ABSTRACT
Fusarium head blight (FHB) is a destructive disease that affects wheat production. Detecting FHB accurately and rapidly is crucial for improving wheat yield. Traditional models are difficult to apply to mobile devices due to large parameters, high computation, and resource requirements. Therefore, this article proposes a lightweight detection method based on an improved YOLOv8s to facilitate the rapid deployment of the model on mobile terminals and improve the detection efficiency of wheat FHB. The proposed method introduced a C-FasterNet module, which replaced the C2f module in the backbone network. It helps reduce the number of parameters and the computational volume of the model. Additionally, the Conv in the backbone network is replaced with GhostConv, further reducing parameters and computation without significantly affecting detection accuracy. Thirdly, the introduction of the Focal CIoU loss function reduces the impact of sample imbalance on the detection results and accelerates the model convergence. Lastly, the large target detection head was removed from the model for lightweight. The experimental results show that the size of the improved model (YOLOv8s-CGF) is only 11.7 M, which accounts for 52.0% of the original model (YOLOv8s). The number of parameters is only 5.7 × 106 M, equivalent to 51.4% of the original model. The computational volume is only 21.1 GFLOPs, representing 74.3% of the original model. Moreover, the mean average precision (mAP@0.5) of the model is 99.492%, which is 0.003% higher than the original model, and the mAP@0.5:0.95 is 0.269% higher than the original model. Compared to other YOLO models, the improved lightweight model not only achieved the highest detection precision but also significantly reduced the number of parameters and model size. This provides a valuable reference for FHB detection in wheat ears and deployment on mobile terminals in field environments.
ABSTRACT
BACKGROUND: Invasive lung adenocarcinoma (LUAD) patients with the micropapillary (MPP) component tend to have extremely poor prognosis. To optimize clinical outcomes, a better understanding of specific concurrent gene alterations and their impact on the prognosis of patients with the MPP component is necessary. METHOD: A total of 621 Chinese patients with surgically resected invasive LUAD who underwent genetic testing for lung cancer were enrolled in this retrospective study. The genomic profiling of major lung cancer-related genes based on next-generation sequencing (NGS) was carried out on formalin-fixed paraffin-embedded tumor samples. RESULT: Among 621 patients with invasive LUAD, 154 (24.8%, 154/621) had the MPP component. We found that PIK3CA (4.5% vs 1.3%), KRAS (9.1% vs 4.7%), and ROS1 (2.6% vs 0.4%) were more frequent in patients with the MPP component than those without the MPP component (P < 0.05). The co-mutation occurred in 66 patients (10.6%, 66/621), of which 19 patients with the MPP component. Most of them were EGFR co-mutations (89.5%, 17/19), including EGFR and PIK3CA, EGFR and ERBB2, and other types. Patients with the MPP component who harbored EGFR co-mutations showed significantly worse recurrence-free survival (RFS) than single EGFR mutation (median RFS 20.1 vs 30.5 months; hazard ratio [HR]: 8.008; 95% confidence interval [CI]: 1.322-48.508). CONCLUSIONS: Patients with the MPP component harbored the co-mutation of driver genes had a higher risk of recurrence after surgery, especially in patients with EGFR co-mutation. EGFR co-mutation was a significant prognostic factor for RFS in patients with the MPP component.
Subject(s)
Adenocarcinoma of Lung , Lung Neoplasms , Humans , Retrospective Studies , Proto-Oncogene Proteins/genetics , Adenocarcinoma of Lung/genetics , Adenocarcinoma of Lung/surgery , Lung Neoplasms/genetics , Lung Neoplasms/surgery , Prognosis , Mutation , Class I Phosphatidylinositol 3-Kinases/genetics , ErbB Receptors/geneticsABSTRACT
Group member prototypicality is a factor in intergroup conflict-not all group members fight for group interests. This study focuses on the role of peripheral group members and the factors that influence their participation. We conducted two studies to examine the effects of group acceptance and self-uncertainty on the relationship between prototypicality and intergroup conflict. Results indicate that group acceptance moderates the relationship between prototypicality and intergroup conflict. Self-uncertainty moderates the effect of the interaction between prototypicality and group acceptance on intergroup conflict. Our findings have theoretical and practical implications for intergroup conflict resolution.
