Accurate measurement of magnetic resonance parkinsonism index by a fully automatic and deep learning quantification pipeline.

OBJECTIVES: This study aims at a fully automatic pipeline for measuring the magnetic resonance parkinsonism index (MRPI) using deep learning methods. METHODS: MRPI is defined as the product of the pons area to the midbrain area ratio and the middle cerebellar peduncle (MCP) width to the superior cerebellar peduncle (SCP) width ratio. In our proposed pipeline, we first used nnUNet to segment the brainstem and then employed HRNet to identify two key boundary points so as to sub-divide the whole brainstem into midbrain and pons. HRNet was also employed to predict the MCP endpoints for measuring the MCP width. Finally, we segmented the SCP on an oblique coronal plane and calculated its width. A total of 400 T1-weighted magnetic resonance images (MRIs) were used to train the nnUNet and HRNet models. Five-fold cross-validation was conducted to evaluate our proposed pipeline's performance on the training dataset. We also evaluated the performance of our proposed pipeline on three external datasets. Two of them had two raters manually measuring the MRPI values, providing insights into automatic accuracy versus inter-rater variability. RESULTS: We obtained average absolute percentage errors (APEs) of 17.21%, 18.17%, 20.83%, and 22.83% on the training dataset and the three external validation datasets, while the inter-rater average APE measured on the first two external validation datasets was 11.31%. Our proposed pipeline significantly improved the MRPI quantification accuracy over a representative state-of-the-art traditional approach (p < 0.001). CONCLUSION: The proposed automatic pipeline can accurately predict MRPI that is comparable with manual measurement. CLINICAL RELEVANCE STATEMENT: This study presents an automated magnetic resonance parkinsonism index measurement tool that can analyze large amounts of magnetic resonance images, enhance the efficiency of Parkinsonism-Plus syndrome diagnosis, reduce the workload of clinicians, and minimize the impact of human factors on diagnosis. KEY POINTS: • We propose an automatic pipeline for measuring the magnetic resonance parkinsonism index from magnetic resonance images. • The effectiveness of the proposed pipeline is successfully established on multiple datasets and comparisons with inter-rater measurements. • The proposed pipeline significantly outperforms a state-of-the-art quantification approach, being much closer to ground truth.

Effects of cyanobacteria- and moss-biocrusts on soil aggregate stability and splash erosion in croplands of the China Mollisols area.

To investigate the effects of biocrusts development on aggregate stability and splash erosion of Mollisols and to understand its function in soil and water conservation, we collected biocrusts (cyano crust and moss crust) samples in croplands during the growing season and measured the differences in aggregate stability between biocrusts and uncrusted soil. The effects of biocrusts on reduction of raindrop kinetic energy were determined and splash erosion amounts were obtained with single raindrop and simulated rainfall experiments. The correlations among soil aggregate stability, splash erosion characteristics, and fundamental properties of biocrusts were analyzed. The results showed that compared to uncrusted soil, the cyano crust and the moss crust decreased the proportion of soil water-stable aggregates <0.25 mm by 10.5% and 21.8%, respectively, while their soil water-stable aggregates 5-10 mm were 4.0 and 8.8 times as that of uncrusted soil. In contrast to uncrusted soil, the macroaggregate content (R0.25), mean weight diameter (MWD), and geometric mean diameter (GMD) of biocrusts were 31.5%, 76.2%, and 33.5% higher, respectively. In addition, biocrusts reduced raindrop kinetic energy by an average of 0.48 J compared to uncrusted soil. The breakthrough raindrop kinetic energy of cyano crust and moss crust were 2.9 and 26.2 times as that of uncrusted soil, while the reduction of raindrop kinetic energy by cyano crust with high biomass was 1.3 and 6.6 times as that of medium and low biomass, respectively. Under the single raindrop and simulated rainfall conditions, biocrusts reduced splash erosion amounts by 47.5% and 79.4%, respectively. The proportion of aggregates >0.25 mm in the splash soil particles of biocrusts (37.9%) was 40.3% lower than that of uncrusted soil, while the proportion of aggregates >0.25 mm decreased as biocrust biomass increased. Moreover, the aggregate stability, splash erosion amount, and fundamental properties of biocrusts were significantly correlated. The MWD of aggregates was significantly and negatively correlated with the splash erosion amount under single raindrop and simulated rainfall conditions, indicating that the improved aggregate stability of surface soil caused by biocrusts accounted for reducing splash erosion. The biomass, thickness, water content, and organic matter content of biocrusts had significant effects on aggregate stability and splash characteristics. In conclusion, biocrusts significantly promoted soil aggregate stability and reduced splash erosion, which had great significance to soil erosion prevention and the conservation and sustainable utilization of Mollisols.

