BACKGROUND:Rapid and accurate segmentation of brain tissue in medical images is of great significance for three-dimensional biomechanical modeling and diagnosis of craniocerebral injuries.Currently,artificial intelligence(AI)-based baseline models exhibit excellent generalization capabilities on large-scale datasets.However,due to the specificity and complexity of craniocerebral tissues,these models have certain limitations in their application to craniocerebral tissue segmentation.Additionally,the scarcity of craniocerebral tissue samples makes it difficult for baseline models to achieve precise segmentation results through fine-tuning.OBJECTIVE:To construct a craniocerebral tissue segmentation model based on texture feature retrieval enhancement to improve segmentation accuracy under a small number of samples.METHODS:Segment Anything in Medical Images(MedSAM)model was selected as the basic framework,and texture features were combined with deep learning to build a brain tissue segmentation model based on texture feature retrieval enhancement(DP-MedSAM).Dice Coefficient and mean intersection over union(MIoU)were selected to evaluate the efficiency of image segmentation results.In comparison with the original MedSAM model,the ablation experiment systematically evaluated the influence of key components on the model performance.The sensitivities of MedSAM,the Segment Anything Model(SAM)for medical image segmentation(SAM-Med2D)and DP-MedSAM in the mandible,left optic nerve,and left parotid gland were compared.RESULTS AND CONCLUSION:(1)By verifying the impact of the number of point prompts on segmentation results on the HaN-Seg dataset,the experimental results indicated that the optimal Dice score was achieved with the addition of three points.(2)DP-MedSAM demonstrated performance improvements compared with MedSAM and SAM-Med2D on two datasets(HaN and Public Domain Database for Computational Anatomy).Especially on the Public Domain Database for Computational Anatomy dataset,in terms of the MIoU metric,DP-MedSAM outperformed MedSAM by 6.59％and SAM-Med2D by 37.35％;in terms of the Dice metric,DP-MedSAM outperformed MedSAM and SAM-Med2D by 4.34％and 25.32％,respectively.(3)The ablation experiment results showed that removing the texture feature extraction module in the DP-MedSAM model,relying solely on original image features,led to a significant decrease in results on the test set.Furthermore,removing the vector cache database and its retrieval enhancement function from the model,which deprived the ability of the model to perform similarity retrieval using an external knowledge base,further reduced model performance.(4)Under conditions of limited data resources,the DP-MedSAM model outperformed the other two models in all evaluation metrics.The DP-MedSAM model performed excellently when processing simple and moderately difficult samples,demonstrating a clear advantage over the other two models and indicating good generalization ability.Processing the fine structures of difficult samples placed higher demands on the model's segmentation capabilities.Although the performance of the DP-MedSAM model declined slightly,it still outperformed the other two models.(5)This study proposes an innovative craniocerebral tissue segmentation model,DP-MedSAM,which improves the baseline model's performance in capturing local details and global structural information in medical images by introducing target region texture feature extraction.Through vector similarity retrieval technology,DP-MedSAM can retrieve the feature vector most similar to the current target region from a pre-constructed vector database,providing more precise guiding information for the segmentation process.

