The Validation of Deep Learning-Based Grading Model for Diabetic Retinopathy.

Purpose: To evaluate the performance of a deep learning (DL)-based artificial intelligence (AI) hierarchical diagnosis software, EyeWisdom V1 for diabetic retinopathy (DR). Materials and Methods: The prospective study was a multicenter, double-blind, and self-controlled clinical trial. Non-dilated posterior pole fundus images were evaluated by ophthalmologists and EyeWisdom V1, respectively. The diagnosis of manual grading was considered as the gold standard. Primary evaluation index (sensitivity and specificity) and secondary evaluation index like positive predictive values (PPV), negative predictive values (NPV), etc., were calculated to evaluate the performance of EyeWisdom V1. Results: A total of 1,089 fundus images from 630 patients were included, with a mean age of (56.52 ± 11.13) years. For any DR, the sensitivity, specificity, PPV, and NPV were 98.23% (95% CI 96.93-99.08%), 74.45% (95% CI 69.95-78.60%), 86.38% (95% CI 83.76-88.72%), and 96.23% (95% CI 93.50-98.04%), respectively; For sight-threatening DR (STDR, severe non-proliferative DR or worse), the above indicators were 80.47% (95% CI 75.07-85.14%), 97.96% (95% CI 96.75-98.81%), 92.38% (95% CI 88.07-95.50%), and 94.23% (95% CI 92.46-95.68%); For referral DR (moderate non-proliferative DR or worse), the sensitivity and specificity were 92.96% (95% CI 90.66-94.84%) and 93.32% (95% CI 90.65-95.42%), with the PPV of 94.93% (95% CI 92.89-96.53%) and the NPV of 90.78% (95% CI 87.81-93.22%). The kappa score of EyeWisdom V1 was 0.860 (0.827-0.890) with the AUC of 0.958 for referral DR. Conclusion: The EyeWisdom V1 could provide reliable DR grading and referral recommendation based on the fundus images of diabetics.

Three-dimensional diabetic macular edema thickness maps based on fluid segmentation and fovea detection using deep learning.

AIM: To explore a more accurate quantifying diagnosis method of diabetic macular edema (DME) by displaying detailed 3D morphometry beyond the gold-standard quantification indicator-central retinal thickness (CRT) and apply it in follow-up of DME patients. METHODS: Optical coherence tomography (OCT) scans of 229 eyes from 160 patients were collected. We manually annotated cystoid macular edema (CME), subretinal fluid (SRF) and fovea as ground truths. Deep convolution neural networks (DCNNs) were constructed including U-Net, sASPP, HRNetV2-W48, and HRNetV2-W48+Object-Contextual Representation (OCR) for fluid (CME+SRF) segmentation and fovea detection respectively, based on which the thickness maps of CME, SRF and retina were generated and divided by Early Treatment Diabetic Retinopathy Study (ETDRS) grid. RESULTS: In fluid segmentation, with the best DCNN constructed and loss function, the dice similarity coefficients (DSC) of segmentation reached 0.78 (CME), 0.82 (SRF), and 0.95 (retina). In fovea detection, the average deviation between the predicted fovea and the ground truth reached 145.7±117.8 µm. The generated macular edema thickness maps are able to discover center-involved DME by intuitive morphometry and fluid volume, which is ignored by the traditional definition of CRT>250 µm. Thickness maps could also help to discover fluid above or below the fovea center ignored or underestimated by a single OCT B-scan. CONCLUSION: Compared to the traditional unidimensional indicator-CRT, 3D macular edema thickness maps are able to display more intuitive morphometry and detailed statistics of DME, supporting more accurate diagnoses and follow-up of DME patients.

Effective treatment of apatinib in desmoplastic small round cell tumor: a case report and literature review.