[Effects of biological soil crusts on solute transport characteristics of sandy and loessal soils on the Loess Plateau, China]. / 黄土高原生物结皮覆盖对风沙土和黄绵土溶质运移的影响.

Biological soil crusts (BSCs) greatly change surface soil structure and nutrient enrichment processes in arid and semiarid regions. However, their impacts on solute transport characteristics and nutrient loss are still not clear. In this study, the solute (Cl- and Ca2+) transport experiments were conducted on soils covered by moss-dominated BSCs and uncrusted soil on sandy and loessal soils on the Loess Plateau, respectively. We analyzed the solute transport characteristics of the BSCs covered soil and uncrusted soil in different soil depths (0-5 cm and 5-10 cm). The BSCs mulching generated delay effects on the solute breakthrough process of 0-5 cm soils. The breakthrough time of Cl- in the BSCs covered soil was 3.83 (sandy soil) and 2.09 times (loessal soil) longer than that in the uncrusted soil. The breakthrough time of Ca2+ in the BSCs covered soil was 2.50 and 2.73 times longer than that in the uncrusted soil. Due to the strong influence of BSCs mulching, the pore volume number of the complete solute breakthrough at 0-5 cm depth was higher than that at 5-10 cm depth in the BSCs covered soils. The breakthrough time of Cl- at 0-5 cm depth was increased by 67.3% (sandy soil) and 51.8% (loessal soil) by the BSCs as compared with that at 5-10 cm depth. The breakthrough time of Ca2+ at 0-5 cm depth was increased by 8.0% and 33.7% by the BSCs. The BSCs reduced soil pore water flow velocity by 37.5%-70.2% compared with the uncrusted soil. Except for the sandy soil at 5-10 cm depth, the BSCs increased the solute dispersion coefficient by 1.73-6.29 times and the degree of dispersion by 2.77-20.95 times compared with the uncrusted soils. After the complete breakthrough of solute, the content of Ca2+ in the BSCs layer (0-2 cm) was 4.14 and 2.58 times higher than that in the uncrusted sandy and loessal soils, respectively. In conclusion, our results indicated that BSCs could reduce the deep percolation and loss of nutrients accumulated in surface soil through improving their solute adsorption and retention abilities, which is of great significance for the improvement of soil fertility and vegetation restoration on degraded land in arid and semiarid regions.

[Distribution of exogenous nitrogen fractions and their fate in moss-dominated biological soil crusts]. / å¤–æºè¾“å ¥æ°®ç´ åœ¨è—“ç±»ç”Ÿç‰©åœŸå£¤ç»“çš®ä¸­å„æ°®ç»„åˆ†çš„åˆ†é ç‰¹å¾ä¸Žå½’è¶‹é€”å¾„.