BACKGROUND:Rapid and accurate segmentation of brain tissue in medical images is of great significance for three-dimensional biomechanical modeling and diagnosis of craniocerebral injuries.Currently,artificial intelligence(AI)-based baseline models exhibit excellent generalization capabilities on large-scale datasets.However,due to the specificity and complexity of craniocerebral tissues,these models have certain limitations in their application to craniocerebral tissue segmentation.Additionally,the scarcity of craniocerebral tissue samples makes it difficult for baseline models to achieve precise segmentation results through fine-tuning.OBJECTIVE:To construct a craniocerebral tissue segmentation model based on texture feature retrieval enhancement to improve segmentation accuracy under a small number of samples.METHODS:Segment Anything in Medical Images(MedSAM)model was selected as the basic framework,and texture features were combined with deep learning to build a brain tissue segmentation model based on texture feature retrieval enhancement(DP-MedSAM).Dice Coefficient and mean intersection over union(MIoU)were selected to evaluate the efficiency of image segmentation results.In comparison with the original MedSAM model,the ablation experiment systematically evaluated the influence of key components on the model performance.The sensitivities of MedSAM,the Segment Anything Model(SAM)for medical image segmentation(SAM-Med2D)and DP-MedSAM in the mandible,left optic nerve,and left parotid gland were compared.RESULTS AND CONCLUSION:(1)By verifying the impact of the number of point prompts on segmentation results on the HaN-Seg dataset,the experimental results indicated that the optimal Dice score was achieved with the addition of three points.(2)DP-MedSAM demonstrated performance improvements compared with MedSAM and SAM-Med2D on two datasets(HaN and Public Domain Database for Computational Anatomy).Especially on the Public Domain Database for Computational Anatomy dataset,in terms of the MIoU metric,DP-MedSAM outperformed MedSAM by 6.59％and SAM-Med2D by 37.35％;in terms of the Dice metric,DP-MedSAM outperformed MedSAM and SAM-Med2D by 4.34％and 25.32％,respectively.(3)The ablation experiment results showed that removing the texture feature extraction module in the DP-MedSAM model,relying solely on original image features,led to a significant decrease in results on the test set.Furthermore,removing the vector cache database and its retrieval enhancement function from the model,which deprived the ability of the model to perform similarity retrieval using an external knowledge base,further reduced model performance.(4)Under conditions of limited data resources,the DP-MedSAM model outperformed the other two models in all evaluation metrics.The DP-MedSAM model performed excellently when processing simple and moderately difficult samples,demonstrating a clear advantage over the other two models and indicating good generalization ability.Processing the fine structures of difficult samples placed higher demands on the model's segmentation capabilities.Although the performance of the DP-MedSAM model declined slightly,it still outperformed the other two models.(5)This study proposes an innovative craniocerebral tissue segmentation model,DP-MedSAM,which improves the baseline model's performance in capturing local details and global structural information in medical images by introducing target region texture feature extraction.Through vector similarity retrieval technology,DP-MedSAM can retrieve the feature vector most similar to the current target region from a pre-constructed vector database,providing more precise guiding information for the segmentation process.

BACKGROUND:It has been confirmed that the mixing of decellularized matrix and polymer electrospinning can not only improve the structural properties of fibers,but also preserve the biological decellularized of decellularized matrix.However,there is no relevant report on the preparation of skin tissue engineering scaffolds by electrospinning polyurethane and decellularized skin matrix. OBJECTIVE:To investigate the reparative effect of a decellularized skin matrix/polyurethane blended fibrous scaffold on rat skin defects. METHODS:Polyurethane electrospun fibrous scaffold and decellularized skin matrix/polyurethane blended fibrous scaffold were fabricated using the electrospinning technique.The fiber structure was observed under scanning electron microscope.Rat adipose mesenchymal stem cells were inoculated on two kinds of scaffolds respectively.The morphology of the scaffolds was observed under scanning electron microscope.Three full-thickness skin defects of 1 cm×1 cm were fabricated on the back of 10 SD rats.Polyurethane electrospun fibrous scaffolds(control group)and decellularized skin matrix/polyurethane blended fibrous scaffolds(experimental group)were implanted in two of the defects,and no material was implanted in the remaining defects(blank control group).The skin wound healing was observed at 1,2,and 3 weeks after operation.At 3 weeks after implantation,the wound was stained with hematoxylin and eosin and the scar area was calculated. RESULTS AND CONCLUSION:(1)Under scanning electron microscope,the two kinds of electrospun fibers were reticulated,and the rat adipose mesenchymal stem cells attached to the fibers on the two kinds of scaffolds,and the adhesion was good.(2)With the extension of the postoperative time,the skin wounds of each group gradually healed.By week 3 after the operation,the skin wounds of the experimental group and the control group were basically healed,and small ulcers could be seen on the wounds of the blank control group.Hematoxylin-eosin staining of skin wounds showed that the epidermal coverage of the wound was basically complete in the control group and the experimental group,and fibroblast growth and inflammatory cell infiltration could be seen in the dermis.In addition,the collagen fibers of the wound in the experimental group were abundant and arranged in a regular order,basically parallel to the epidermal surface.The wound epidermis of blank control group was still defective.The scar area of the experimental group was smaller than that of the other two groups(P<0.05,P<0.01).(3)These results indicate that the decellularized skin matrix/polyurethane blended fibrous scaffold can effectively repair full-thickness skin defects and improve scar formation in rats.

