1.Automatic recognition and segmentation of brachial plexus in ultrasonic images based on deep learning
Duo SHI ; Han ZHANG ; Peipei LIU ; Ruichao ZHANG ; Qingyu LIU ; Hao SUN ; Xiaofang FU ; Mengjie DOU ; Junpu HU ; Changqin SUN ; Keyan LI ; Jianqiu HU ; Guangquan ZHOU ; Ligang CUI ; Ping ZHOU ; Faqin LYU
Chinese Journal of Ultrasonography 2025;34(9):737-744
Objective:To propose a deep learning(DL)-based ultrasound imaging auxiliary tool for automatic segmentation and recognition of the brachial plexus(BP),and to enhance the accuracy and safety of clinical procedures.Methods:It was a multicenter study that collected 773 healthy subjects from Peking University Third Hospital and its branch campuses,the Third Medical Center of the Chinese PLA General Hospital,and Shanghai Eighth People's Hospital between August 2024 and February 2025. Brachial plexus(BP)images in the interscalene groove were captured used high-frequency ultrasound by senior sonographers,a dataset comprising 1 289 standardized images were constructed and the improved model(CHA-TransUNet)was trained. The test set was input into 6 different models(CHA-TransUNet,R50-Unet,TransUnet,SegFormer,SwinUnet,MISSFormer)for segmentation. Segmentation accuracy was evaluated using metrics including the Dice similarity coefficient(DSC),95% Hausdorff distance(HD95)and mean intersection over union(mIoU),and was compared with the segmentation results of 3 ultrasound physicians with varying experience levels(junior physicians and senior physicians)to validate the model's segmentation efficacy.Results:The CHA-TransUNet model established based on a dataset of 1 289 standardized images achieved segmentation results for the BP with a DSC of 90.15%,mIoU of 91.02%,and HD95 of 8.08. Its accuracy was higher than other mainstream models(DSC:90.15% vs. 87.60%,87.77%,81.35%,84.78%,84.55%),significantly better than junior physicians(DSC:90.15% vs. 68.73%, Z=-127.76, P<0.001),and approached the level of senior physician(DSC:90.15% vs. 86.15%, Z=-31.33, P=0.549). The model demonstrated superior boundary recognition in complex anatomical structures(e.g.,C6/C7 nerve roots)compared to ultrasound physicians(junior and senior)(HD95:8.08 vs. 26.34,17.44,56.80). Conclusions:This study proposes an analysis model for BP ultrasound images,CHA-TransUNet. This model achieves segmentation and recognition of the BP with relatively complex pathways and structures. The model exhibits high accuracy and stability,outperforming current mainstream network models and junior physicians while approaching the performance level of senior physicians. It assists junior physicians or trainees in more accurately identifying and localizing the BP.
2.Automatic recognition and segmentation of brachial plexus in ultrasonic images based on deep learning
Duo SHI ; Han ZHANG ; Peipei LIU ; Ruichao ZHANG ; Qingyu LIU ; Hao SUN ; Xiaofang FU ; Mengjie DOU ; Junpu HU ; Changqin SUN ; Keyan LI ; Jianqiu HU ; Guangquan ZHOU ; Ligang CUI ; Ping ZHOU ; Faqin LYU
Chinese Journal of Ultrasonography 2025;34(9):737-744
Objective:To propose a deep learning(DL)-based ultrasound imaging auxiliary tool for automatic segmentation and recognition of the brachial plexus(BP),and to enhance the accuracy and safety of clinical procedures.Methods:It was a multicenter study that collected 773 healthy subjects from Peking University Third Hospital and its branch campuses,the Third Medical Center of the Chinese PLA General Hospital,and Shanghai Eighth People's Hospital between August 2024 and February 2025. Brachial plexus(BP)images in the interscalene groove were captured used high-frequency ultrasound by senior sonographers,a dataset comprising 1 289 standardized images were constructed and the improved model(CHA-TransUNet)was trained. The test set was input into 6 different models(CHA-TransUNet,R50-Unet,TransUnet,SegFormer,SwinUnet,MISSFormer)for segmentation. Segmentation accuracy was evaluated using metrics including the Dice similarity coefficient(DSC),95% Hausdorff distance(HD95)and mean intersection over union(mIoU),and was compared with the segmentation results of 3 ultrasound physicians with varying experience