Nitrogen (N) labeled with 15N was evenly added into plots of moss-dominated biological soil crusts (BSCs) and bare soil on the Chinese Loess Plateau. After that, the surface BSCs and bare soil samples were continuously collected within 1-30 days. The 15N content of each N fraction in soil, microorganisms, and mosses was measured for each sample. The effects of BSCs on soil N fate and cycling was determined through analyzing the differences in the distribution of 15N fractions between the BSCs and bare soil. Our results showed that: 1) The 15N content of total N (TN), microbial biomass N (MBN), and dissolved organic N (DON) in the BSCs was 2.9, 17.5, and 9.0 times higher than that in the bare soil, respectively. The 15N content of moss plants in the BSCs was 4.73 mg kg-1. 2) The residual rate of 15N in the BSCs and bare soil was 13.0% and 3.3%, respectively, indicating that the N fixing and holding ability of BSCs was four times higher than that of bare soil. The percentage of each 15N fraction in T15N in the BSCs was in the order of MBN (54.3%)>moss plant N (22.5%)>DON (6.2%), while that in the bare soil was in the order of MBN (11.5%)>DON (2.6%). Over all, microorganisms and mosses in the BSCs had 65.3% higher capacity of N fixation as compared with the bare soil. 3) The transferred amount and storage capacity of MB15N in the BSCs were 17.2 and 20.5 times higher than that in the bare soil, respectively. Accordingly, the turnover rate of MB15N in the BSCs and bare soil was 5.8 and 7.2 times per month, respectively, with the turnover time of MB15N in the BSCs being 1.2 times longer than that in bare soil. In conclusion, BSCs fix and hold more N than bare soil and change the distribution of each N fraction, implying that BSCs play a critical role in N cycling in dryland ecosystems.

SRR intronic variation inhibits expression of its neighbouring SMG6 gene and protects against temporal lobe epilepsy.

D-serine is a predominant N-methyl-D-aspartate receptor co-agonist with glutamate, and excessive activation of the receptor plays a substantial role in epileptic seizures. Serine racemase (SRR) is responsible for transforming L-serine to D-serine. In this study, we aimed to investigate the genetic roles of SRR and a neighbouring gene, nonsense-mediated mRNA decay factor (SMG6), in temporal lobe epilepsy (TLE). Here, a total of 496 TLE patients and 528 healthy individuals were successfully genotyped for three SRR tag single nucleotide polymorphisms. The frequencies of the GG genotype at rs4523957 T > G were reduced in the TLE cases in the initial cohort (cohort 1) and were confirmed in the independent cohort (cohort 2). An analysis of all TLE cases in cohort 1 + 2 revealed that the seizure frequency and drug-resistant incidence were significantly decreased in carriers of the GG genotype at rs4523957. Intriguingly, the activity of the SMG6 promoter with the mutant allele at rs4523957 decreased by 22% in the dual-luciferase assay, and up-regulated expression of SMG6 was observed in an epilepsy rat model. This study provides the first demonstration that the GG genotype is a protective marker against TLE. In particular, variation at rs4523957 likely inhibits SMG6 transcription and plays a key role against susceptibility to and severity of TLE. The significance of SMG6 hyperfunction in epileptic seizures deserves to be investigated in future studies.

Promoter Variants of the ADAM10 Gene and Their Roles in Temporal Lobe Epilepsy.

Previous evidence has indicated that downregulated ADAM10 gives rise to epileptic seizures in Alzheimer's disease, and this study investigated the association of ADAM10 with temporal lobe epilepsy (TLE) from a genetic perspective. A total of 496 TLE patients and 528 healthy individuals were enrolled and genotyped for ADAM10 promoter variants (rs653765 G > A and rs514049 A > C). The alleles, genotypes, and haplotypes were then compared with clarify the association of these variants with TLE and their impacts upon age at onset, initial seizure types before treatments, and responses to drug treatments. In cohorts I, II, and I + II, the frequencies of the A allele and AA genotype at rs514049 were consistently increased in the cases compared with the controls (p = 0.020 and p = 0.009; p = 0.008 and p = 0.009; p = 0.000 and p = 0.000; q = 0.003 and q = 0.002, respectively). In contrast, the frequency of the AC haplotype (rs653765-rs514049) decreased in cohorts I + II (p = 0.013). Further analyses of the TLE patients indicated that the AA genotype functioned as a predisposing factor to drug-resistant TLE and the AC haplotype as a protective factor against generalized tonic-clonic seizures (GTCS) and drug-resistant TLE. This study is the first to demonstrate an association of the ADAM10 promoter variants with TLE. In particular, the AA genotype and AC haplotype showed their effects upon GTCS and drug-resistant TLE.