ABSTRACT
The widespread application of copper (Cu) -based fertilizers and pesticides could increase the accumulation of Cu in kiwifruit. According to a global survey, red- and yellow-fleshed kiwifruit contained more elevated amounts of Cu than green-fleshed kiwifruit due to weaker disease resistance and higher use of Cu pesticides. Intriguingly, our research revealed that external and endogenous ascorbic acid (AsA) reduced the phenotypic and physiological injury of Cu toxicity in kiwifruit. Cu stress assays and transcriptional analysis have shown that Cu treatment for 12 h significantly increased the AsA content in kiwifruit leaves and up-regulated key genes involved in AsA biosynthesis, such as GDP-L-galactose phosphorylase3 (GGP3) and GDP-mannose-3',5'-epimerase (GME). Overexpressing GGP3 in transgenic kiwifruit significantly increased the endogenous AsA content of kiwifruit, which was beneficial in mitigating Cu toxicity by decreasing levels of reactive oxygen species, malondialdehyde, and electrolyte leakage, as well as reducing damage to the chloroplast structure and photosystem II. This study presented a novel strategy to ameliorate plant Cu stress by increasing the endogenous antioxidant (AsA) content through transgenesis.
Subject(s)
Copper , Pesticides , Copper/toxicity , Ascorbic Acid/pharmacology , Biological Assay , ChloroplastsABSTRACT
In the sulfotransferase (SULT) superfamily, members of the SULT1 family mainly catalyse the sulfonation reaction of phenolic compounds, which is involved in the phase II metabolic detoxification process and plays a key role in endocrine homeostasis. A coding variant rs1059491 in the SULT1A2 gene has been reported to be associated with childhood obesity. This study aimed to investigate the association of rs1059491 with the risk of obesity and cardiometabolic abnormalities in adults. This caseâcontrol study included 226 normal weight, 168 overweight and 72 obese adults who underwent a health examination in Taizhou, China. Genotyping of rs1059491 was performed by Sanger sequencing in exon 7 of the SULT1A2 coding region. Chi-squared tests, one-way ANOVA, and logistic regression models were applied. The minor allele frequencies of rs1059491 in the overweight combined with obesity and control groups were 0.0292 and 0.0686, respectively. No differences in weight and body mass index were detected between the TT genotype and GT + GG genotype under the dominant model, but the levels of serum triglycerides were significantly lower in G-allele carriers than in non-G-allele carriers (1.02 (0.74-1.32) vs. 1.35 (0.83-2.13) mmol/L, P = 0.011). The GT + GG genotype of rs1059491 versus the TT genotype reduced the risk of overweight and obesity by 54% (OR 0.46, 95% CI 0.22-0.96, P = 0.037) after adjusting for sex and age. Similar results were observed for hypertriglyceridaemia (OR 0.25, 95% CI 0.08-0.74, P = 0.013) and dyslipidaemia (OR 0.37, 95% CI 0.17-0.83, P = 0.015). However, these associations disappeared after correction for multiple tests. This study revealed that the coding variant rs1059491 is nominally associated with a decreased risk of obesity and dyslipidaemia in southern Chinese adults. The findings will be validated in larger studies including more detailed information on genetic background, lifestyle and weight change with age.
Subject(s)
Arylsulfotransferase , Dyslipidemias , Obesity , Overweight , Adult , Humans , Alleles , Arylsulfotransferase/genetics , Body Mass Index , Case-Control Studies , Dyslipidemias/genetics , East Asian People , Genotype , Overweight/genetics , Polymorphism, Single Nucleotide , Obesity/geneticsABSTRACT
The high mortality rate associated with gastric cancer (GC) has resulted in an urgent need to identify novel therapeutic targets for GC. This study aimed to investigate whether GAIP interacting protein, C terminus 1 (GIPC1) represents a therapeutic target and its regulating mechanism in GC. GIPC1 expression was elevated in GC tissues, liver metastasis tissues, and lymph node metastases. GIPC1 knockdown or GIPC1 blocking peptide blocked the platelet-derived growth factor receptor (PDGFR)/PI3K/AKT signaling pathway, and inhibited the proliferation and migration of GC cells. Conversely, GIPC1 overexpression markedly activated the PDGFR/PI3K/AKT signaling pathway, and promoted GC cell proliferation and migration. Furthermore, platelet-derived growth factor subunit BB (PDGF-BB) cytokines and the AKT inhibitor attenuated the effect of differential GIPC1 expression. Moreover, GIPC1 silencing decreased tumor growth and migration in BALB/c nude mice, while GIPC1 overexpression had contrasting effects. Taken together, our findings suggest that GIPC1 functions as an oncogene in GC and plays a central role in regulating cell proliferation and migration via the PDGFR/PI3K/AKT signaling pathway.
