Enhanced metagenomic deep learning for disease prediction and consistent signature recognition by restructured microbiome 2D representations.

Metagenomic analysis has been explored for disease diagnosis and biomarker discovery. Low sample sizes, high dimensionality, and sparsity of metagenomic data challenge metagenomic investigations. Here, an unsupervised microbial embedding, grouping, and mapping algorithm (MEGMA) was developed to transform metagenomic data into individualized multichannel microbiome 2D representation by manifold learning and clustering of microbial profiles (e.g., composition, abundance, hierarchy, and taxonomy). These 2D representations enable enhanced disease prediction by established ConvNet-based AggMapNet models, outperforming the commonly used machine learning and deep learning models in metagenomic benchmark datasets. These 2D representations combined with AggMapNet explainable module robustly identified more reliable and replicable disease-prediction microbes (biomarkers). Employing the MEGMA-AggMapNet pipeline for biomarker identification from 5 disease datasets, 84% of the identified biomarkers have been described in over 74 distinct works as important for these diseases. Moreover, the method also discovered highly consistent sets of biomarkers in cross-cohort colorectal cancer (CRC) patients and microbial shifts in different CRC stages.

[Repeatability of three-dimensional measurements based on cone-beam CT images].

OBJECTIVE: To investigate the repeatability of three-dimensional (3-D) cephalometric measurements for the clinical application of 3-D cephalometry. METHODS: Forty-nine measurements that widely used in traditional cephalometric analyses were defined in 3-D cone-beam CT (CBCT) images. Three examiners identified landmarks on CBCT images of 17 subjects with normal occlusion, respectively, and 3-D measurements were exported automatically by software SimPlant. Inter-examiner reliability correlation coefficients (ICC) were obtained for all measurements. RESULTS: Repeatability of 36 measurements was high (ICC value greater than 0.9), including SNA, SNB. Repeatability of 11 measurements was moderate (ICC value between 0.8 and 0.9), including CoL-GoL, CoL-MSP. Repeatability of 2 measurements was low (ICC value lower than 0.8), including Gn-MSP and MPR-MSP. CONCLUSIONS: Most 3-D cephalometric measurements based on CBCT had high repeatability. However, some 3-D cephalometric measurements had limited repeatability.

[Three-dimensional analysis of craniofacial growth in adolescents with normal occlusion in Beijing].

OBJECTIVE: To analyze craniofacial growth three-dimensionally for adolescents with normal occlusion in Beijing. METHODS: One hundred and twenty-six adolescents with normal occlusion were selected according to the criteria. The sample was divided into four age groups (53 within 4 years, 30 within 7 years, 27 within 10 years and 16 within 13 years). Information of growth was collected. Three-dimensional cephalometric system based on cone-bean CT was established. RESULTS: From 4 to 13 years Co-A increased (14.55 ± 1.15) mm on average on the left and (13.66 ± 1.14) mm on the right, and Co-Gn increased (22.89 ± 1.40) mm on the left and (22.82 ± 1.38) mm on the right; and U1-NA increased (2.20 ± 0.44) mm on the left and (1.60 ± 0.46) mm on the right; and CoL-CoR and GoL-GoR increased (13.31 ± 1.21) mm and (18.59 ± 1.40) mm, and N-Me increased (18.03 ± 1.32) mm.SN-PP and SN-MPL basically remained unchanged. CONCLUSIONS: Adolescents with normal occlusion in Beijing grew obviously in three-dimensions and developed harmoniously.