Objective:To develop a deep learning-based segmentation model for anterior segment ultrasound biomicroscopy (UBM) images to automatically segment the anterior segment tissues of patients with primary angle-closure glaucoma (PACG).Methods:A single-center retrospective case series was conducted.A small-scale dataset comprised 468 UBM images of the anterior chamber angle closure from 156 patients with PACG who underwent the UBM examination at Tianjin Medical University Eye Hospital between July 12, 2022, and February 20, 2023.The UBM images were randomly split into a training dataset of 228 images and a testing dataset of 152 images using a random seed method in a ratio of 6∶4.The models were trained using the PSPNet model with MobileNet V2 and ResNet50 as backbones, the DeepLab v3+ model with MobileNet V2 and Xception as backbones, and the SegFormer model with MiT-B0 and MiT-B2 as backbones.The testing dataset was used for result prediction and to achieve segmentation of four regions: the cornea and sclera, iris, ciliary body, and anterior lens surface.To evaluate the performance of the models in segmenting the anterior segment structures, multiple metrics were assessed, including the mean intersection over union (mIoU), Dice coefficient, precision, recall, false negative rate, and specificity.A comparative analysis of the test results across the different models was subsequently performed.This study adhered to the Declaration of Helsinki.The study protocol was approved by the Ethics Committee of Tianjin Medical University Eye Hospital (No.2023KY-05).Results:The two models with the best segmentation performance were PSPNet and DeepLab v3+ .The PSPNet model with ResNet50 as the backbone achieved the mIoU of 85.11%, Dice coefficient of 91.38%, precision of 91.83%, recall of 90.94%, false negative rate of 9.06%, and specificity of 98.89%.The DeepLab v3+ model with MobileNet V2 as the backbone achieved an mIoU of 85.84%, Dice coefficient of 92.01%, precision of 92.67%, recall of 91.36%, false negative rate of 8.64%, and specificity of 98.90%.Among the five key metrics, mIoU, Dice coefficient, recall, false negative rate, and specificity, DeepLab v3+ exhibited the best segmentation performance.In addition, the DeepLab v3+ model with Xception as the backbone had the highest precision among all models, reaching 92.77%.Conclusions:The deep learning-based DeepLab v3+ model achieves precise segmentation of anterior segment tissue structures in PACG anterior segment UBM image segmentation, providing auxiliary support for clinical diagnosis.

Four cases of typical faults of Siemens MOBILETT XP Digital bedside DR were introduced.The causes for the problems were analyzed,and some countermeasures were put forward accordingly.References were provided for medical engineers to treat similar faults.[Chinese Medical Equipment Journal,2025,46(3):118-120]

OBJECTIVE: To investigate the clinical characteristics and prognosis of patients with IgD multiple myeloma (MM). METHODS: The clinical data of 8 patients with IgD MM admitted to Gansu Provincial Hospital from September 2013 to February 2023 were collected, and their clinical characteristics and prognosis were retrospectively analyzed and summarized. RESULTS: Among the 8 enrolled patients, there were 4 males and 4 females, with a median age of 60 (44-74) years. All patients had symptoms of renal insufficiency and anemia. There were 3 cases of bone invasion, 3 cases of splenomegaly, 7 cases of IgD-λ type, and 1 case of IgD-κ type. FISH examination was performed in 7 cases, and 6 of them were positive for 1q21 . There were 6 cases in DS stage III and 2 cases in DS stage II; According to ISS staging, there were 6 cases in stage III, 1 case in stage II, and 1 case in stage I; According to R-ISS staging, there were 5 cases in stage III and 3 cases in stage II. All patients received bortezomib-based combination chemotherapy, with 1 case undergoing autologous stem cell transplantation (ASCT) and 2 cases receiving daratumumab in combination. The median treatment period was 6 (1-15) cycles. The short-term efficacy was evaluated after 4-6 courses of treatment. Among the 6 patients with assessable efficacy, 1 case experienced disease progression (PD), and 5 cases achieved complete remission (CR). The median follow-up time was 26 (11-33) months, and the median progression-free survival (PFS) and median overall survival (OS) of the patients were 11.25 (3-26) months and 18.5 (4-33) months, respectively. Among the 8 patients, 4 cases died. Among the deceased patients, 3 cases were in R-ISS stage III and 3 cases were 1q21 positive. 2 of the 5 patients with early CR died due to disease progression. CONCLUSION The incidence of IgD MM is low, the symptoms of early renal damage, blood system damage and bone erosion in IgD MM patients are obvious, and the median survival time is short. ASCT and / or daratumumab may bring lasting relief for IgD MM patients, but large-scale clinical studies are still needed.
Humans ; Multiple Myeloma/therapy* ; Middle Aged ; Male ; Female ; Aged ; Prognosis ; Immunoglobulin D ; Adult ; Retrospective Studies

