Objective:To analyze the colonization rate and molecular types of Staphylococcus aureus (SA) on the tracheotomy wounds of neonates at early postoperative stage in neonatal intensive care unit (NICU). Methods:This was a case series study. Patients who were admitted and underwent tracheotomy in NICU of Beijing Children′s Hospital, Capital Medical University from January 1 st 2020 to December 31 st 2023 were enrolled. Swabs on the skin around the incision or on the nasal mucosa were collected and cultured at 24, 72 and 168 h after operation. Coagulase test and Staphytect Plus kits were used for SA identification. The nuc gene amplification and molecular types of SA were assessed by PCR. The patients were divided into SA colonization group and non-colonization group based on the presence or absence of SA colonization, and into infection group and non-infection group based on the presence or absence of infection. Demographic data, hospitalization information, colonization and infection status of SA were collected from the digital medical record system of the hospital. Differences between groups were analyzed using the independent sample t test or Fisher exact test. Results:Totally 19 patients were enrolled, among whom 13 were male. The gestation age was 39.0 (38.1, 40.0) weeks, and the birth weight was 3 150 (2 600, 3 400) g. Tracheotomy was done at 8.2(4.1, 19.6) days after diagnosis and indication confirmed. Corrected gestational age of patients on the operation day was 43.6 (42.2, 45.4) weeks. The NICU stay time was (34.0±3.1) days. SA colonization was confirmed around the incision of 8 patients. Out of the 18 strains of colonized bacteria, 10 were methicillin-resistant Staphylococcus aureus (MRSA). The most common molecular type of MRSA was ST59-SCCmec Ⅳ-t437 strain (8 strains). A total of 10 patients presented typical clinical manifestations of bacterial infection at the lungs, 3 patients in the blood stream and 2 patients in the central nervous system. Among 10 patients with bacterial infection, 3 patients were MRSA positive by boby fluid culture and affected by the ST59-SCCmec Ⅳ-t437 strain. The infection rate was different between patients with or without SA colonization on the tracheotomy incision (7/8 vs. 3/11, P=0.020). Conclusions:The colonization rate and infection rate were high on the tracheotomy incision in neonates. The major type was MRSA, and the most common molecular strain was ST59-SCCmec Ⅳ-t437 .

Fetal ultrasound standard plane detection was introduced in terms of its importance and problems encountered.The deep learning techniques applied in fetal standard plane detection were reviewed,including transfer learning,modified basic network,hybrid network and multi-task network.The problems encountered by the deep learning techniques during the application were analyzed,and the future research directions were envisioned.[Chinese Medical Equipment Journal,2025,46(5):91-101]

This study was aimed at investigating infections with Giardia and Cryptosporidium in Ochotona curzoniae in Zoige County,Sichuan Province.O.curzoniae were captured in five townships of Zoige County(Dazhasi,Axi,Hongxing,Tangke,and Maixi)between March and December of 2023.DNA from the gastrointestinal contents was subjected to nested PCR to amplify Giardia bg,gdh,and tpi genes,and the Cryptosporidium SSU rRNA gene.The sequences of PCR-PCR products were analyzed and compared.Phylogenetic trees were constructed to determine the protozoa species and genotypes.A total of 114 O.curzoniae animals were captured,among which 44 samples showed bg gene positivity,and 14 samples showed gdh gene positivity for Giardia.The total detection rate was 43.9％(50/114),and two assemblages were detected(assem-blage E and a new assemblage tentatively termed assemblage OC1);the positivity rate for Cryptosporidium was 7.0％(8/114),and three new genotypes were observed.Mixed infection with Cryptosporidium and Giardia was present in some sam-ples,with a detection rate of 3.5％(4/114).Giardia lamblia and Giardia sp.(REG-1,REG-2)were prevalent in O.curzoni-ae in Zoige County in Sichuan province;assemblage E was the dominant assemblage,and the new assemblage OC1 was pres-ent;and Cryptosporidium sp.(REG-1,REG-2,and REG-3)were identified.In summary,future monitoring of Giardia and Cryptosporidium should be further strengthened in Zoige to provide detailed data for promoting local public health.