levels(junior physicians and senior physicians)to validate the model's segmentation efficacy.Results:The CHA-TransUNet model established based on a dataset of 1 289 standardized images achieved segmentation results for the BP with a DSC of 90.15%,mIoU of 91.02%,and HD95 of 8.08. Its accuracy was higher than other mainstream models(DSC:90.15% vs. 87.60%,87.77%,81.35%,84.78%,84.55%),significantly better than junior physicians(DSC:90.15% vs. 68.73%, Z=-127.76, P<0.001),and approached the level of senior physician(DSC:90.15% vs. 86.15%, Z=-31.33, P=0.549). The model demonstrated superior boundary recognition in complex anatomical structures(e.g.,C6/C7 nerve roots)compared to ultrasound physicians(junior and senior)(HD95:8.08 vs. 26.34,17.44,56.80). Conclusions:This study proposes an analysis model for BP ultrasound images,CHA-TransUNet. This model achieves segmentation and recognition of the BP with relatively complex pathways and structures. The model exhibits high accuracy and stability,outperforming current mainstream network models and junior physicians while approaching the performance level of senior physicians. It assists junior physicians or trainees in more accurately identifying and localizing the BP.
3.Application of DDPM in artificial intelligence image data augmentation of medical device
Pengfei HAO ; Qingyu LI ; Rui CHAI ; Xi CHEN ; Qinghua SONG ; Naishui HAN ; Ke ZHANG
China Medical Equipment 2024;21(3):154-158
Medical device imaging data augmentation is a method of expanding existing datasets by generating new data samples,which is of great significance for improving the performance of artificial intelligence(AI)medical device-related models and clinical application effects.However,traditional data augmentation methods are usually limited by the quality,realism,and diversity of generated samples.Denoising diffusion probabilistic model(DDPM)is a generative model based on the noise diffusion process,and its main idea is to generate samples with high quality by modelling the sampling process of the target distribution as a process of progressive denoising from the noise distribution.The basic principles and working mechanisms of DDPM were reviewed,the application scenarios of this method in AI medical device data augmentation were analyzed,and its advantages,challenges,and future development directions were explored to provide a reference for the field of AI medical device data augmentation.
4.NDFIP1 limits cellular TAZ accumulation via exosomal sorting to inhibit NSCLC proliferation.
Yirui CHENG ; Xin LU ; Fan LI ; Zhuo CHEN ; Yanshuang ZHANG ; Qing HAN ; Qingyu ZENG ; Tingyu WU ; Ziming LI ; Shun LU ; Cecilia WILLIAMS ; Weiliang XIA
Protein & Cell 2023;14(2):123-136
NDFIP1 has been previously reported as a tumor suppressor in multiple solid tumors, but the function of NDFIP1 in NSCLC and the underlying mechanism are still unknown. Besides, the WW domain containing proteins can be recognized by NDFIP1, resulted in the loading of the target proteins into exosomes. However, whether WW domain-containing transcription regulator 1 (WWTR1, also known as TAZ) can be packaged into exosomes by NDFIP1 and if so, whether the release of this oncogenic protein via exosomes has an effect on tumor development has not been investigated to any extent. Here, we first found that NDFIP1 was low expressed in NSCLC samples and cell lines, which is associated with shorter OS. Then, we confirmed the interaction between TAZ and NDFIP1, and the existence of TAZ in exosomes, which requires NDFIP1. Critically, knockout of NDFIP1 led to TAZ accumulation with no change in its mRNA level and degradation rate. And the cellular TAZ level could be altered by exosome secretion. Furthermore, NDFIP1 inhibited proliferation in vitro and in vivo, and silencing TAZ eliminated the increase of proliferation caused by NDFIP1 knockout. Moreover, TAZ was negatively correlated with NDFIP1 in subcutaneous xenograft model and clinical samples, and the serum exosomal TAZ level was lower in NSCLC patients. In summary, our data uncover a new tumor suppressor, NDFIP1 in NSCLC, and a new exosome-related regulatory mechanism of TAZ.