BACKGROUND: Desmoplastic small round cell tumor (DSRCT) is a rare malignant sarcoma with poor prognosis due to lack of effective treatments. Apatinib is a new potent oral small-molecule tyrosine kinase inhibitor, and targets the intracellular domain of vascular endothelial growth factor receptor 2 (VEGFR-2). In this study, we presented a case of intra-abdominal DSRCT which was effectively treated by apatinib. CASE PRESENTATION: A 32-year-old man was admitted due to increasing urination frequency and palpable mass in right lower abdomen for 2 months. The mass was resected and diagnosed DSRCT. The patient refused chemotherapy and radiotherapy,and used Chinese medicine only. Six months after the surgery, the patient re-hospitalized due to growing abdominal mass and ascites. Intraperitoneal cisplatin treatment showed little effect. Apatinib was then recommended. Apatinib revealed outstanding effect on reducing mass size and ascites during 2-month treatment. Apatinib therapy continued for additional 2 months, and the patient was in good condition. The only toxicity was hand-food syndrome, which was controllable and well tolerated. CONCLUSION: It is the first report that apatinib is effective on DSRCT. This report may provide an additional option for the treatment of metastatic DSRCT.

Method for Removing Spectral Contaminants to Improve Analysis of Raman Imaging Data.

The spectral contaminants are inevitable during micro-Raman measurements. A key challenge is how to remove them from the original imaging data, since they can distort further results of data analysis. Here, we propose a method named "automatic pre-processing method for Raman imaging data set (APRI)", which includes the adaptive iteratively reweighted penalized least-squares (airPLS) algorithm and the principal component analysis (PCA). It eliminates the baseline drifts and cosmic spikes by using the spectral features themselves. The utility of APRI is illustrated by removing the spectral contaminants from a Raman imaging data set of a wood sample. In addition, APRI is computationally efficient, conceptually simple and potential to be extended to other methods of spectroscopy, such as infrared (IR), nuclear magnetic resonance (NMR), X-Ray Diffraction (XRD). With the help of our approach, a typical spectral analysis can be performed by a non-specialist user to obtain useful information from a spectroscopic imaging data set.

Identification of candidate target genes for human peripheral arterial disease using weighted gene co­expression network analysis.

The aim of the present study was to identify the potential treatment targets of peripheral arterial disease (PAD) and provide further insights into the underlying mechanism of PAD, based on a weighted gene co­expression network analysis (WGCNA) method. The mRNA expression profiles (accession. no. GSE27034), which included 19 samples from patients with PAD and 18 samples from normal control individuals were extracted from the Gene Expression Omnibus database. Subsequently, the differentially expressed genes (DEGs) were obtained using the Limma package and the co­expression network modules were screened using the WGCNA approach. In addition, the protein­protein interaction network for the DEGs in the most significant module was constructed using Cytoscape software. Functional enrichment analyses of the DEGs in the most significant module were also performed using the Database for Annotation, Visualization and Integrated Discovery and Kyoto Encyclopedia of Genes and Genomes (KEGG) Orthology­Based Annotation System, respectively. A total of 148 DEGs were identified in PAD, which were used to construct the WGCN, in which two modules (gray module and turquoise module) were identified, with the gray module exhibiting a higher gene significance (GS) value than the turquoise module. In addition, a co­expression network was constructed for 60 DEGs in the gray module. The functional enrichment results showed that the DEGs in the gray module were enriched in five Gene Ontology terms and four KEGG pathways. For example, cyclin­dependent kinase inhibitor 1A (CDKN1A), FBJ murine osteosarcoma viral oncogene homolog (FOS) and prostaglandin­endoperoxide synthase 2 (PTGS2) were enriched in response to glucocorticoid stimulus. The results of the present study suggested that DEGs in the gray module, including CDKN1A, FOS and PTGS2, may be associated with the pathogenesis of PAD, by modulating the cell cycle, and may offer potential for use as candidate treatment targets for PAD.

[Application of FTIR Microspectroscopy in the Study of Lignocellulosic Cell Walls].

