Objective To investigate the impact of different application methods of tissue compensators(bolus)on the skin dose delivered to the chest wall following radical mastectomy for breast cancer.Methods A retrospective analysis was conducted on 60 female patients who underwent radical mastectomy and required chest wall radiotherapy at the hospital between January 2023 and March 2025.The Pinnacle3 9.10 radiotherapy planning system(TPS)was utilized to design two VMAT dual semi-arc radiotherapy plans for each patient,with a prescribed target dose of 50 Gy delivered in 2 Gy fractions over 25 sessions.In Plan 1,a Bolus was applied and optimized during the first 15 fractions,and subsequently removed for the remaining 10 fractions without re-optimization.The sub-field configuration and dose weighting from the initial optimization were retained,and only dose recalculations were performed.The final treatment plan combined both the Bolus-included and Bolus-excluded phases.In contrast,Plan 2 involved the application and optimization of Bolus during the first 15 fractions,followed by its removal and re-optimization of the plan for the last 10 fractions.The two optimized plans were then combined for the overall treatment delivery.Data from the two plan groups were analyzed using a paired sample t-test with SPSS 29.0 software.Results There was a statistically significant difference(P＜0.05)in skin Dmean,V52.5,and V55;heart Dmean,V5,V30,and V40;affected lung Dmean,V5,and V20;PRVcord Dmean and Dmax;healthy breast Dmean,V5,and V10;affected humeral head Dmean and V30;as well as PTV Dmean,V50,V55,D2％,D98％,CI,and MU.Moreover,the dose distribution on the target layer and the DVH curves showed marked differences.However,no statistically significant difference was observed in PTV HI(P=0.125).Conclusion The combination of the two optimized plans,consisting of 15 fractions with bolus and 10 fractions without bolus,more accurately reflects the dose distri-bution within the planned target area and organs at risk,thereby providing enhanced protection for the patient's chest wall skin.

OBJECTIVE To evaluate the impact of bundle intervention measures on incidence of catheter-related in-fections in patients of neurosurgery intensive care unit(NICU).METHODS The related data regarding to incidence rates of hospital-associated infections and utilization rates of catheters were collected from the NICU patients of Renji Hospital,Shanghai Jiaotong University from 2007 to 2022.The effect of the bundle intervention measures on reducing the incidence of catheter-related infections was analyzed by means of interrupted tune series(ITS)method.RESULTS There was significant difference in the decline trend of incidence of the infections during the study(the trend x2=91.978,P＜0.001).ITS analysis showed that the utilization rate of catheters only presented the decline trend after the intervention(β2=-0.667,P=0.042),the utilization rate of ventilator tubes only de-clined after the intervention(β3=-26.375,P＜0.001).Th utilization rate of central venous catheters showed a downward trend before the intervention(β1=-2.525,P＜0.001)but did not change significantly after the inter-vention.The incidence of central line-associated bloodstream infection(CLABSI)was decreased(RR=0.268,95％CI:0.124 to 0.581),however,there was significant difference in the change of slope(RR=0.912,95％CI:0.730 to 1.140).The incidence of catheter-associated urinary tract infection(CAUTI)showed a downward trend(RR=0.796,95％CI:0.647 to 0.975),however,there was no significant difference in the change of the level(RR=1.543,95％CI:0.816 to 3.005).The incidence of ventilator-associated pneumonia(VAP)showed a downward trend(RR=0.829,95％CI:0.767 to 0.895),however,there was no significant difference in the change of the level(RR=0.817,95％CI:0.595 to 1.114).CONCLUSIONS The bundle intervention measures may effectively reduce the incidence of catheter-related infections.The incidence rates of the catheter-related infections vary in the chan-ging trends after the interventions are taken,indicating that it is necessary to comprehensively take the character-istics of catheter-related infections into account when carrying out the bundle intervention measures and continu-ously optimize the control strategies.