Objective:To develop PixelSIFT-UNet, a novel semantic segmentation model that integrates deep learning with scale-invariant feature transform (SIFT) algorithm to improve the segmentation accuracy of oral mucosal lesions.Methods:This investigation utilized 838 standard clinical white light images of oral mucosal diseases acquired from January 2020 to December 2022 at the Stomatology Hospital Zhejiang University School of Medicine. Randomization was achieved through Python′s random.seed function implementation. The random sample function was subsequently applied for sampling distribution. The dataset was stratified into three subsets with a 6∶2∶2 ratio: training ( n=506), validation ( n=166), and testing ( n=166). Lesion boundaries were annotated using Labelme software, and a PixelSIFT-UNet-based deep learning model was developed with VGG-16 and ResNet-50 backbone networks. Model parameters were optimized using the validation set, and performance metrics [including Dice coefficient, mean intersection over union (mIoU), mean pixel accuracy (mPA), and Precision] were assessed on the test set. The model′s performance was benchmarked against conventional semantic segmentation frameworks (U-Net and PSPNet). Results:The developed PixelSIFT-UNet model could achieve precise segmentation of three common oral mucosal lesions: oral lichen planus, oral leukoplakia, and oral submucous fibrosis. Utilizing VGG-16 as the backbone network, the model achieved Dice coefficient, mIoU, mPA, and Precision values of 0.642, 0.699, 0.836, and 0.792, respectively. Implementation with ResNet-50 backbone network yielded metrics of 0.668, 0.733, 0.872 and 0.817, demonstrating significant improvements across all performance indicators compared to conventional U-Net model (relevant metrics: 0.662, 0.717, 0.861 and 0.809) and PSPNet model (relevant metrics: 0.671, 0.721, 0.858 and 0.813).Conclusions:The proposed PixelSIFT-UNet architecture demonstrates superior performance in oral mucosal lesion segmentation tasks, surpassing conventional semantic segmentation models and providing robust quantitative improvements in segmentation accuracy.

Objective:To analyze the colonization rate and molecular types of Staphylococcus aureus (SA) on the tracheotomy wounds of neonates at early postoperative stage in neonatal intensive care unit (NICU). Methods:This was a case series study. Patients who were admitted and underwent tracheotomy in NICU of Beijing Children′s Hospital, Capital Medical University from January 1 st 2020 to December 31 st 2023 were enrolled. Swabs on the skin around the incision or on the nasal mucosa were collected and cultured at 24, 72 and 168 h after operation. Coagulase test and Staphytect Plus kits were used for SA identification. The nuc gene amplification and molecular types of SA were assessed by PCR. The patients were divided into SA colonization group and non-colonization group based on the presence or absence of SA colonization, and into infection group and non-infection group based on the presence or absence of infection. Demographic data, hospitalization information, colonization and infection status of SA were collected from the digital medical record system of the hospital. Differences between groups were analyzed using the independent sample t test or Fisher exact test. Results:Totally 19 patients were enrolled, among whom 13 were male. The gestation age was 39.0 (38.1, 40.0) weeks, and the birth weight was 3 150 (2 600, 3 400) g. Tracheotomy was done at 8.2 (4.1, 19.6) days after diagnosis and indication confirmed. Corrected gestational age of patients on the operation day was 43.6 (42.2, 45.4) weeks. The NICU stay time was (34.0±3.1) days. SA colonization was confirmed around the incision of 8 patients. Out of the 18 strains of colonized bacteria, 10 were methicillin-resistant Staphylococcus aureus (MRSA). The most common molecular type of MRSA was ST59-SCCmec Ⅳ-t437 strain (8 strains). A total of 10 patients presented typical clinical manifestations of bacterial infection at the lungs, 3 patients in the blood stream and 2 patients in the central nervous system. Among 10 patients with bacterial infection, 3 patients were MRSA positive by boby fluid culture and affected by the ST59-SCCmec Ⅳ-t437 strain. The infection rate was different between patients with or without SA colonization on the tracheotomy incision (7/8 vs. 3/11, P=0.020). Conclusions:The colonization rate and infection rate are high on the tracheotomy incision in neonates. The major type is MRSA, and the most common molecular strain is ST59-SCCmec Ⅳ-t437 .