Subject(s)
Stomach Neoplasms , Humans , Animals , Mice , Stomach Neoplasms/genetics , Phosphatidylinositol 3-Kinases , Proto-Oncogene Proteins c-akt , Mice, Nude , Signal Transduction , Adaptor Proteins, Signal TransducingABSTRACT
Since the ambiguous boundary of the lesion and inter-observer variability, white matter hyperintensity segmentation annotations are inherently noisy and uncertain. On the other hand, the high capacity of deep neural networks (DNN) enables them to overfit labels with noise and uncertainty, which may lead to biased models with weak generalization ability. This challenge has been addressed by leveraging multiple annotations per image. However, multiple annotations are often not available in a real-world scenario. To mitigate the issue, this paper proposes a supervision augmentation method (SA) and combines it with ensemble learning (SA-EN) to improve the generalization ability of the model. SA can obtain diverse supervision information by estimating the uncertainty of annotation in a real-world scenario that per image have only one ambiguous annotation. Then different base learners in EN are trained with diverse supervision information. The experimental results on two white matter hyperintensity segmentation datasets demonstrate that SA-EN gets the optimal accuracy compared with other state-of-the-art ensemble methods. SA-EN is more effective on small datasets, which is more suitable for medical image segmentation with few annotations. A quantitative study is presented to show the effect of ensemble size and the effectiveness of the ensemble model. Furthermore, SA-EN can capture two types of uncertainty, aleatoric uncertainty modeled in SA and epistemic uncertainty modeled in EN.
ABSTRACT
Whole-brain segmentation from T1-weighted magnetic resonance imaging (MRI) is an essential prerequisite for brain structural analysis, e.g., locating morphometric changes for brain aging analysis. Traditional neuroimaging analysis pipelines are implemented based on registration methods, which involve time-consuming optimization steps. Recent related deep learning methods speed up the segmentation pipeline but are limited to distinguishing fuzzy boundaries, especially encountering the multi-grained whole-brain segmentation task, where there exists high variability in size and shape among various anatomical regions. In this article, we propose a deep learning-based network, termed Multi-branch Residual Fusion Network, for the whole brain segmentation, which is capable of segmenting the whole brain into 136 parcels in seconds, outperforming the existing state-of-the-art networks. To tackle the multi-grained regions, the multi-branch cross-attention module (MCAM) is proposed to relate and aggregate the dependencies among multi-grained contextual information. Moreover, we propose a residual error fusion module (REFM) to improve the network's representations fuzzy boundaries. Evaluations of two datasets demonstrate the reliability and generalization ability of our method for the whole brain segmentation, indicating that our method represents a rapid and efficient segmentation tool for neuroimage analysis.
ABSTRACT
Neuroimaging-driven brain age estimation has become popular in measuring brain aging and identifying neurodegenerations. However, the single estimated brain age (gap) compromises regional variations of brain aging, losing spatial specificity across diseases which is valuable for early screening. In this study, we combined brain age modeling with Shapley Additive Explanations to measure brain aging as a feature contribution vector underlying spatial pathological aging mechanism. Specifically, we regressed age with volumetric brain features using machine learning to construct the brain age model, and model-agnostic Shapley values were calculated to attribute regional brain aging for each subject's age estimation, forming the brain age vector. Spatial specificity of the brain age vector was evaluated among groups of normal aging, prodromal Parkinson disease (PD), stable mild cognitive impairment (sMCI), and progressive mild cognitive impairment (pMCI). Machine learning methods were adopted to examine the discriminability of the brain age vector in early disease screening, compared with the other two brain aging metrics (single brain age gap, regional brain age gaps) and brain volumes. Results showed that the proposed brain age vector accurately reflected disorder-specific abnormal aging patterns related to the medial temporal and the striatum for prodromal AD (sMCI vs. pMCI) and PD (healthy controls [HC] vs. prodromal PD), respectively, and demonstrated outstanding performance in early disease screening, with area under the curves of 83.39% and 72.28% in detecting pMCI and prodromal PD, respectively. In conclusion, the proposed brain age vector effectively improves spatial specificity of brain aging measurement and enables individual screening of neurodegenerative diseases.