The development and utilization of lignocellulosic biomass resource becomes a major subject of widespread concern at home and abroad. To make better use of this resource and reduce cost, a thorough understanding of recalcitrance mechanism as well as chemical compositional and structural characteristics of lignocellulosic cell walls seems desiderated. With the property of fast testing, high sensitivity and simple sample preparation, FTIR microspectroscopy can provide microanalysis and nondestructive test. It has been proved to accurately present compositional and structural information of lignocellulosic cell walls and be suitable for the study of topochemistry of main components at the cellular level. After a brief introduction of the principle of FTIR microspectroscopy and its detailed experimental procedure, this review provides an overview on the feasibility of FTIR microspectroscopy in monitoring as well as evaluating the composition and structure of lignocellulosic cell walls from three aspects, including in situ distribution, changes after pretreatment and molecular orientation of the major components in cell walls. Some prospects, for example development of the magnification of the microscope and the spatial resolution of chemical imaging, combination with various analysis methods and instruments are also proposed. This technology can act as a new approach in characterizing lignocellulosic cell walls.

Comparison between better and poorly differentiated locally advanced gastric cancer in preoperative chemotherapy: a retrospective, comparative study at a single tertiary care institute.

BACKGROUND: Gastric cancer is the third leading cause of cancer-related mortality in China, and the long-term survival for locally advanced gastric cancer is very poor. Simple surgery cannot yield an ideal result because of the high recurrence rate after tumor resection. Preoperative chemotherapy could help to reduce tumor volume, improve the R0 resection rate (no residual tumor after surgery), and decrease the risk of local tumor recurrence. The aim of this study was to evaluate the influence of pathological differentiation in the effect of preoperative chemotherapy for patients with locally advanced gastric cancer. METHODS: Patients with locally advanced gastric cancer (n = 32) received preoperative chemotherapy under the XELOX (capecitabine plus oxaliplatin) regimen. According to pathological examination, patients' tumors were classified into better (well and moderate) and poorly differentiated (lower differentiated and undifferentiated) groups, and the clinical response rate, type of gastrectomy, and negative tumor residual rate were compared between the two groups of patients. Morphological changes and toxic reactions were monitored after chemotherapy. RESULTS: The results showed that the clinical response rate in the better differentiated group was significantly higher than that in the poorly differentiated group (100% versus 25%, P = 0.000). The partial gastrectomy rate in the better differentiated group was significantly higher than that in the poorly differentiated group (87.5% versus 25% P = 0.000). A significant shrinking of tumor and necrosis of tumor tissues caused by chemotherapy could be observed. CONCLUSIONS: In conclusion, the better differentiated group with locally advanced gastric cancer is suitable for preoperative chemotherapy under the XELOX regimen, and as a result of effective preoperative chemotherapy, much more gastric tissue can be preserved for the better differentiated group.

[Predictive value of tumor vessel density from enhanced spiral CT scan for lymph node and distant metastasis of colorectal cancer].

OBJECTIVE: To test whether the tumor vessel density (TVD) from the enhanced spiral CT can preoperatively predict nodal status and distant metastasis of colorectal cancer. METHODS: Forty cases of colorectal cancer patients who received surgical treatment were included in this study. The three dimensional tumor vessels were reconstructed by an enhanced CT 64-slice spiral CT and its AW4.4 image processing platform. The TVD was measured by the 1000 high-resolution color graphics pathological analysis system. The TVD level was compared between different tumor size, classification, and TNM stage. The postoperative pathological staging was taken as golden standard. RESULTS: The sensitivity, specificity and accuracy for direct prediction of lymph node metastasis by the enhanced CT 64-slice spiral CT was 74.1%(20/27), 53.8%(7/13) and 67.5%(27/40) respectively. The TVD from the reconstructed three dimensional tumor vessels in the group with lymph node metastasis was significantly higher than that without metastasis(0.070±0.046 vs. 0.037±0.013, P<0.05). The TVD in the distant metastasis group was significantly higher than that without distant metastasis (0.130±0.032 vs. 0.049±0.030, P<0.01). No difference of TVD was found between different tumor size, invasion depth, and differentiation type. CONCLUSION: TVD level from the reconstructed three dimensional tumor vessels can indicate lymph node and distant metastasis of colorectal cancer.

Laparoscopic versus open resection of gastric gastrointestinal stromal tumors.