Humans
;
Carcinoma, Non-Small-Cell Lung/metabolism*
;
Carrier Proteins/metabolism*
;
Cell Line
;
Cell Proliferation
;
Exosomes/metabolism*
;
Lung Neoplasms/genetics*
;
Membrane Proteins/metabolism*
;
Transcriptional Coactivator with PDZ-Binding Motif Proteins/metabolism*
5.Analysis on drug resistance in people infected with HIV-1 CRF55_01B before antiviral therapy in China
Shan ZHENG ; Chang SONG ; Mengze GAN ; Qingyu HAN ; Jing HU ; Aobo DONG ; Lingjie LIAO ; Yi FENG ; Yuhua RUAN ; Hui XING
Chinese Journal of Experimental and Clinical Virology 2021;35(2):194-198
Objective:To explore the resistance to protease inhibitors (PI), nucleoside reverse transcriptase inhibitors (NRTIs), non-nucleoside reverse transcriptase inhibitors (NNRTIs) and integrase inhibitors (INSTIs) of HIV-1 CRF55_01B and the transmission of drug-resistant strains among HIV-1 CRF55_01B infected patients before antiviral treatment in China.Methods:HIV-1 RNA was extracted from plasma samples of the patients infected with CRF55_01B in the national surveillance of HIV drug resistance before antiviral treatment in 2018. A 1 056 bp gene fragment of protease/reverse transcriptase (PR/RT) region and an 846 bp gene fragment of integrase (IN) region were obtained and sequenced. Drug resistance was analyzed by using all drugs included in the Stanford University HIV db Program, HIV-1 molecular network analysis was performed with software HIV-TRACE and polymorphism mutations of CRF55_01B integrase gene region were analyzed.Results:A total of 178 samples from 26 provinces, municipalities and autonomous regions in China were analyzed, and 170 sequences of CRF55_01B PR/RT region and 170 sequences of IN region of corresponding samples were obtained. The drug resistance rate was 15.3% (26/170). The drug resistance rates of PIs, NRTIs, NNRTIs and INSTIs were 1.2% (2/170), 1.2% (2/170), 15.3% (26/170), 0.6% (1/170), respectively. The level of drug resistance was mostly low. NNRTIs drug resistance mutations were mainly V179D/E co-appeared with other mutations, and 84.1% (143/170) of the infected patients carrying V179D/E alone showed potential drug resistance. INSTIs drug resistance mutation was G163R, and showed low resistance to EVG and RAL. The molecular network access rate was 30.0%(51/170)according to the 0.9% gene distance threshold. The resistant strains were transmitted between men with homosexual transmission and heterosexually transmitted people, and both carried resistant mutations E138G and V179E. In the integrase region, CRF55_01B and CRF01_AE and B subtypes showed high mutation frequency difference in 5 sites (T215A、G134N、I135V, K136R and L101I/V).Conclusions:Before antiviral treatment, CRF55_01B infected patients in China had a high resistance to NNRTIs. Strains carrying both E138G and V179E resistance mutations were transmitting in clusters. The prevalence of CRF55_01B integrase inhibitor resistant strains is low, but some genetic polymorphisms with high mutation rate in the integrase gene region have potential influence on drug sensitivity. The influence of drug resistance of new recombinant strains on antiviral therapy in China needs to be further monitored and analyzed.
6.Study on the mechanism of Flos Puerariae and Semen Hoveniae in treatment of alcoholic liver injury based on network pharmacology and molecular docking.
Yanan WANG ; Xiaoming YAN ; Qingyu ZHANG ; Aihua SONG ; Fei HAN
Journal of Zhejiang University. Medical sciences 2020;49(6):714-724
OBJECTIVE:
To explore the mechanism of Flos Puerariae and Semen Hoveniae in treatment of alcoholic liver injury (ALI) based on network pharmacology and molecular docking.
