Texture analysis based on U-Net neural network for intracranial hemorrhage identification predicts early enlargement.

BACKGROUND AND OBJECTIVE: Early hemorrhage enlargement in hypertensive cerebral hemorrhage indicates a poor prognosis. This study aims to predict the early enlargement of cerebral hemorrhage through the intelligent texture analysis of cerebral hemorrhage after segmentation. METHODS: A total of 54 patients with hypertensive intracerebral hemorrhage were selected and divided into enlarged hematoma (enlarged group) and non-enlarged hematoma (negative group). The U-Net Neural network model and contour recognition were used to extract the brain parenchymal region, and Mazda texture analysis software was used to extract regional features. The texture features were reduced by Fisher coefficient (Fisher), classification error probability combined average correlation coefficients (POE + ACC), and mutual information (MI) to select the best feature parameters. B11 module was used to analyze the selected features. The misclassified rate of feature parameters screened by different dimensionality reduction methods was calculated. RESULTS: The neural network based on U-Net can accurately identify the lesion of cerebral hemorrhage. Among the 54 patients, 18 were in the enlarged group and 36 in the negative group. The parameters of gray level co-occurrence matrix and gray level run length matrix can be used to predict the enlargement of intracerebral hemorrhage. Among the features screened by Fisher, POE + ACC and MI, the texture features of MI showed the lowest misclassified rate, which was 0. CONCLUSION: The texture analysis based on U-Net neural network is helpful to predict the early expansion of hypertensive cerebral hemorrhage, and the parameters of gray level co-occurrence matrix and gray level run length matrix under MI dimensionality reduction have the most excellent predictive value.

Anatomic comparison of veins of Labbé between autopsy, digital subtraction angiography and computed tomographic venography.

OBJECTIVE: The drainage portion of the vein of Labbé varies, and it is difficult to predict whether the operation is likely to damage this vein. The aim of this study was to correlate the microanatomy of the vein of Labbé with digital subtraction angiography (DSA) and computed tomographic venography (CTV), in order to provide a basis for the preservation of the vein of Labbé during a supratentorial surgical approach. METHODS: A total of 30 human cadavers (60 sides) and 61 living patients (110 sides) were examined in this study. Each cadaver head was injected with blue latex via the superior sagittal sinus and the internal jugular veins. The venograms of each patient were obtained from the venous phases of DSA (60 sides for 36 patients) or CTV (50 sides for 25 patients). RESULTS: The patients were divided into four subgroups based on the location where a vein entered the dural sinus: the transverse sinus group, the tentorial group, the petrosal group, and the upper-transverse sinus group. The veins of Labbé in transverse sinus group and petrosal group directly entered dural sinus. The veins of Labbé in tentorial group and upper-transverse sinus group indirectly entered transverse sinus via the tentorium sinus or the upper-transverse sinus. These sinuses were meningeal veins running through two layers of the cerebral dura mater. The length of meningeal veins in these groups was 10.0 ± 7.2 mm. The veins of Labbé were mainly localized around the STP junction, which was the confluence of sigmoid sinus, transverse sinus, and superior petrosal sinus. The distance between the dural entrance of veins and the STP junction was 16.8 ± 10.2 mm. There was no significant difference in the results of the DSA and CTV examinations when compared to the observations in cadavers. CONCLUSIONS: Preoperative venograms are useful to design an individualized surgical approach for the preservation of the vein of Labbé. In general, the supratentorial median approach has the least chance to damage this vein. However, when preoperative venograms show that the vein of Labbé is too close to the confluence of sinuses or the meningeal veins are too long, an alternative approach should be chosen.

Computed tomographic angiography of the superficial cerebral venous anastomosis based on volume rendering, multi-planar reconstruction, and integral imaging display.

As damage to the superficial cerebral venous anastomosis may create catastrophic complications even after successful surgery, it is important to visualize and determine the normal features of the venous anastomosis with computed tomographic angiography. A total of 90 patients underwent a 64-detector row helical CT scan of head. The superficial cerebral venous anastomosis was reconstructed by volume rendering, multi-planar reconstruction, and integral display algorithm. In particular, we examined the vein of Trolard, the vein of Labbe, and the vein of Sylvian, in order to analyze the venous anastomosis. The superficial cerebral venous anastomosis varied across different individuals, and in this study, six types of anastomosis were found. In 28 % of patients, no venous anastomosis was found in the unilateral cerebral hemisphere. The display rate of the vein of Trolard, the vein of Labbe, and the vein of Sylvian in contributing to venous anastomosis was 70, 80, and 91 %, respectively. The number of vein of Trolard and vein of Labbe on the left side was greater than that of those on the right side. We implemented the 64-detector row helical CT as a rapid and noninvasive method to investigate the superficial cerebral venous anastomosis in our group of patients. We performed substantial image processing for the visualization of the superficial cerebral venous anastomosis; this would not only enable the early diagnosis of cerebral venous disease, but also protect the cerebral vein during neurosurgical intervention.