The aims of this study were to explore whether laparoscopic surgical resections of gastric gastrointestinal stromal tumors (GISTs) would produce better perioperative and similar oncologic outcomes compared with open surgical resection in Chinese patients. Thirty-six gastric GISTs cases were divided into a minimally invasive laparoscopic group and open resection group, depending on the surgical approach that was used. The general preoperative information, operative time, incision length, intraoperative blood loss, postoperative time to first flatulence, postoperative complications, postoperative hospital stay, total hospitalization costs, and such follow-up data as recurrence, metastasis, and mortality rates were compared between two groups. Among the 36 gastric GISTs, 15 received laparoscopic surgical treatment (laparoscopy group, n=15), and 21 received routine open resection treatment (open resection group, n=21). The laparoscopy group and the open resection group showed statistically significant differences (P<0.05) in incision length (7.8±2.3 vs. 16.9±3.8 cm), postoperative time to first flatulence (3.8±1.3 vs. 5.1±2.1 d), postoperative hospitalization time (7.6±2.5 vs. 11.3±3.7 d), and total cost of hospitalization (RMB 28,239±5,521 vs. RMB 23,761±5,362). There were no statistically significant differences (P>0.05) between the laparoscopy group and the open resection group in operative time (147.8±59.3 vs. 139.2±62.1 min) and intraoperative blood loss (149.8±98.9 vs. 154.2±99.3 mL). Both groups had no postoperative complications, no recurrence and metastasis, and no postoperative mortality. There were no statistically significant differences between the two groups in postoperative complications, postoperative recurrence and metastasis, and postoperative mortality. In conclusion, compared with open resection, the laparoscopic resection of gastric GISTs offers the advantages of less trauma, faster recovery, and shorter hospital stay.

[Application of proteomics in the research of discrepancy proteins in gastric cancer].

OBJECTIVE: To explore the discrepancy proteins in gastric cancer by proteome analysis. METHODS: Total proteins of gastric cancer tissues and matched normal gastric epithelial tissues were separated respectively by two-dimensional polyacrylamide gel electrophoresis (2-DE). Mass spectrometry was used to test the differentially expressed proteins. RESULTS: One thousand one hundred and forty-seven protein spots from gastric cancer tissue and 1079 spots from the normal tissue were gained. Out of 164 different protein spots, 41 were only expressed in gastric cancer tissue, 27 were unique in normal tissue, 39 were up-regulated and 57 were down-regulated in gastric cancer. Seven proteins, which were highly expressed in gastric cancer tissue, were identified. CONCLUSION: Different protein spots between gastric cancer tissues and normal gastric epithelial tissue were gained by proteomics. The 7 discrepancy proteins were further identified. It establishes the foundation of finding specific gastric cancer proteins, which act as biomarkers for the diagnosis and prognosis of gastric cancer.

Metabolic effects of telmisartan in spontaneously hypertensive rats.

The favorable metabolic effects of telmisartan are supposedly related to the changes in carbohydrate and lipid metabolism driven by peroxisome proliferators-activated receptor-gamma (PPARgamma). The fatty acid translocase CD36 is one of the PPARgamma targets that mediate these actions. We studied the metabolic effects of telmisartan in the NIH-derived strain of spontaneously hypertensive rats (SHR/NIH), which harbors a deletion mutation in CD36, in comparison to the original SHRs (SHR/Izm), which express wild-type CD36. In SHR/Izm, administration of telmisartan was associated with significantly lower serum levels of free fatty acids (42%), triglycerides (29%), glucose (11%), insulin (31%), and lower hepatic triglyceride (17%) levels, as well as larger epididymal fat pads (1.19-fold) than in SHR/NIH. Additionally, insulin-stimulated glucose incorporation into epididymal fat tissues was significantly augmented in SHR/Izm (1.33-fold) compared with SHR/NIH. In the epididymal fat pads of SHR/Izm treated with telmisartan, CD36 mRNA transcript (1.55-fold) and protein expression (1.37-fold) were also significantly enhanced. However, after 4 weeks of treatment with telmisartan, in SHR/NIH only serum free fatty acid levels were slightly reduced (20%). Overall, these results showed marked discrepancies in the metabolic actions of telmisartan in SHR/Izm and SHR/NIH and further supported the involvement of CD36 in the actions of this drug, suggesting that this pharmacogenetic interaction may be of particular importance in CD36-deficient patients.