METHODS:
The information of chemical constituents and targets of Flos Puerariae and Semen Hoveniae was collected from TCMSP and Swiss databases, and the threshold values of oral bioavailability (OB) ≥ 30%, drug likeness (DL) ≥0.18 were used to screen the potential active compounds. The GeneCard and DrugBank databases were used to obtain the targets corresponding to ALI. The common targets were queried using Venn Diagram, and the network of PPI and Gene Ontology (GO) functional enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis were performed through DAVID and Reactome database. Autodock Vina software was used for molecular docking of potential ingredients and key targets.
RESULTS:
A total of 21 potential active compounds and 431 therapeutic targets were gathered in Flos Puerariae and Semen Hoveniae, which involved 273 biological functions, 90 KEGG pathways and 362 Reactome pathways. The GO functions involved protein binding, ATP binding, etc.; the KEGG pathways mainly included PI3K-Akt signaling pathway and TNF signaling pathway; the Reactome pathways contained signal transduction and immune system, etc. The results of molecular docking showed that 21 potential active ingredients had good affinity with the core targets Akt1, TP53 and IL-6.
CONCLUSIONS
The network pharmacology and molecular docking analysis demonstrate the synergetic effect of Flos Puerariae and Semen Hoveniae with multi-compounds, multi-targets and multi-pathways in the treatment of ALI; and also predict the possible medicinal substance, key targets and pathways, which provides clues for the new drug development and mechanism research.
Animals
;
Computer Simulation
;
Drugs, Chinese Herbal/therapeutic use*
;
Lepidoptera/chemistry*
;
Liver/drug effects*
;
Liver Diseases, Alcoholic/therapy*
;
Molecular Docking Simulation
;
Phosphatidylinositol 3-Kinases/metabolism*
;
Plant Extracts/therapeutic use*
;
Rhamnaceae/chemistry*
;
Signal Transduction/drug effects*
8.The analysis of modified hamstring post surgery for the treatment of spastic cerebral palsy with knee flexion deformity
Qingyu WANG ; Zhiqiang XU ; Yan CHEN ; Fan FAN ; Bingrui ZHAO ; Lanwei YU ; Lianghua LI ; Dawei HAN
Chinese Journal of Orthopaedics 2017;37(12):713-720
Objective To explore the curative effect of knee flexion deformity on spastic cerebral palsy treatment method.Methods All of 30 patients with spastic cerebral and knee flexion deformity were randomly divided into two groups:traditional operation group and modified operation group,using the commonly used operation (In 15 cases,with traditional hamstring post surgery) and) modified operation (In 15 cases,with modified hamstring post surgery),two groups of patients were treated with Ilizarov external fixation drafting device in the correction of knee flexion deformity after soft tissue release.Adjustment began 7 days after the soft tissue release operation of external fixation,and stretched approximately 0.2 degrees each time,3 times/d,until knee flexion deformity was corrected to hyperextension for 10 degree and maintained for 3 weeks,and the flexion contracture degree of knee joint was measured every 2 weeks;then the Ilizarov external fixation drafting device was removed.Then wear a walking straight leg brace for more than 3 months,until the knee walking function is good.Clinical evaluation included the joint pain index,walking function index,knee flexion deformity degree and range of joint motion score of Dimeglio standard.Results Flexion contracture deformity of knee joint in 30 cases (60 knees) was corrected when Ilizarov external fixator was removed,knee extension to 0 degrees-5 degrees.The two groups of patients had the knee flexion angle range of-1.2 degrees to 13.3 degrees,with an average of (7.32°±3.41°) after removed of the walking straight leg brace,in which 4 cases (8 joints) got recurrent deformity of 10°-15° at the time of removing of the walking straight leg brace.Knee activity significantly was improved at the end of treatment.60 cases of knee joint activity were close to normal,with flexion of 100 degrees to 135 degrees,extension of 0 degrees to 10 degrees.Two groups of patients were statistically significantly improved before and after surgery.Curative effect comparison:The walking function index of the modified operation group was obviously superior to that of the traditional operation group,there was significant statistical difference.There were no statistically significant differences in pain index,knee flexion,and range of joint motion.Conclusion For the treatment of flexion deformity of the knee joint in spastic cerebral palsy,traditional surgery using the semitendinosus and gracilis,post and semimembranosus lysis,combined with the Ilizarov draft external fixation could improve the walking function of the patients,simplify the surgical incision and reduce trauma.As a result,modified hamstring post surgery is an ideal,effective treatment method.
