Mild cognitive impairment (MCI) is the prodromal stage of Alzheimer’s disease (AD). Identifying MCI subjects who are at high risk of converting to AD is crucial for effective treatments. In this study, a deep learning approach based on convolutional neural networks (CNN), is designed to accurately predict MCI-to-AD conversion with magnetic resonance imaging (MRI) data. First, MRI images are prepared with age-correction and other processing. Second, local patches, which are assembled into 2.5 dimensions, are extracted from these images. Then, the patches from AD and normal controls (NC) are used to train a CNN to identify deep learning features of MCI subjects. After that, structural brain image features are mined with FreeSurfer to assist CNN. Finally, both types of features are fed into an extreme learning machine classifier to predict the AD conversion. The proposed approach is validated on the standardized MRI datasets from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) project. This approach achieves an accuracy of 79.9% and an area under the receiver operating characteristic curve (AUC) of 86.1% in leave-one-out cross validations. Compared with other state-of-the-art methods, the proposed one outperforms others with higher accuracy and AUC, while keeping a good balance between the sensitivity and specificity. Results demonstrate great potentials of the proposed CNN-based approach for the prediction of MCI-to-AD conversion with solely MRI data. Age correction and assisted structural brain image features can boost the prediction performance of CNN.

Pregnancy is a physiological process with pronounced hormonal fluctuations in females, and relatively little is known regarding how pregnancy influences the ecological shifts of supragingival microbiota. In this study, supragingival plaques and salivary hormones were collected from 11 pregnant women during pregnancy (P1, ≤14 weeks; P2, 20–25 weeks; P3, 33–37 weeks) and the postpartum period (P4, 6 weeks after childbirth). Seven non-pregnant volunteers were sampled at the same time intervals. The microbial genetic repertoire was obtained by 16S rDNA sequencing. Our results indicated that the Shannon diversity in P3 was significantly higher than in the non-pregnant group. The principal coordinates analysis showed distinct clustering according to gestational status, and the partial least squares discriminant analysis identified 33 genera that may contribute to this difference. There were differentially distributed genera, among which Neisseria, Porphyromonas, and Treponema were over-represented in the pregnant group, while Streptococcus and Veillonella were more abundant in the non-pregnant group. In addition, 53 operational taxonomic units were observed to have positive correlations with sex hormones in a redundancy analysis, with Prevotella spp. and Treponema spp. being most abundant. The ecological events suggest that pregnancy has a role in shaping an at-risk-for-harm microbiota and provide a basis for etiological studies of pregnancy-associated oral dysbiosis.

Bone tissue engineering is an area of regenerative medicine that attempts to repair bone defects. Seed cells such as dental pulp stem cells (DPSCs) and adipose tissue-derived stem cells (ADSCs) are two of the most well-characterized cells for bone regeneration because their use involves few ethical constraints and they have the ability to differentiate into multiple cell types, secreting growth factors and depositing mineral. However, bone regeneration ability of these cells remains unclear. This study aimed to compare the bone formation capacity of DPSCs and ADSCs in vitro and in vivo. Studies revealed that DPSCs had enhanced colony-forming ability, higher proliferative ability, stronger migration ability and higher expression of angiogenesis-related genes. They also secreted more vascular endothelial growth factor compared to ADSCs. In contrast, ADSCs grew more slowly compared to DPSCs but exhibited greater osteogenic differentiation potential, higher expression of osteoblast marker genes, and greater mineral deposition. Furthermore, after DPSCs and ADSCs were implanted into a mandibular defect of a rat for 6 weeks, ADSCs showed visible bone tissue as early as week 1 and promoted faster and greater bone regeneration compared to the DPSC group. These results suggest that ADSCs might be more useful than DPSCs for bone regeneration.

Osteomodulin (OMD), a member of the small leucine-rich proteoglycan family, distributes in mineralized tissues and is positively regulated by bone morphogenetic protein 2 (BMP2). However, the exact function of OMD during mineralization and its association with BMP2 remain poorly understood. Herein, the expression pattern of OMD during osteogenesis was investigated in human dental pulp stem cells. Silencing OMD gene significantly suppressed the alkaline phosphatase activity, mineralized nodule formation and osteogenesis-associated gene transcription. Besides, OMD could enhance BMP2-induced expression of SP7 and RUNX2 with concentration dependence in vitro. Rat mandibular bone defect model revealed that scaffolds injected with the combination of OMD and suboptimal BMP2 exhibited more mature and abundant mineralized bone than that treated with OMD or suboptimal BMP2 alone. Mechanistically, OMD could bind to BMP2 via its terminal leucine-rich repeats and formed complexes with BMP2 and its membrane receptors, thus promoting BMP/SMAD signal transduction. In addition, OMD was a putative target gene of SMAD4, which plays a pivotal role in this pathway. Collectively, these data elucidate that OMD may act as a positive coordinator in osteogenesis through BMP2/SMADs signaling.