Objective:To investigate the impact of early invasive arterial blood pressure (IBP) monitoring on survival and neurological outcomes in out-of-hospital cardiac arrest (OHCA) patients undergoing extracorporeal cardiopulmonary resuscitation (ECPR).Methods:This retrospective cohort study analyzed 44 OHCA patients receiving ECPR between January 2021 and January 2023. Patients were divided into: Early intervention group : IBP established within 3 min of ECMO initiation; Late intervention group : IBP established after ICU admission. Baseline characteristics, ECMO parameters, and clinical outcomes were compared. Multivariable logistic regression (adjusted for age, initial rhythm, etc.) and Spearman's correlation were used.Results:This study included a total of 44 patients treated with OHCA and ECPR, divided into an early intervention group of 23 cases and a late intervention group of 21 cases. The early intervention group showed significantly higher: Survival to discharge (43.5% vs. 9.5%, P<0.05), Good neurological recovery (CPC 1-2: 34.8% vs. 9.5%, P<0.05).Early intervention independently predicted survival (adjusted OR=18.84, 95% CI:1.97-179.98, P=0.01). Stratified analysis by pH (cutoff 7.0) demonstrated consistent benefits in both pH>7.0 ( aOR=0.392, 95% CI:0.106-0.678) and pH≤7.0 subgroups ( aOR=0.385, 95% CI: 0.075-0.695; interaction P=0.183). Early IBP positively correlated with CPC scores ( ρ=0.40, P=0.007). Conclusions:Early IBP monitoring significantly improves survival and neurological outcomes in OHCA-ECPR patients, supporting its integration into standardized protocols.

OBJECTIVE: This study aimed to investigate the impact of glycemic control and diabetes duration on subsequent myocardial infarction (MI) in patients with both coronary heart disease (CHD) and type 2 diabetes (T2D). METHODS: We conducted a retrospective cohort study of 33,238 patients with both CHD and T2D in Shenzhen, China. Patients were categorized into 6 groups based on baseline fasting plasma glucose (FPG) levels and diabetes duration (from the date of diabetes diagnosis to the baseline date) to examine their combined effects on subsequent MI. Cox proportional hazards regression models were used, with further stratification by age, sex, and comorbidities to assess potential interactions. RESULTS: Over a median follow-up of 2.4 years, 2,110 patients experienced MI. Compared to those with optimal glycemic control (FPG < 6.1 mmol/L) and shorter diabetes duration (< 10 years), the fully-adjusted hazard ratio ( HR) (95% Confidence Interval [95% CI]) for those with a diabetes duration of ≥ 10 years and FPG > 8.0 mmol/L was 1.93 (95% CI: 1.59, 2.36). The combined effects of FPG and diabetes duration on MI were largely similar across different age, sex, and comorbidity groups, although the excess risk of MI associated with long-term diabetes appeared to be more pronounced among those with atrial fibrillation. CONCLUSION Our study indicates that glycemic control and diabetes duration significant influence the subsequent occurrence of MI in patients with both CHD and T2D. Tailored management strategies emphasizing strict glycemic control may be particularly beneficial for patients with longer diabetes duration and atrial fibrillation.