9.Regulating the EMT of human gastric cancer cell line in vitro through miRNA sponge
Suli LI ; Fang ZHOU ; Qingyu ZHANG ; Wenliang JIA ; Anling ZHANG ; Lei HAN ; Chunsheng KANG
Chinese Journal of Clinical Oncology 2014;45(11):684-688
Objective:To explore the effect and mechanism of miRNA sponge on the epithelial-mesenchymal transition (EMT) of gastric carcinoma cell lines SGC7901. Methods:Synthetic ZEB2 3'UTR plasmid and siRNA targeting ZEB2 were transfected into the SGC7901 cell line by Lipofectamine 2000. Real-time quantitative polymerase chain reaction was performed to evaluate the expres-sion levels of miR-200a/b/c. Finally, the migratory, invasive, and proliferative activities of the gastric carcinoma cells in vitro were ana-lyzed by the scratch test, the Transwell cell invasion, and the cell cloning assay. The expression of the target protein was detected by Western blot. Results:Compared with the control group, the expressions of miR-200a/b/c significantly decreased, and their migration, invasion, and proliferation capabilities were considerably higher after they were transfected with ZEB2 3'UTR. Although the expres-sions of miR-200a/b/c significantly increased, the migratory, invasive, and proliferative activities of SGC7901 cells also degraded after they were transfected with siRNA targeting ZEB2. The expression of ZEB2 increased, and that of E-cadherin decreased at the protein level after they were transfected with ZEB2 3'UTR. The protein expression of Vimentin in SGC7901 cells significantly increased. The indicators show the opposite trend when cells were transfected with siZEB2, and the differences between the control and mutation groups were insignificant. Conclusion:ZEB2 3'UTR can regulate EMT course by regulating the miR-200a/b/c expression in gastric car-cinoma, consequently regulating the invasion and migration of carcinoma cells.
10.Effects of ZEB2 3′UTR Transfection on Proliferation, Invasion and Migration in Human Gastric Epithelial Cell GES-1
Suli LI ; Fang ZHOU ; Qingyu ZHANG ; Wenliang JIA ; Anling ZHANG ; Lei HAN ; Chunsheng KANG
Tianjin Medical Journal 2014;(5):401-405
Objective To explore the effects of zinc finger E-box binding protein (ZEB)2 3′UTR gene transfection on proliferation, invasion and migration in human gastric epithelial cell line GES-1. Methods The synthetic ZEB2 3′UTR and miR-200b micmics were transfected into GES-1 cell line by lipofectamine 2000. We set up control grop, the mutation group and ZEB2 3′UTR group. Real-time quantitative PCR was performed to evaluate the expression levels of miR-200a/b/c and ZEB1/ZEB2 mRNAs after transfection.And then we set up control group, ZEB2 3′UTR group, ZEB2 3′UTR+negative control group and ZEB2 3′UTR+miR-200b micmics group. The protein expression levels of ZEB1, ZEB2, matrix metallopro-teinases (MMP) 2/9 and proliferating cell nuclear antigen (PCNA) were detected by Western blot assay. The invasion and mi-gration capability were analyzed by transwell assay and wound healing test. MTT assay was used to detect the proliferation ability. Results Compared with control group and mutation group, the expressions of miR-200a/b/c were significantly de-creased, especially for miR-200b. And the expressions of ZEB1/ZEB2 were significantly increased at both mRNA and pro-tein levels after transfected with the ZEB2 3′UTR, enhancing the capability of migration,invasion,and proliferation (P <0.05). Compared with ZEB2 3′UTR group, the capabilities of proliferation,invasion and migration were significantly lower in combined group. Conclusion ZEB2 3′UTR can increase the ability of cell proliferation, invasion and metastasis through regulating the levels of miR-200a/b/c, and then influence the regulation of transcription of the target gene, which could lead to malignant transformation of GES-1 cells.

Result Analysis
Print
Save
E-mail