Subject(s)
Alzheimer Disease , Cognitive Dysfunction , Neurodegenerative Diseases , Humans , Alzheimer Disease/pathology , Magnetic Resonance Imaging/methods , Brain/diagnostic imaging , Brain/pathology , Cognitive Dysfunction/pathology , Aging/pathology , Neurodegenerative Diseases/diagnostic imaging , Neurodegenerative Diseases/pathologyABSTRACT
Respiratory syncytial virus (RSV) is the most common pathogen causing acute lower respiratory tract infection (LRTI) in children. RSV usually peaks in winter and declines by early spring in China. The outbreak of coronavirus disease 2019 (COVID-19) was reported to bring changes to the transmission pattern of respiratory pathogens including RSV. Here in this paper, we analyzed RSV-positive nasopharyngeal aspirates from inpatients in the Children's Hospital of Fudan University from October 2019 to October 2021 and compared the clinical features of the RSV-positive patients before and during COVID-19. We found an atypical upsurge of RSV infection in the late summer of 2021 after a major suppression in 2020. RSV B was the main subtype spreading among children throughout the study. Phylogenetic analysis revealed that all RSV A strains belonged to ON1 genotype and all RSV B strains were BA9 genotype. Deduced amino acid analysis displayed different substitutions in the RSV strains observed before and during COVID-19. Demographic analysis suggested that males and infants aged under 5 months were the main populations infected with RSV by gender and age, respectively. Less severe clinical outcomes were observed in patients during COVID-19 than before the pandemic, especially in RSV B-positive patients. Our findings described the epidemiological changes in RSV infection brought by COVID-19, which further underscored the importance of continuous surveillance of RSV in the shadow of COVID-19 at both local and global scales.
ABSTRACT
BACKGROUND: Maffucci syndrome (MS) is a rare, nonhereditary congenital mesodermal dysplasia characterized by multiple enchondromas and hemangiomas, associated with an increased risk of developing malignant tumors. Given their rarity, the pathogenesis of these tumors has not been clarified, and there is no standard treatment. CASE PRESENTATION: We present a case of a 45-year-old man with MS to supplement the clinical manifestations and explore the molecular mechanism of MS. The patient underwent amputation surgery to inhibit tumor development and was diagnosed with MS with 1-2 grade giant chondrosarcoma in the left ankle. In addition, the whole exon analysis results revealed isocitrate dehydrogenase 1 (IDH1) R132C mutation in chondrosarcoma lesions but not in blood DNA. CONCLUSIONS: This case report showed MS complicated by giant chondrosarcoma in the left ankle with an IDH1 R132C mutation, which is appropriate to monitor the development of MS pathology and other concomitant lesions.
Subject(s)
Bone Neoplasms , Chondrosarcoma , Enchondromatosis , Ankle/pathology , Bone Neoplasms/complications , Bone Neoplasms/genetics , Bone Neoplasms/surgery , Chondrosarcoma/complications , Chondrosarcoma/genetics , Chondrosarcoma/surgery , Enchondromatosis/complications , Enchondromatosis/genetics , Enchondromatosis/surgery , Humans , Isocitrate Dehydrogenase/genetics , Male , Middle Aged , MutationABSTRACT
Objectives: [18F]9-fluoropropyl-(+)-dihydrotetrabenazine ([18F]-FP-DTBZ) positron emission tomography (PET) provides reliable information for the diagnosis of Parkinson's disease (PD). In this study, we proposed a multi-atlas-based [18F]-FP-DTBZ PET image segmentation method for PD quantification assessment. Methods: A total of 99 subjects from Xuanwu Hospital of Capital Medical University were included in this study, and both brain PET and magnetic resonance (MR) scans were conducted. Data from 20 subjects were used to generate atlases, based on which a multi-atlas-based [18F]-FP-DTBZ PET segmentation method was developed especially for striatum and its subregions. The proposed method was compared with the template-based method through striatal subregion parcellation performance and the standard uptake value ratio (SUVR) quantification accuracy. Discriminant analysis between healthy controls (HCs) and PD patients was further performed. Results: Segmentation results of the multi-atlas-based method showed better consistency than the template-based method with the ground truth, yielding a dice coefficient of 0.81 over 0.73 on the full striatum. The SUVRs calculated by the multi-atlas-based method had an average interclass correlation coefficient (ICC) of 0.953 with the standardized result, whereas the template-based method only reached 0.815. The SUVRs of HCs were generally higher than that of patients with PD and showed significant differences in all of the striatal subregions (all p < 0.001). The median and posterior putamen performed best in discriminating patients with PD from HCs. Conclusion: The proposed multi-atlas-based [18F]-FP-DTBZ PET image segmentation method achieved better performance than the template-based method, indicating great potential in improving accuracy and efficiency for PD diagnosis in clinical routine.
