Segmentation of meibomian glands based on deep learning / 国际眼科杂志(Guoji Yanke Zazhi)

AIM: To explore the application value of deep learning technology in automatic meibomian glands segmentation. METHODS:Infrared meibomian gland images were collected and 193 of them were picked out for establishing the database. The images were manually labeled by three clinicians. UNet++ network and automatic data expansion strategy were introduced to construct the automatic meibomian glands segmentation model. The feasibility and effectiveness of the proposed segmentation model were analyzed by precision, sensitivity, specificity, accuracy and intersection over union.RESULTS: Taking manual labeling as the gold standard, the presented method segment the glands effectively and steadily with accuracy, sensitivity and specificity of 94.31%, 82.15% and 96.13% respectively. On the average, only 0.11s was taken for glands segmentation of single image.CONCLUSIONS: In this paper, deep learning technology is introduced to realize automatic segmentation of meibomian glands, achieving high accuracy, good stability and efficiency. It would be quite useful for calculation of gland morphological parameters, the clinical diagnosis and screening of related diseases, improving the diagnostic efficiency.

Screening disease-associated proteins from sera of patients with rheumatoid arthritis: a comparative proteomic study / 中华医学杂志(英文版)

<p><b>BACKGROUND</b>Rheumatoid arthritis (RA) is an autoimmune disease characterized by chronic inflammation at the synovial membrane. Although great progress has been made recently in exploring the etiology and pathogenesis of RA, its molecular pathological mechanism remains to be further defined and it is still a great challenge in determining the diagnosis and in choosing the appropriate therapy in early patients. This study was performed to screen candidate RA-associated serum proteins by comparative proteomics to provide research clues to early diagnosis and treatment of RA.</p><p><b>METHODS</b>Sera isolated from 6 RA patients and 6 healthy volunteers were pooled respectively and high-abundance proteins were depleted by Plasma 7 Multiple Affinity Removal System. The protein expression profiles between the two groups were then compared by two-dimensional gel electrophoresis (2-DE) and the proteins over/under-expressed by more than 3-fold were identified by mass spectrometry analysis. To validate the differential expression levels of the identified proteins between the two groups, ELISA was performed in two of the identified proteins in individual sera from 32 RA patients and 32 volunteers.</p><p><b>RESULTS</b>Eight proteins which over/under-expressed in sera of RA patients were identified. Among them, chain A of transthyretin (TTR) was under-expressed, while serum amyloid A protein, apolipoprotein A (ApoA)-IV, ApoA-IV precursor, haptoglobin 2, ceruloplasmin (Cp), immunoglobulin superfamily 22 and HT016 were over-expressed. ELISA test confirmed that Cp expressed remarkably higher while TTR obviously lower in RA group compared with volunteer group.</p><p><b>CONCLUSION</b>There were 8 identified proteins differentially expressed between RA group and volunteer group, which might be candidate RA-associated proteins and might be promising diagnostic indicators or therapeutic targets for RA.</p>

Value of the peripheral blood B-cells subsets in patients with ankylosing spondylitis / 中华医学杂志(英文版)

<p><b>BACKGROUND</b>The role of B-cell remains an enigma in the pathogenesis of ankylosing spondylitis (AS). This study aimed to investigate the distributions of B-cells and subsets in peripheral blood of AS patients and observe their changes in etanercept-treated AS patents.</p><p><b>METHODS</b>We detected the proportions of CD19(+) B-cell, naive B-cell (CD19(+)CD27-), memory B-cell (CD19(+)CD27dim) and plasmablast (CD19(+)CD27high) in peripheral blood of 66 patients with AS (39 at active stage, 27 at stable stage; 35 patients with peripheral joint involvement, 31 patients with axial involvement alone), 30 patients with rheumatoid arthritis (RA) and 30 healthy volunteers using flow cytometry. And then we observed the changes of the above indexes of 39 active AS patients treated with etanercept in a randomized, double-blind, placebo-controlled trial.</p><p><b>RESULTS</b>(1) Percentages of CD19(+) B-cells in active or peripheral joint involvement AS patients increased more obviously than those in stable or axial involvement alone AS patients (both P = 0.001), and percentage of CD19(+)CD27high B-cells in AS patients with peripheral joint involvement was significantly higher than that in cases with axial involvement alone or healthy volunteers (P = 0.005 and 0.006, respectively); (2) The percentage of CD19(+) B-cells in AS patients was positively correlated with Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) scores, Patient's Global Assessment (PGA) scores, total back pain scores and nocturnal back pain scores (r = 0.270, 0.255, 0.251 and 0.266, P = 0.029, 0.039, 0.042 and 0.031, respectively); (3) At week 6 and week 12, there were no statistical differences of the percentages of B-cells and subsets between etanercept group and placebo group of AS patients (P > 0.05); the percentage of CD19(+) B-cells in etanercept group was higher than that in healthy volunteers at week 12 (t = 3.320, P = 0.003).</p><p><b>CONCLUSIONS</b>Misbalance is present in B-cells and some subsets in peripheral blood of active AS patients with peripheral joint involved. B-cells might play an important role in the pathogenesis of AS patients. The high percentage of CD19(+) B-cells in active AS patients cannot be down-regulated after 12-week etanercept treatment.</p>