OBJECTIVE: To analyze the clinical characteristics and prognosis of patients with T-cell large granular lymphocytic leukemia (T-LGLL). METHODS: The clinical data of 7 patients with T-LGLL in Gansu Provincial Hospital from March 2016 to June 2023 were analyzed retrospectively. RESULTS: Among the 7 patients, 5 were male and 2 were female, with a median age of 51(28-83) years old. At the onset of illness, 6 cases showed symptoms of fatigue and anemia, 4 cases had enlarged lymph nodes, and 5 cases had splenomegaly. Examination showed that 4 cases were antinuclear antibody(ANA) positive, 5 cases were anemia. The median hemoglobin (Hb) level was 83(61-151) g/L, the median white blood cell count (WBC) was 5.6(2.0-8.7)×10⁹ /L, and the median percentage of lymphocytes in peripheral blood was 66.2(13.9-89.1)%. There were 3 cases with extremely active bone marrow hyperplasia, 2 cases with active hyperplasia, and 2 cases with decreased hyperplasia. There were 5 cases with mild myelofibrosis (MF-1), and 1 case with moderate myelofibrosis (MF-2). The median percentage of T cells was 64.3 (31.5-80.6)%. 5 cases showed the classic immunophenotype (CD3 ⁺ CD4^- CD8 ⁺), 6 cases were CD57 ⁺, 3 cases were TCRα/β ⁺, and 3 cases were TCRγ/δ ⁺. TCRG rearrangement was detected in 5 cases.The median follow-up time was 55(4-87) months, one patient died of heart disease, and the other 6 patients are surviving. CONCLUSION The incidence of T-LGLL is low. The initial symptoms of T-LGLL include anemia, fatigue, lymph node enlargement, splenomegaly, and higher percentage of lymphocytes in peripheral blood, the percentage of abnormal T cells in bone marrow was significantly increased. Analysis of flow cytometric immunophenotyping, TCR gene rearrangement, and hot spot genes such as STAT3 and STAT5b, can improve the diagnostic accuracy.
Humans ; Leukemia, Large Granular Lymphocytic/diagnosis* ; Male ; Middle Aged ; Female ; Aged ; Prognosis ; Adult ; Aged, 80 and over ; Retrospective Studies

Objective:To develop a deep learning-based segmentation model for anterior segment ultrasound biomicroscopy (UBM) images to automatically segment the anterior segment tissues of patients with primary angle-closure glaucoma (PACG).Methods:A single-center retrospective case series was conducted.A small-scale dataset comprised 468 UBM images of the anterior chamber angle closure from 156 patients with PACG who underwent the UBM examination at Tianjin Medical University Eye Hospital between July 12, 2022, and February 20, 2023.The UBM images were randomly split into a training dataset of 228 images and a testing dataset of 152 images using a random seed method in a ratio of 6∶4.The models were trained using the PSPNet model with MobileNet V2 and ResNet50 as backbones, the DeepLab v3+ model with MobileNet V2 and Xception as backbones, and the SegFormer model with MiT-B0 and MiT-B2 as backbones.The testing dataset was used for result prediction and to achieve segmentation of four regions: the cornea and sclera, iris, ciliary body, and anterior lens surface.To evaluate the performance of the models in segmenting the anterior segment structures, multiple metrics were assessed, including the mean intersection over union (mIoU), Dice coefficient, precision, recall, false negative rate, and specificity.A comparative analysis of the test results across the different models was subsequently performed.This study adhered to the Declaration of Helsinki.The study protocol was approved by the Ethics Committee of Tianjin Medical University Eye Hospital (No.2023KY-05).Results:The two models with the best segmentation performance were PSPNet and DeepLab v3+ .The PSPNet model with ResNet50 as the backbone achieved the mIoU of 85.11%, Dice coefficient of 91.38%, precision of 91.83%, recall of 90.94%, false negative rate of 9.06%, and specificity of 98.89%.The DeepLab v3+ model with MobileNet V2 as the backbone achieved an mIoU of 85.84%, Dice coefficient of 92.01%, precision of 92.67%, recall of 91.36%, false negative rate of 8.64%, and specificity of 98.90%.Among the five key metrics, mIoU, Dice coefficient, recall, false negative rate, and specificity, DeepLab v3+ exhibited the best segmentation performance.In addition, the DeepLab v3+ model with Xception as the backbone had the highest precision among all models, reaching 92.77%.Conclusions:The deep learning-based DeepLab v3+ model achieves precise segmentation of anterior segment tissue structures in PACG anterior segment UBM image segmentation, providing auxiliary support for clinical diagnosis.