The association between three promoter polymorphisms of IL-1 and stroke: a meta-analysis.

AIM: The association between Interleukin-1 (IL-1) gene polymorphism and stroke remains controversial. The present study was designed to clarify this relationship through a pooled analysis of the numerous epidemiological studies focusing on this association. METHODS: Published data addressing the association between polymorphism of the IL-1 gene and stroke were selected from electronic databases. A total of 21 studies from 19 publications including 5280 stroke patients and 5699 controls were included in this meta-analysis which detect whether IL-1α-889C/T, IL-1ß-511C/T and IL-1 RN polymorphism were associated with stroke susceptibility. RESULT: The combined results of overall analysis revealed that there was a significant association between IL-1α-889C/T polymorphism and stroke (allele model: OR = 1.39, 95% CI = 1.14-1.68, P < 0.001; recessive model: OR = 1.56, 95% CI = 1.27-1.90, P < 0.001; dominant model: OR = 1.45, 95% CI = 1.13-1.87, P = 0.004; additive model: OR = 1.70, 95% CI = 1.37-2.10, P < 0.001). On a subgroup analysis by ethnicity of study population, significant association was found in Asians (allele model: OR = 1.25 95% CI = 1.12-1.39, P < 0.001; recessive model: OR = 1.36, 95% CI = 1.09-1.69, P = 0.007; dominant model: OR = 1.29, 95% CI = 1.12-1.48, P < 0.001; additive model: OR = 1.42, 95% CI = 1.13-1.80, P = 0.003). There are no associations of IL-1ß-511C/T and IL-RN polymorphisms on stroke risk were identified. CONCLUSION: In conclusion, our results suggested that IL-1α-889C/T polymorphism is associated with stroke risk, especially in Asians. However, no associations of IL-1ß-511C/T and IL-RN polymorphisms on stroke risk in overall analysis or subgroup analysis.

The association between three promoter polymorphisms of IL-10 and inflammatory bowel diseases (IBD): a meta-analysis.

AIM: To assess the relationship of the Interleukin-10 (IL-10) -1082G/A (rs1800896), -819C/T (rs1800871) and -592C/A (rs1800872) polymorphism with inflammatory bowel disease (IBD) by means of meta-analysis. METHODS: Published data addressing the association between polymorphism of the IL-10 with Crohn's disease (CD) and Ulcerative colitis (UC) were selected from electronic databases. A total of 17 studies including 4132 cases and 5109 controls were included in this meta-analysis which detected whether -1082G/A, -819C/T and -592C/A polymorphism were associated with CD or UC susceptibility. RESULT: The IL-10 -819C/T and -519C/A variant allele observed a significant association with UC (OR 1.16, 95%CI 1.03-1.31 and OR 1.19, 95%CI 1.03-1.38) not CD while there is no significant association between -1082G/A and UC or CD. CONCLUSION: The IL-10 -819C/T and -592C/A polymorphisms contribute to susceptibility to UC, but IL-10 -1082G/A polymorphism neither associated with CD nor UC.

Serial imaging changes of cerebral Whipple's disease: from onset to the end.

A case with multiple cerebral Whipple disease is reported. The neuroradiological evolution from multiple macrophagic encephalitis to diffuse cerebral infarct in this case is described.

[Protective effect of ligustrazine on insulin resistance after local cerebral ischemia of rat].

OBJECTIVE: To study the protective effect of Ligustrazine on IR after local cerebral ischemia of rat. METHOD: Models of rat IR after local cerebral ischemia were prepared by electrocagulation of the middle cerebral artery, and changes of serum insulin, tomer necrosis factor-alpha (TNF-alpha), plasma endothelin-1 (ET-1), nitric oxide (NO) and nitric oxide synthase (NOS) were observed 2 weeks after the ischemia. RESULT: Ligustrazine could significantly reduce serum insulin (P < 0.01), the content of plasma ET-1 (P < 0.01) and serum TNF-alpha (P < 0.01), the activity of brain tissue NO and NOS (P < 0.01). The drug also increased insulin sensitivity indexes (ISI). CONCLUSION: The protective effects of Ligustrazion on IR cerebral ischemia may be related to decreasing ET-1 content in plasma, TNF-alpha content in serum, NO content and NOS activities in tissue.