[Expression of PI3K pathway proteins in refractory epilepsy associated with cortical malformation development].

OBJECTIVE: To investigate the expression of TSC1, TSC2, p-mTOR, p-4E-BP1, p-p70S6K and p-S6 in refractory epilepsy associated malformation of cortical development (MCD) tissues. METHODS: A total of 43 cases of refractory epilepsy were involved in the study, and all the patients were treated in Xuanwu Hospital during 2005 - 2008, including focal cortical dysplasia type IIa (11 cases) and type IIb (11 cases), tuberous sclerosis complex (10 cases) and ganalioglioma (11 cases), and other 12 cases were used as control. These cases were divided into 7 study groups and immunohistochemical EnVision method was used. To detect the location and intensity of TSC1, TSC2, p-mTOR, p-4E-BP1, p-p70S6K and p-S6 expression in every group. Then the Image-Pro Plus 6.0 image processing and analysis software were used to measure the number, area, integrating absorbance (IA) of positive cells in every samples. The statistical software SPSS 16.0 was used to analyze the data. RESULTS: The immunolocalization of TSC1 and TSC2 was similar. It could be observed the expression of various levels in the cytoplasm of dysmorphic neurons, balloon cells, giant cells, ganglioglioma cells and normal neurons. TSC1 staining in normal neurons was more notably than others but TSC2 staining in giant cells was weaker than other samples. p-mTOR mainly presented in giant cells, which could also be observed in astrocyte. P-4E-BP1 presented in the cytoplasm and nuclear membrane of balloon cells, giant cells and ganglioglioma cells, the staining of giant cells was stronger than balloon cells, but their staining were weaker than ganglioglioma cells. P-p70S6K mainly expressed in giant cells and less commonly presented in balloon cells. P-S6 typically presented in all abnormal glioneuronal cells and it nearly did not present in the normal neurons of N-CTX group. CONCLUSIONS: PI3K pathway, at least in part, involves in the occurrence of MCD, and may play an important role in the pathogenesis.

[Brain tumors in patients with intractable epilepsy: a clinicopathologic study of 35 cases].

OBJECTIVE: To study the clinicopathologic features of brain tumors occurring in patients with medically intractable epilepsy. METHODS: The clinical, radiologic and pathologic features of brain tumors occurring in 35 patients with intractable epilepsy encountered during the period from January, 2005 to April, 2008 in Xuanwu Hospital were retrospectively reviewed. RESULTS: The mean age of seizure onset and duration of disease were 14.3-year-old and 8.6 years, respectively. Abnormal signals were observed in 94.3% of cases (33/35) by magnetic resonance imaging. The histologic types of brain tumors included ganglioglioma (13/35, WHO grade I and 6/35, WHO grade II), dysembryoplastic neuroepithelial tumor (3/35, WHO grade I), pleomorphic xanthoastrocytoma (3/35, WHO grade II), diffuse astrocytoma (1/35, WHO grade II), oligoastrocytoma (1/35, WHO grade II), angiocentric glioma (1/35, WHO grade I) and meningioangiomatosis (1/35). The 6 remaining cases showed features seen in between glioneuronal hamartoma and mixed neuronal-glial tumor. Most of these tumors were located in the temporal lobe (27/35) and associated with focal cortical dysplasia. Immunohistochemical study showed a remarkable expression of CD34 in gangliogliomas. CONCLUSIONS: Brain tumors in patients with medically intractable epilepsy are almost always benign and located in the temporal lobe. Most of them represent mixed neuronal-glial tumors and some show transitional features in-between glioneuronal hamartoma and mixed neuronal-glial neoplasm. The similar morphologic pattern and biological behavior of glioneuronal hamartoma and mixed neuronal-glial tumor may suggest a common pathogenetic mechanism.