Four cases of typical faults of Siemens MOBILETT XP Digital bedside DR were introduced.The causes for the problems were analyzed,and some countermeasures were put forward accordingly.References were provided for medical engineers to treat similar faults.[Chinese Medical Equipment Journal,2025,46(3):118-120]

Objective To investigate the safety and efficacy of novel bioabsorbable vascular scaffold(BVS)in the treatment of patients with complex coronary artery disease.Methods This was a retrospective,matched,single-center observational study.45 patients with coronary atherosclerotic cardiopathy received BVS treatment in the cardiovascular medicine department Department of the Affiliated Hospital of Guangdong Medical University from June 2020 to June 2021(BVS),and 45 patients treated with drug-eluting stents(DES)group were selected according to matching study requirements during the same period.Baseline,surgical,and follow-up data were compared between the two groups to evaluate safety and efficacy.The main measures of safety were:surgical time,intraoperative adverse events,etc.,and the end point of efficacy was target lesion failure(TLF),including cardiac death,target vessel myocardial infarction,and ischa-driven target lesion revascularization.Results A total of 90 patients were enrolled in this study,all of whom were followed up for at least 3 years.There were 20 cases of bifurcation lesions and 25 cases of diffuse long lesions in the two groups,and 50 cases of imaging were reviewed among the 90 patients.The proportion of stable coronary heart disease,history of diabetes,history of hypertension,history of smoking,pre-dilated balloon pressure and postoperative diastolic blood pressure in BVS group was higher than that in DES group,and the proportion of family history was lower than that in DES group(all P＜0.05).There were no statistically significant differences in the rates of cardiac death,target vessel myocardial infarction,and ischemia-driven revascularization of target lesions between the two groups(all P＞0.05).Binary Logistic regression model analysis showed that the diameter stenosis ratio of target lesions was an independent risk factor for intrastent restenosis(OR 2.786,95％CI 1.096-7.081,P=0.031).Conclusions Compared with traditional DES,BVS implantation has consistent safety and efficacy in the treatment of complex coronary artery disease within 3 years.The diameter stenosis ratio of target lesions was an independent risk factor for intrastent restenosis.

Objective To analyze the molecular markers of various nuclei in the human basal ganglia and the differentially expressed genes(DEGs)among different nuclei,gender,and Parkinson's disease(PD),followed by the biological function annotations of the DEGs.Methods Forty-five specimens of basal ganglia from 10 human postmortem brains were divided into control and PD groups,and the control group was further categorized into female and male groups.RNA from each sample was extracted for high-throughput transcriptome sequencing.Bioinformatic analysis was conducted to identify molecular markers of each nuclei in the control group,nuclei-specific,gender-specific,and PD-specific DEGs,followed by gene enrichment analysis and functional annotation.Results Sequencing analysis revealed top DEGs such as DRD1,FOXG1,and FAM183A in the caudate;SLC6A3,EN1,SLC18A2,and TH in the substantia nigra;MEPE and FGF10 in the globus pallidus;and SLC17A6,PMCH,and SHOX2 in the subthalamic nucleus.In them,putamen showed some overlapping DEGs with caudate,such as DRD1 and FOXG1.A significant number of DEGs were identified among different nuclei in the control group,with the highest number between caudate and globus pallidus(9321),followed by putamen and globus pallidus(6341),caudate and substantia nigra(6054),and substantia nigra and subthalamic nucleus(44).Gene enrichment analysis showed that downregulated DEGs between caudate and globus pallidus were significantly enriched in processes like myelination of neurons and cell migration.Upregulated DEGs between putamen and globus pallidus were enriched processes like chemical synaptic transmission and regulation of membrane potential,while downregulated DEGs were enriched in myelination and cell adhesion.Upregulated DEGs between caudate and substantia nigra were enriched in processes like chemical synaptic transmission and axonal conduction,while downregulated DEGs were enriched in myelination of neurons.Totally 468,548,1402,333,and 341 gender-specific upregulated DEGs and 756,988,2532,444,and 1372 downregulated DEGs were identified in caudate,putamen,substantia nigra,globus pallidus,and subthalamus nucleus.Gene enrichment analysis revealed upregulated DEGs mostly enriched in pathways related to immune response and downregulated DEGs in chemical synaptic transmission.At last,709,852,276,507,and 416 PD-specific upregulated DEGs and 830,2014,1218,836,and 1730 downregulated DEGs were identified in caudate,putamen,substantia nigra,globus pallidus,and subthalamus nucleus.Gene enrichment analysis revealed upregulated DEGs mostly enriched in apoptotic regulation and downregulated DEGs in chemical synaptic transmission and action potential regulation.Conclusion We identified and analysed the molecular markers of different human basal ganglia nuclei,as well as DEGs among different nuclei,different gender,and between control and PD.