Objective:To construct a thyroid nodule segmentation model using ultrasound dynamic images and explore its potential for assisting in the screening of thyroid nodules.Methods:A total of 126 patients with thyroid nodules（comprising 150 nodules）who were diagnosed and treated at Xuzhou Cancer Hospital from April 2024 to December 2024 were prospectively enrolled. Two-dimensional ultrasound was performed to capture short-axis and long-axis video images of thyroid nodules，forming a dynamic ultrasound image dataset. The dataset was divided into training，validation，and test sets in a ratio of 6∶1∶3. After the training loss curve converged，the model that performed well on the validation set was selected for testing. Three-fold cross-validation was employed for training and testing. All 300 ultrasound videos were divided into three subsets. In each experiment，two subsets were used as the training set，and one subset was used as the test set to evaluate the model's generalization ability. A collaborative spatiotemporal diffusion model was established based on the dynamic trends and tissue texture details of thyroid nodules. Six widely used segmentation metrics were employed to evaluate the model's application capabilities.Results:The study included 126 patients with 150 thyroid nodules，300 dynamic ultrasound images，and video lengths of 3-4 seconds per nodule，resulting in 12 312 segmented images. The size of the thyroid nodules was（10.7 ± 10.6）mm（transverse diameter）×（8.4 ± 6.3）mm（anteroposterior diameter）. Among the nodules，62（41.3%）had clear boundaries，while 88（58.7%）had indistinct boundaries；61（40.7%）exhibited regular shapes，while 89（59.3%）were irregular；66（44.0%）had a taller-than-wide aspect ratio；and 70（46.7%）showed microcalcifications. The collaborative diffusion model based on dynamic ultrasound image segmentation achieved the following scores：a Jaccard score of（69.22 ± 0.03）%，a Dice score of（79.16 ± 0.18）%，a Precision score of（86.70 ± 0.17）%，a Recall score of（77.82 ± 0.04）%，an Sα score of（85.26 ± 0.01）%，and an Eθmn score of（90.58 ± 0.17）%. Compared to other models，this model demonstrated significant improvements across all evaluation metrics，achieving the highest values in each metric with increments of over 8% and 1%，respectively. Conclusions:The collaborative diffusion model with a dynamic controller，constructed based on dynamic ultrasound images of thyroid nodules，demonstrates excellent performance in ultrasound image segmentation. It improves the accuracy of thyroid nodule screening，thereby providing a valuable auxiliary diagnostic tool for clinical practice.

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.

OBJECTIVES: We propose a multi-feature fusion model based on manually extracted features and deep learning features from endoscopic images for grading rebleeding risk of peptic ulcers. METHODS: Based on the endoscopic appearance of peptic ulcers, color features were extracted to distinguish active bleeding (Forrest I) from non-bleeding ulcers (Forrest II and III). The edge and texture features were used to describe the morphology and appearance of the ulcers in different grades. By integrating deep features extracted from a deep learning network with manually extracted visual features, a multi-feature representation of endoscopic images was created to predict the risk of rebleeding of peptic ulcers. RESULTS: In a dataset consisting of 3573 images from 708 patients with Forrest classification, the proposed multi-feature fusion model achieved an accuracy of 74.94% in the 6-level rebleeding risk classification task, outperforming the experienced physicians who had a classification accuracy of 59.9% (P<0.05). The F1 scores of the model for identifying Forrest Ib, IIa, and III ulcers were 90.16%, 75.44%, and 77.13%, respectively, demonstrating particularly good performance of the model for Forrest Ib ulcers. Compared with the first model for peptic ulcer rebleeding classification, the proposed model had improved F1 scores by 5.8%. In the simplified 3-level risk (high-risk, low-risk, and non-endoscopic treatment) classification task, the model achieved F1 scores of 93.74%, 81.30%, and 73.59%, respectively. CONCLUSIONS The proposed multi-feature fusion model integrating deep features from CNNs with manually extracted visual features effectively improves the accuracy of rebleeding risk classification for peptic ulcers, thus providing an efficient diagnostic tool for clinical assessment of rebleeding risks of peptic ulcers.
Humans ; Deep Learning ; Peptic Ulcer ; Risk Assessment ; Peptic Ulcer Hemorrhage ; Recurrence