ABSTRACT
The brain activities and the underlying wiring diagrams are vulnerable in multiple sclerosis (MS). Also, it remains unknown whether the complex coupling between these functional and structural brain properties would be affected. To address this issue, we adopted graph frequency analysis to quantify the high-order structural-functional interactions based on a combination of brain diffusion and functional MRI data. The structural-functional decoupling index was proposed to measure how much brain regional functional activity with different graph frequency was organized atop the underlying wiring diagram in MS. The identified patterns in MS included (1) disruption of inherent structural-functional coupling in the somatomotor network (ß = 0.05, p = 0.03), and (2) excessive decrease of decoupling in the subcortical (ß = -0.10, p = 0.02), visual (ß = -0.04, p = 0.01), and dorsal attention networks (ß = -0.12, p = 0.03). Besides, this structural-functional coupling signature in the somatomotor network was associated with cognitive worsening of MS patients (ß = -24.31, p = 0.006). Overall, our study unveiled a unique signature of brain structural-functional reorganization in MS.
Subject(s)
Multiple Sclerosis , Brain/diagnostic imaging , Brain Mapping , Humans , Magnetic Resonance Imaging , Multiple Sclerosis/diagnostic imaging , Nerve Net/diagnostic imagingABSTRACT
BACKGROUND: Seeds are the economic basis of oilseed crops, especially soybeans, the most widely cultivated oilseed crop worldwide. Seed development is accompanied by a multitude of diverse cellular processes, and revealing the underlying regulatory activities is critical for seed improvement. RESULTS: In this study, we profiled the transcriptomes of developing seeds at 20, 25, 30, and 40 days after flowering (DAF), as these stages represent critical time points of seed development from early to full development. We identified a set of highly abundant genes and highlighted the importance of these genes in supporting nutrient accumulation and transcriptional regulation for seed development. We identified 8925 differentially expressed genes (DEGs) that exhibited temporal expression patterns over the course and expression specificities in distinct tissues, including seeds and nonseed tissues (roots, stems, and leaves). Genes specific to nonseed tissues might have tissue-associated roles, with relatively low transcript abundance in developing seeds, suggesting their spatially supportive roles in seed development. Coexpression network analysis identified several underexplored genes in soybeans that bridge tissue-specific gene modules. CONCLUSIONS: Our study provides a global view of gene activities and biological processes critical for seed formation in soybeans and prioritizes a set of genes for further study. The results of this study help to elucidate the mechanism controlling seed development and storage reserves.
Subject(s)
Gene Expression Regulation, Plant , Glycine max , Gene Expression Profiling , Seeds/genetics , Glycine max/genetics , TranscriptomeABSTRACT
Pinpointing the brain dysconnectivity in idiopathic rapid eye movement sleep behaviour disorder (iRBD) can facilitate preventing the conversion of Parkinson's disease (PD) from prodromal phase. Recent neuroimage investigations reported disruptive brain white matter connectivity in both iRBD and PD, respectively. However, the intrinsic process of the human brain structural network evolving from iRBD to PD still remains largely unknown. To address this issue, 151 participants including iRBD, PD and age-matched normal controls were recruited to receive diffusion MRI scans and neuropsychological examinations. The connectome-wide association analysis was performed to detect reorganization of brain structural network along with PD progression. Eight brain seed regions in both cortical and subcortical areas demonstrated significant structural pattern changes along with the progression of PD. Applying machine learning on the key connectivity related to these seed regions demonstrated better classification accuracy compared to conventional network-based statistic. Our study shows that connectome-wide association analysis reveals the underlying structural connectivity patterns related to the progression of PD, and provide a promising distinct capability to predict prodromal PD patients.
