4HPR local delivery significantly inhibited OSCC tumor size, mitotic indices and expression of the endothelial marker, erythroblast transformation-specific-related gene with concurrent increases in tumor apoptosis (cleaved caspase-3). Collectively, these data show that 4HPR suppresses invasion at multiple sites including 'outside-in' signaling, cell-ECM interactions and suppression of MMPs. These functions are also essential for physiologic function. Regulation is therefore essential and reinforces the pharmacologic advantage of local delivery chemopreventive formulations. .

Quercetin Potentiates the NGF-Induced Effects in Cultured PC 12 Cells: Identification by HerboChips Showing a Binding with NGF.

Dementia is a persistent disorder of the mental processes and is strongly related to depression. However, the performance of current antidepression medicine is far from satisfactory. Herbal extract provides an excellent source to identify compounds for possible drug development against depression. Here, HerboChips were employed to search herbal compounds that could bind nerve growth factor (NGF). By screening over 500 types of herbal extracts, the water extract of Ginkgo Folium, the leaf of Ginkgo biloba, showed a strong binding to NGF. The herbal fractions showing NGF binding were further isolated and enriched. By using LC-MS/MS analysis, one of the NGF binding fractions was enriched, which was further identified as quercetin, a major flavonoid in Ginkgo Folium. Quercetin, similar to Ginkgo Folium extract, could enhance the effect of NGF in cultured PC 12 cells, including potentiation of neurite outgrowth and phosphorylation of Erk-1/2. This is the first report of discovering an NGF binding compound by using HerboChips from herbal extracts, which could be further developed for antidepression application.