The principles, traits, and proceedings of the data-driven methods including principal component analysis, independent component analysis and canonical analysis are summarized. And the aspects of the methods adapted to functional magnetic resonance images(fMRI) data are emphasized.

Objective To build the linear models for exploring relations between behavior data and functional magnetic resonance image(fMRI) signals of brain during cognitive task and to validate whether it is reasonable.Methods The linear models of behavior data and fMRI signals were built,and the functional regions of brain were detected by tests of corresponding parameters.Experimental data of Stroop tasks were used to study the effects of the models by comparing with the results of SPM.Results The results of Stroop data showed that dorsal lateral prefrontal cortex(Brodmann 9/46),and superior frontal median cortex(Brodmann 8/9) were associated with response time of Stroop tasks,and accorded with SPM results and other reports.Conclusion The models can quantitatively analyze the relations of response time and fMRI signals,providing a new way to explore functional images of cognition.

To develop a method of evaluating bone mass with CT numbers (Hounsfield unit, HU), volumetric mean of CT numbers (HUm) was defined to describe bone mass for the study of bone loss caused by microgravity. Ten femurs in the tail-suspended rats experiment were scanned by CT scanner. And the RUm of the intact, proximal, middle and distal femurs were obtained from these CT data. Then regressive analysis was performed between HUm and bone mineral density (BMD) measured by DXA in those regions. When HU threshold is equal to 400, the correlation between BMD and HUm of the intact femur is the best in all cases (R = 0.887, P <0.001), and the correlation between BMD and HUm of the proximal femur (R = 0.833, P <0.01) is better than those of the middle femur (R=0.683, P<0.05) and distal femur (R = 0.744, P<0.05). Volumetric mean of CT numbers has a good linear relationship with BMD; it can accurately describe bone mass and can be used to evaluate the results of the tail-suspended rats experiment.
Animals ; Bone Density ; Femur ; diagnostic imaging ; Radiographic Image Interpretation, Computer-Assisted ; Rats ; Tomography, X-Ray Computed ; methods

Objective:To evaluate whether underdilated stent could reduce the occurrence of hepatic encephalopathy (HE) after transjugular intrahepatic portosystemic shunt (TIPS) creation.Methods:A total of 197 patients with decompensated liver cirrhosis, who had underwent TIPS creation at Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, were analyzed retrospectively, including 110 males and 87 females with age 25-79 (54±11) years old. Uncovered and covered stents with 8 mm diameter were implanted in all subjects, and then dilated by balloon catheters with 6 mm or 8 mm diameter. The patients were divided into two groups, including underdilated group (6 mm, n=105) and control group (8 mm, n=92).Kaplan-Meier curves were used to illustrate cumulative rate of HE, and the differences were assessed with the log-rank test. Multivariate analyses with a Cox regression model were conducted to explore the risk factors for HE. Results:During a median follow-up period of 29 (12-54) months, 16 (15.2%) patients developed HE in the underdilated group and 27 (29.3%) patients in the control group. There was a significant difference in the cumulative rate of HE ( P=0.014), but no statistical differences were found in terms of variceal rebleeding, shunt dysfunction and survival between the two groups ( P=0.608, P=0.659, P=0.968). In multivariated analysis, group assignment (underdilated vs. control, HR=0.291, 95% CI 0.125-0.674, P=0.004) was identified as an independent risk factor for HE after TIPS creation. Conclusion:Underdilated TIPS could reduced the risk of HE compared with completely dilated TIPS, with comparable risk of variceal rebleeding, shunt dysfunction and mortality. And it is worthy of applying this technique to a large sample of patients in clinical practice.

Objective To analyze the CT signs of acute blunt or penetrating intestinal injury,and to improve the diagnostic accuracy of multi spiral CT for intestinal injury.Methods The CT and clinical data of 63 patients with intestinal injury confirmed by clinical surgical exploration who underwent emergency CT scan were collected,and the CT findings and surgical findings were comparatively analyzed.Results There were 63 cases of intestinal injury,of which 26 cases were complicated with mesenteric injury.The direct CT signs of intestinal injury included intestinal wall thickening sign and intestinal discontinuity sign,which accounted for 64%(40/63)and 17%(10/63),respectively.The indirect CT signs of intestinal injury included intraperitoneal/retroperitoneal gas sign,intraperitoneal/retroperitoneal effusion sign,intramural air,and portal venous gas,which accounted for 72%(45/63),88%(55/63),7%(5/63)and 5%(3/63),respectively.Conclusion Recognizing the CT signs of intestinal injury,such as intestinal wall thickening sign,intestinal discontinuity sign,intraperitoneal/retroperitoneal gas sign,intraperitoneal/retroperitoneal effusion sign,intramural air,and portal venous gas can help to make the early and correct diagnosis of intestinal injury if combined with clinical practice.

As the only translational factor that plays a critical role in two translational processes (elongation and ribosome regeneration), GTPase elongation factor G (EF-G) is a potential target for antimicrobial agents. Both Mycobacterium smegmatis and Mycobacterium tuberculosis have two EF-G homologous coding genes, MsmEFG1 (MSMEG_1400) and MsmEFG2 (MSMEG_6535), fusA1 (Rv0684) and fusA2 (Rv0120c), respectively. MsmEFG1 (MSMEG_1400) and fusA1 (Rv0684) were identified as essential genes for bacterial growth by gene mutation library and bioinformatic analysis. To investigate the biological function and characteristics of EF-G in mycobacterium, two induced EF-G knockdown strains (Msm-ΔEFG1(KD) and Msm-ΔEFG2(KD)) from Mycobacterium smegmatis were constructed by clustered regularly interspaced short palindromic repeats interference (CRISPRi) technique. EF-G2 knockdown had no effect on bacterial growth, while EF-G1 knockdown significantly retarded the growth of mycobacterium, weakened the film-forming ability, changed the colony morphology, and increased the length of mycobacterium. It was speculated that EF-G might be involved in the division of bacteria. Minimal inhibitory concentration assay showed that inhibition of EF-G1 expression enhanced the sensitivity of mycobacterium to rifampicin, isoniazid, erythromycin, fucidic acid, capreomycin and other antibacterial agents, suggesting that EF-G1 might be a potential target for screening anti-tuberculosis drugs in the future.
Antitubercular Agents/pharmacology* ; Bacterial Proteins/metabolism* ; Drug Resistance ; Mycobacterium smegmatis/metabolism* ; Peptide Elongation Factor G/pharmacology*