This study aims to provide evidence for clinical practice by systematically reviewing the efficacy and safety of Gusongbao preparation in the treatment of primary osteoporosis(POP). The relevant papers were retrieved from four Chinese academic journal databases and four English academic journal databases(from inception to May 31, 2022). The randomized controlled trial(RCT) of Gusongbao preparation in the treatment of POP was included after screening according to the inclusion and exclusion criteria. The quality of articles was evaluated using risk assessment tools, and the extracted data were subjected to Meta-analysis in RevMan 5.3. A total of 657 articles were retrieved, in which 15 articles were included in this study, which involved 16 RCTs. A total of 3 292 patients(1 071 in the observation group and 2 221 in the control group) were included in this study. In the treatment of POP, Gusongbao preparation+conventional treatment was superior to conventional treatment alone in terms of increasing lumbar spine(L2-L4) bone mineral density(MD=0.03, 95%CI[0.02, 0.04], P<0.000 01) and femoral neck bone mineral density, reducing low back pain(MD=-1.69, 95%CI[-2.46,-0.92], P<0.000 1) and improving clinical efficacy(RR=1.36, 95%CI[1.21, 1.53], P<0.000 01). Gusongbao preparation was comparable to similar Chinese patent medicines in terms of improving clinical efficacy(RR=0.95, 95%CI[0.86, 1.04], P=0.23). Gusongbao preparation was inferior to similar Chinese patent medicines in reducing traditional Chinese medicine syndrome scores(MD=1.08, 95%CI[0.44, 1.71], P=0.000 9) and improving Chinese medicine syndrome efficacy(RR=0.89, 95%CI[0.83, 0.95], P=0.000 4). The incidence of adverse reactions of Gusongbao preparation alone or combined with conventio-nal treatment was comparable to that of similar Chinese patent medicines(RR=0.98, 95%CI[0.57, 1.69], P=0.94) or conventio-nal treatment(RR=0.73, 95%CI[0.38, 1.42], P=0.35), and the adverse reactions were mainly gastrointestinal discomforts. According to the available data, Gusongbao preparation combined with conventional treatment is more effective than conventional treatment alone in increasing lumbar spine(L2-L4) bone mineral density and femoral neck bone mineral density, reducing low back pain, and improving clinical efficacy. The adverse reactions of Gusongbao preparation were mainly gastrointestinal discomforts, which were mild.
Humans ; Bone Density ; Low Back Pain ; Medicine, Chinese Traditional ; Osteoporosis/drug therapy*

In this study, alkali-soluble polysaccharide was extracted from Poria residue, and the structure of alkali-soluble polysaccharide was characterized by Fourier transform infrared spectroscopy (FTIR), X-ray powder diffraction (XRD), and differential scanning calorimetry (DSC). The physical morphology of alkali-soluble polysaccharide and ethyl cellulose (EC) was investigated by scanning electron microscopy (SEM), and the focus on angle of repose, bulk density, tapped density, Carr index, interparticle porosity, cohesion index, Hausner ratio, etc. The physical fingerprints were drawn, and the powder properties were evaluated by multivariate analysis. Diclofenac sodium extended-release tablets were prepared by direct compression method using alkali-soluble polysaccharide and EC as insoluble backbone materials to evaluate the basic properties of the extended-release tablets, investigate the in vitro drug release behavior and study the release mechanism. The results showed that alkali-soluble polysaccharide is a semi-crystalline polymer with smooth lamellar structure, and its stacking and compressibility are stronger than EC. The in vitro release experiments showed that the slow release performance of alkali-soluble polysaccharide is stronger than EC, and the release behavior of the prepared slow release tablets is in accordance with the Higuchi model. The pore structure is formed inside the tablets during the release process, and the release mode is pore diffusion release. The results of this study are of great significance for the development of new slow-release materials and the rational use of resources.