Search

To construct and validate a deep learning radiomics (DLR) model based on X-ray images for predicting and distinguishing acute and chronic osteoporotic vertebral fractures (OVFs). A total of 942 cases (1076 vertebral bodies) with both vertebral X-ray examination and MRI scans were included in this study from three hospitals. They were divided into a training cohort (n = 712), an internal validation cohort (n = 178), an external validation cohort (n = 111), and a prospective validation cohort (n = 75). The ResNet-50 model architecture was used for deep transfer learning (DTL), with pre-training performed on RadImageNet and ImageNet datasets. DTL features and radiomics features were extracted from lateral X-ray images of OVFs patients and fused together. A logistic regression model with the least absolute shrinkage and selection operator was established, with MRI showing bone marrow edema as the gold standard for acute OVFs. The performance of the model was evaluated using receiver operating characteristic curves. Eight machine learning classification models were evaluated for their ability to distinguish between acute and chronic OVFs. The Nomogram was constructed by combining clinical baseline data to achieve visualized classification assessment. The predictive performance of the best RadImageNet model and ImageNet model was compared using the Delong test. The clinical value of the Nomogram was evaluated using decision curve analysis (DCA). Pre-training resulted in 34 and 39 fused features after feature selection and fusion. The most effective machine learning algorithm in both DLR models was Light Gradient Boosting Machine. Using the Delong test, the area under the curve (AUC) for distinguishing between acute and chronic OVFs in the training cohort was 0.979 and 0.972 for the RadImageNet and ImageNet models, respectively, with no statistically significant difference between them (P = 0.235). In the internal validation cohort, external validation cohort, and prospective validation cohort, the AUCs for the two models were 0.967 vs 0.629, 0.886 vs 0.817, and 0.933 vs 0.661, respectively, with statistically significant differences in all comparisons (P < 0.05). The deep learning radiomics nomogram (DLRN) was constructed by combining the predictive model of RadImageNet with clinical baseline features, resulting in AUCs of 0.981, 0.974, 0.895, and 0.902 in the training cohort, internal validation cohort, external validation cohort, and prospective validation cohort, respectively. Using the Delong test, the AUCs for the fused feature model and the DLRN in the training cohort were 0.979 and 0.981, respectively, with no statistically significant difference between them (P = 0.169). In the internal validation cohort, external validation cohort, and prospective validation cohort, the AUCs for the two models were 0.967 vs 0.974, 0.886 vs 0.895, and 0.933 vs 0.902, respectively, with statistically significant differences in all comparisons (P < 0.05). The Nomogram showed a slight improvement in predictive performance in the internal and external validation cohort, but a slight decrease in the prospective validation cohort (0.933 vs 0.902). DCA showed that the Nomogram provided more benefits to patients compared to the DLR models. Compared to the ImageNet model, the RadImageNet model has higher diagnostic value in distinguishing between acute and chronic OVFs. Furthermore, the diagnostic performance of the model is further improved when combined with clinical baseline features to construct the Nomogram. [ABSTRACT FROM AUTHOR]

• Multiple discriminative bacterial taxa were identified among the MDD subgroups. • Fecal Streptococcus was highly correlated with both HAMD and HAMA scores. • Fecal indole-3-carboxyaldehyde proved useful to discriminate depressive symptoms. Major depressive disorder (MDD) is a heterogeneous mental disorder for which the precise assessment of symptom severity remains challenging. Studies have consistently found that the microbiota–gut–brain (MGB) axis is profoundly altered in MDD, but whether MGB-relevant clinical parameters are applicable to depression subphenotyping remains largely unexplored. In this prospective study, we assessed the taxonomic and metabolic signatures of fecal microbiota from 45 unmedicated MDD patients and explored their associations with the severity of depression and anxiety symptoms as measured by Hamilton depression scale-17 (HAMD-17) and Hamilton anxiety scale-14 (HAMA-14), respectively. The global microbial compositions of MDD patients with mild, moderate and severe symptoms were largely similar. Nevertheless, multiple discriminative bacterial taxa could be identified among the subgroups across the genus to species level. The abundance of fecal Streptococcus was highly correlated with both HAMD and HAMA scores. Patients with severe depression symptoms showed significantly higher abundance of Phascolarctobacterium and Akkermansia , while enrichment of Akkermansia , Coprococcus and Streptococcus were observed with severe anxiety symptoms. In addition, fecal microbial metabolite indole-3-carboxyaldehyde proved useful to discriminate the severity of depression or anxiety symptoms. Together, our results support the utility of microbial taxa and metabolites as potential MGB-based biomarker panel for stratifying the symptom severity of MDD patients. [ABSTRACT FROM AUTHOR]

Glycosylation, one of the most important post-translational modifications of proteins, plays an irreplaceable role in the whole process of spermatogenesis, sperm-egg recognition, and fertilization. Herein, we mapped the first bovine sperm N-linked glycoproteome and a total of 1188 N -glycosylated sites on 626 proteins were identified. Bioinformatics analysis revealed that bovine sperm N- glycosylated proteins were classified into "extracellular region" and "lysosome" groups based on cellular component annotation and enrichment of glycoproteins with proteolytic and reproductive functions. Notably, cysteines were highly enriched in the canonical N -glycosylation motifs N-!P-[S/T/C] and the conservative motifs N–C-[S/T] were also significantly enriched, indicating these modifications play extraordinary roles in bovine spermatogenesis and maturation. The percentage of cysteine at the second position relative to modified asparagine was 7.5%, much higher than that of the previously reported N-linked glycoproteome. A total of 120 cysteine enriched N -glycoproteins were identified, which had significantly upregulated metalloendopeptidase activity and metal ion binding compared with the whole bovine sperm glycoproteome. Strikingly, 15 of 58 N–C-[S/T] motif-containing glycoproteins had a disintegrin and metalloproteinase (ADAM) protein domain. Thus, we hypothesized that ADAM-containing conserved free cysteine residues in N-linked glycoprotein motifs may be key cysteine-switches and may have extraordinary roles in bovine spermatozoa. In conclusion, almost all bovine sperm glycoproteins have enzyme activity, participate in proteolysis, and play indispensable roles in spermatogenesis, sperm-egg recognition, and eventual fertilization. The mapping of N -glycosylation on bovine sperm may provide a new means to explore potential biomarkers for improving sperm quality and fertility. - The first bovine sperm N-glycoproteomic was mapped with identification of 1188 N-glycosylated sites on 626 proteins. - Motifs N–C-[S/T] are significantly enriched in bovine sperm N-glycoproteins and the percentage of cysteine in motifs N-!P-[S/T/C] is extremely high. - ADAM containing conserved free cysteine residues in N-glycoprotein motifs may be the key cysteine-switch and have extraordinary roles in bovine spermatozoa. - Mapping of bovine sperm N-glycoproteomic is useful to explore potential biomarkers for improving sperm quality and fertility [ABSTRACT FROM AUTHOR]