Humans ; Diabetes Mellitus, Type 2/blood* ; Male ; Female ; Middle Aged ; Aged ; Coronary Disease/complications* ; Myocardial Infarction/etiology* ; Retrospective Studies ; China/epidemiology* ; Glycemic Control ; Blood Glucose ; Adult ; Risk Factors ; Time Factors

Objective:To construct a visual analysis system for the operation efficiency of medical equipment,improve the quality of medical services while improving the management level and efficiency of medical equipment.Method:Using the hospital LAN and Internet of Things technology,using advanced indoor positioning technology,three-dimensional high-precision map technology,installing active Bluetooth technology tags in movable high-value medical examination equipment,using LoRa long-distance wireless communication technology,real-time collection of equipment operation status,use frequency,fault information and other data,docking the hospital's business system,with the help of data visualization technology,designed to be able to real-time data,Historical data is integrated into a visual medical equipment operation efficiency analysis system that analyzes graphical interfaces in different dimensions.Results:The system realized the tracking,positioning,monitoring and traceability of the operation status and real-time movement trajectory of medical equipment through maps,ring charts,bar charts,line charts,stand-alone operation status charts,equipment operation status rankings,etc.,providing multi-dimensional,accurate and real-time equipment operation information for the use and management of medical equipment to ensure the high-quality development of the hospital.Conclusion:Through the accurate analysis of the operation efficiency of medical equipment,hospitals can better control medical costs and improve economic and social benefits.

Understory planting of medicinal plants is a new planting mode that connects Chinese herbal medicine(CHM) with forest resources.The complex and variable understory environmental factors will inevitably affect the yield and quality of understory CHM.This research summarized the research progress on understory planting of medicinal plants based on forest types and environmental factors within the forest from the perspectives of understory light, air temperature and humidity, soil characteristics, and the interaction between crops within the forest.The results showed that the complex and variable light, temperature and humidity, and soil factors(such as fertility, acidity and alkalinity, and microorganisms) under the forest could affect the yield and quality of medicinal plants to varying degrees through physiological activities such as photosynthesis and respiration, resulting in a significant increase or decrease in yield and quality compared to open field cultivation.In addition, the competition or mutual benefit between different crops within the forest could lead to differences in the yield and quality of understory medicinal plants compared to open field cultivation.A reasonable combination of planting could achieve resource sharing and complementary advantages.Therefore, conducting systematic research on the effects of understory environmental factors on the yield and content of medicinal plants with different growth and development characteristics can provide theoretical guidance and technical references for formulating comprehensive strategies for understory planting of medicinal plants, such as selecting suitable medicinal plant varieties, optimizing planting density, and conducting reasonable forest management, thus contributing to the sustainable development and ecological protection of CHM.