Objective:To develop PixelSIFT-UNet, a novel semantic segmentation model that integrates deep learning with scale-invariant feature transform (SIFT) algorithm to improve the segmentation accuracy of oral mucosal lesions.Methods:This investigation utilized 838 standard clinical white light images of oral mucosal diseases acquired from January 2020 to December 2022 at the Stomatology Hospital Zhejiang University School of Medicine. Randomization was achieved through Python′s random.seed function implementation. The random sample function was subsequently applied for sampling distribution. The dataset was stratified into three subsets with a 6∶2∶2 ratio: training ( n=506), validation ( n=166), and testing ( n=166). Lesion boundaries were annotated using Labelme software, and a PixelSIFT-UNet-based deep learning model was developed with VGG-16 and ResNet-50 backbone networks. Model parameters were optimized using the validation set, and performance metrics [including Dice coefficient, mean intersection over union (mIoU), mean pixel accuracy (mPA), and Precision] were assessed on the test set. The model′s performance was benchmarked against conventional semantic segmentation frameworks (U-Net and PSPNet). Results:The developed PixelSIFT-UNet model could achieve precise segmentation of three common oral mucosal lesions: oral lichen planus, oral leukoplakia, and oral submucous fibrosis. Utilizing VGG-16 as the backbone network, the model achieved Dice coefficient, mIoU, mPA, and Precision values of 0.642, 0.699, 0.836, and 0.792, respectively. Implementation with ResNet-50 backbone network yielded metrics of 0.668, 0.733, 0.872 and 0.817, demonstrating significant improvements across all performance indicators compared to conventional U-Net model (relevant metrics: 0.662, 0.717, 0.861 and 0.809) and PSPNet model (relevant metrics: 0.671, 0.721, 0.858 and 0.813).Conclusions:The proposed PixelSIFT-UNet architecture demonstrates superior performance in oral mucosal lesion segmentation tasks, surpassing conventional semantic segmentation models and providing robust quantitative improvements in segmentation accuracy.

Objective To explore the predictive value of dose surface histogram(DSH)in image guided radiotherapy(IGRT)for radiotherapy-induced acute radiation proctitis(ARP)in cervical cancer(CCA).Methods Totally 380 patients with CCA IGRT admitted to some hospital from May 2019 to May 2023 were selected prospectively and randomly divided into a control group(n=1 80)and an experimental group(n=200).The patients in the 2 groups were followed up and the incidence rates of ARP were counted,and rectal dose distribution was evaluated using dose volume histogram(DVH)in the control group and DSH in the experimental group.The predictive values of DVH and DSH for ARP were evaluated and compared using ROC curves.Statistical analysis was performed using SPSS 21.0 software.Results The two groups did not have statistically significant difference in the incidence rate of ARP(P＞0.05),while there were significant differences in the evaluation indicators of the rectal dose distribution(P＜0.05).V40,V50,S40 and S50 proved to have low predictive values for grade Ⅰ-Ⅳ ARP with AUC 0.700(P＜0.05);V60 and S60 had moderate predictive values for grade Ⅰ-Ⅳ ARP with AUC greater than 0.700 and less than or equal to 0.900(P＜0.05);V70,V78,S70 and S7s showed high predictive values for grade Ⅰ-Ⅳ ARP with AUC higher than 0.900(P＜0.05).Delong's test results indicated that DVH and DSH had no significant differences in AUC when used to predict gradeⅠ-Ⅳ ARP(allP＞0.05).Conclusion DSH is essentially the same as DVH when used for the prediction of grade Ⅰ-Ⅳ ARP due to CCA IGRT,and thus can be used for the supplementation and optimization of radiotherapy planning systems.[Chinese Medical Equipment Journal,2025,46(3):48-53]

Objective To explore the predictive value of dose surface histogram(DSH)in image guided radiotherapy(IGRT)for radiotherapy-induced acute radiation proctitis(ARP)in cervical cancer(CCA).Methods Totally 380 patients with CCA IGRT admitted to some hospital from May 2019 to May 2023 were selected prospectively and randomly divided into a control group(n=1 80)and an experimental group(n=200).The patients in the 2 groups were followed up and the incidence rates of ARP were counted,and rectal dose distribution was evaluated using dose volume histogram(DVH)in the control group and DSH in the experimental group.The predictive values of DVH and DSH for ARP were evaluated and compared using ROC curves.Statistical analysis was performed using SPSS 21.0 software.Results The two groups did not have statistically significant difference in the incidence rate of ARP(P＞0.05),while there were significant differences in the evaluation indicators of the rectal dose distribution(P＜0.05).V40,V50,S40 and S50 proved to have low predictive values for grade Ⅰ-Ⅳ ARP with AUC 0.700(P＜0.05);V60 and S60 had moderate predictive values for grade Ⅰ-Ⅳ ARP with AUC greater than 0.700 and less than or equal to 0.900(P＜0.05);V70,V78,S70 and S7s showed high predictive values for grade Ⅰ-Ⅳ ARP with AUC higher than 0.900(P＜0.05).Delong's test results indicated that DVH and DSH had no significant differences in AUC when used to predict gradeⅠ-Ⅳ ARP(allP＞0.05).Conclusion DSH is essentially the same as DVH when used for the prediction of grade Ⅰ-Ⅳ ARP due to CCA IGRT,and thus can be used for the supplementation and optimization of radiotherapy planning systems.[Chinese Medical Equipment Journal,2025,46(3):48-53]