Subject(s)
Connectome , Parkinson Disease , REM Sleep Behavior Disorder , Brain/diagnostic imaging , Humans , Magnetic Resonance Imaging , Parkinson Disease/diagnostic imagingABSTRACT
MAIN CONCLUSION: QTL mapping of stem diameter was carried out in three RIL populations using a high-density genetic map, and candidate genes related to stem diameter were predicted. Stem diameter is an important agronomic trait affecting soybean lodging and productivity. However, this trait is underexploited, and the underlying genetic mechanism in soybean remains unclear. In this study, three recombinant inbred line (RIL) populations, including 156 F10 lines from Nannong 94-156 × Bogao (N × B), 127 F9 lines from Dongnong 50 × Williams 82 (D × W), and 146 F9 lines from Suinong 14 × Enrei (S × E), were used to identify QTLs for soybean stem diameter across multiple environments. Phenotype analysis revealed that stem diameter exhibited strong positive correlations with plant height and 100-seed weight, two of the most important yield components. A total of 12 QTLs for stem diameter were identified on eight chromosomes across three RIL populations and five environments. The most influential QTL that was stably identified across all the populations and environments, q11, explained 12.58-26.63% of the phenotypic variation. Detection of several environment-specific QTLs, including q14, q16, and q20, suggests that environments may also have important effects in shaping the natural variation in soybean stem diameter. Furthermore, we predicted candidate genes underlying the QTLs and found that several promising candidate genes may be responsible for the variation in stem diameter in soybean. Overall, the markers/genes linked closely or underlying the major QTLs may be used for marker-assisted selection of soybean varieties to enhance lodging resistance and even yield. Our results lay the foundation for the fine mapping of stem development-related genes to reveal the molecular mechanisms.
Subject(s)
Glycine max , Quantitative Trait Loci , Chromosome Mapping , Genetic Linkage , Phenotype , Quantitative Trait Loci/genetics , Seeds , Glycine max/geneticsABSTRACT
Recent studies have confirmed that white matter hyperintensities (WMHs) accumulated in strategic brain regions can predict cognitive impairments associated with Alzheimer's disease (AD). The knowledge of white matter anatomy facilitates lesion-symptom mapping associated with cognition, and provides important spatial information for lesion segmentation algorithms. However, deep learning-based methods in the white matter hyperintensity (WMH) segmentation realm do not take full advantage of anatomical knowledge in decision-making and lesion localization processes. In this paper, we proposed an anatomical knowledge-based MRI deep learning pipeline (AU-Net), handcrafted anatomical-based spatial features developed from brain atlas were integrated with a well-designed U-Net configuration to simultaneously segment and locate WMHs. Manually annotated data from WMH segmentation challenge were used for the evaluation. We then applied this pipeline to investigate the association between WMH burden and cognition within another publicly available database. The results showed that AU-Net significantly improved segmentation performance compared with methods that did not incorporate anatomical knowledge (p < 0.05), and achieved a 14-17% increase based on area under the curve (AUC) in the cohort with mild WMH burden. Different configurations for incorporating anatomical knowledge were evaluated, the proposed stage-wise AU-Net-two-step method achieved the best performance (Dice: 0.86, modified Hausdorff distance: 3.06 mm), which was comparable with the state-of-the-art method (Dice: 0.87, modified Hausdorff distance: 3.62 mm). We observed different WMH accumulation patterns associated with normal aging and cognitive impairments. Furthermore, the characteristics of individual WMH lesions located in strategic regions (frontal and parietal subcortical white matter, as well as corpus callosum) were significantly correlated with cognition after adjusting for total lesion volumes. Our findings suggest that AU-Net is a reliable method to segment and quantify brain WMHs in elderly cohorts, and is valuable in individual cognition evaluation.
Subject(s)
Cognitive Dysfunction , Deep Learning , White Matter , Aged , Brain/diagnostic imaging , Cognitive Dysfunction/diagnostic imaging , Humans , Magnetic Resonance Imaging , White Matter/diagnostic imagingABSTRACT
White matter hyperintensities (WMH) are important biomarkers for cerebral small vessel disease and closely associated with other neurodegenerative process. In this paper, we proposed a fully automatic WMH segmentation method based on U-net architecture. CRF were combined with U-net to refine segmentation results. We used a new anatomical based spatial feature produced by brain tissue segmentation based on T1 image, along with intensities of T1 and T2-FLAIR images to train our neural network. We compared 8 forms of automated WMH segmentation methods, range from traditional statistical learnng methods to deep learning based methods, with different architecture and used different features. Results showed our proposed method achieved best performance in terms of most metrics, and the inclusion of anatomical based spatial features strongly increase the segmentation performance.