Plants, Medicinal/growth & development* ; Forests ; Soil/chemistry* ; Environment ; Ecosystem ; Temperature

Objective:To investigate the changes in serum levels of interleukin-1β (IL-1β) and hypoxia-inducible factor-2α (HIF-2α) in patients with intracranial aneurysms before and after surgery, and their relationship with clinical prognosis.Methods:A prospective research method was used, a total of 120 patients with intracranial aneurysms who underwent endovascular embolization treatment in Xi'an NO.9 Hospital from December 2019 to December 2022 were selected as the study subjects. According to the prognosis after one-year follow-up, they were divided into poor prognosis group and good prognosis group. The general data of the two groups and the changes of serum IL-1β and HIF-2α levels before and after surgery were compared and analyzed. The changes of serum IL-1β and HIF-2α before and after surgery in patients with intracranial aneurysms and their relationship with the prognosis of patients were analyzed.Results:Among the 120 patients, 98 (81.67%) had a good prognosis and 22 (18.33%) had a poor prognosis. The proportion of large and wide necked aneurysms in the poor prognosis group was higher than that in the good prognosis group: 40.91% (9/22) vs. 10.20% (10/98), 45.45% (10/22) vs. 20.41%(20/98), with statistical significant differences ( P<0.05). On postoperative day 7, the serum levels of IL-1 β and HIF-2 α in the poor prognosis group were higher than those in the good prognosis group: (62.58 ± 6.12) ng/L vs. (56.95 ± 5.33) ng/L, (101.62 ± 10.55) ng/L vs. (92.70 ± 7.82) ng/L, with statistical significant differences ( P<0.05). Multivariate Logistic regression analysis showed that the size of aneurysms, including giant aneurysms, wide necked aneurysms, and high levels of serum IL-1β and HIF-2α 7 d after surgery, were independent risk factors for clinical prognosis in patients with intracranial aneurysms ( P<0.05). The working characteristic curve of the subjects was drawn, and the results showed that serum IL-1β and HIF-2α had certain predictive value for the clinical prognosis of intracranial aneurysm patients 7 d after surgery, but their sensitivity was relatively low. The parallel experiment method was used to jointly predict the samples. The results showed that the area under the curve for predicting the clinical prognosis of intracranial aneurysm patients with serum IL-1β and HIF-2α at 7 d after surgery was 0.867, with sensitivity and specificity of 86.40% and 74.50%, respectively, indicating high predictive value. Conclusions:The levels of serum IL-1β and HIF-2α are significantly elevated in patients with intracranial aneurysms after surgery, and are closely related to their clinical prognosis, which can help predict the clinical prognosis of intracranial aneurysm patients.

Objective To analyze the predictive value of microRNA-34a-5p(miR-34a-5p)expression in pancreatic cancer tissue on postoperative poor prognosis.Methods The surgically resected pancreatic cancer tissues and normal tissues adjacent to cancer from 123 patients with pancreatic cancer were collected to detect the expression of miR-34a-5p.The expression of miR-34a-5p in pancreatic cancer tissues for patients with different clinicopathological characteristics were compared.The patients were divided into the poor prognosis group and the good prognosis group according to their prognosis,and the clinical data of patients between the two groups was compared.The risk factors of poor prognosis for patients with pancreatic cancer were analyzed by Cox regression model,and the predictive value of miR-34a-5p expression in pancreatic cancer tissues on poor prognosis of patients was analyzed by receiver operating characteristic(ROC)curve.Results The expression of miR-34a-5p in pancreatic cancer tissues was lower than that in normal tissues adjacent to cancer(P<0.05).There were statistically significant differences in the expression of miR-34a-5p in pancreatic cancer tissues of patients with different differentiation degrees,TNM stages,and lymph node metastasis(P<0.05).The proportions of low differentiation,TNM stage for stage Ⅲ,lymph node metastasis and incisal margin of R1,and levels of carbohydrate antigen 199(CA199),neutrophils/lymphocytes ratio(NLR)and platelet/lymphocyte ratio(PLR)for patients in the poor prognosis group were higher than those in the good prognosis group(P<0.05),while miR-34a-5p expression was lower than that in the good prognosis group(P<0.05).Cox regression analysis showed that low differentiation,TNM stage for stage Ⅲ,lymph node metastasis,incisal margin of R1,decreased expression of miR-34a-5p and increased levels of CA199,NLR and PLR were risk factors of poor prognosis for patients with pancreatic cancer(P<0.05).ROC curve analysis showed that the optimal cut-off value of miR-34a-5p expression in pancreatic cancer tissue for predicting poor prognosis of patients was 0.48,the sensitivity was 78.82%,the specificity was 89.47%and the area under the curve was 0.855,with good predictive value.Conclusion The expression of miR-34a-5p in pancreatic cancer tissue is lower than that in normal tissue adjacent to cancer,and its expression is related to the differentiation degree,TNM stage and lymph node metastasis,which is also a risk factor and predictor of poor prognosis for patients with pancreatic cancer.