Objective: To explore the influencing factors for implementation of weight management services in primary medical and healthcare institutions, so as to provide references for implementing sustainable services of weight management. Methods: From May to June 2025, Pinghu City, Zhejiang Province was selected as the survey site. Personnel responsible for weight management in primary medical and healthcare institutions were selected as the survey subjects using a combined method of purposive sampling and snowball sampling. Based on the five core domains of the consolidated framework for implementation research (CFIR), a semi-structured interview outline for weight management services in primary medical and healthcare institutions was designed. Original data was collected through face-to-face semi-structured interviews. Interview data was organized and analyzed using framework analysis. Factors affecting weight management services were quantitatively analyzed by referencing CFIR's structural rating criteria. Results: A total of 21 participants completed interviews, covering positions in nutrition, endocrinology, traditional Chinese medicine, general practice, maternal health, and public health. There were 9 males and 12 females. Fifteen participants (71.43%) were aged 35 years and above, 18 (85.71%) held a bachelor's degree or higher, and 15 (71.43%) were frontline medical staff. Fifteen factors affecting weight management services were identified across five domains: innovation, outer setting, inner setting, individuals, and implementation process. Six barrier factors were identified: difficulties in policy implementation, time-consuming interventions, limited incentive measures, lack of professional skills, unclear weight-loss plans and goal setting, and imperfect follow-up and evaluation mechanisms. Three neutral factors were identified: the development and refinement of policies and regulations, the implementation of weight management training, and the optimization of the referral process within integrated healthcare systems (medical alliances / communities). Six facilitating factors were identified: the relatively significant advantages of lifestyle interventions, collaboration and coordination across multiple departments, cooperative communication among different units within the institution, the inherent convenience of primary care settings, a strong sense of professional responsibility, and the establishment of multidisciplinary teams. Conclusions The delivery of weight management services in primary medical and healthcare institutions is influenced by a wide array of factors across multiple domains. It requires policy support, multi-department coordination, a practice-oriented training system, optimized team resource allocation, incentives, and improved professional skills of medical staff to jointly promote long-term implementation.

Objective:To construct a thyroid nodule segmentation model using ultrasound dynamic images and explore its potential for assisting in the screening of thyroid nodules.Methods:A total of 126 patients with thyroid nodules（comprising 150 nodules）who were diagnosed and treated at Xuzhou Cancer Hospital from April 2024 to December 2024 were prospectively enrolled. Two-dimensional ultrasound was performed to capture short-axis and long-axis video images of thyroid nodules，forming a dynamic ultrasound image dataset. The dataset was divided into training，validation，and test sets in a ratio of 6∶1∶3. After the training loss curve converged，the model that performed well on the validation set was selected for testing. Three-fold cross-validation was employed for training and testing. All 300 ultrasound videos were divided into three subsets. In each experiment，two subsets were used as the training set，and one subset was used as the test set to evaluate the model's generalization ability. A collaborative spatiotemporal diffusion model was established based on the dynamic trends and tissue texture details of thyroid nodules. Six widely used segmentation metrics were employed to evaluate the model's application capabilities.Results:The study included 126 patients with 150 thyroid nodules，300 dynamic ultrasound images，and video lengths of 3-4 seconds per nodule，resulting in 12 312 segmented images. The size of the thyroid nodules was（10.7 ± 10.6）mm（transverse diameter）×（8.4 ± 6.3）mm（anteroposterior diameter）. Among the nodules，62（41.3%）had clear boundaries，while 88（58.7%）had indistinct boundaries；61（40.7%）exhibited regular shapes，while 89（59.3%）were irregular；66（44.0%）had a taller-than-wide aspect ratio；and 70（46.7%）showed microcalcifications. The collaborative diffusion model based on dynamic ultrasound image segmentation achieved the following scores：a Jaccard score of（69.22 ± 0.03）%，a Dice score of（79.16 ± 0.18）%，a Precision score of（86.70 ± 0.17）%，a Recall score of（77.82 ± 0.04）%，an Sα score of（85.26 ± 0.01）%，and an Eθmn score of（90.58 ± 0.17）%. Compared to other models，this model demonstrated significant improvements across all evaluation metrics，achieving the highest values in each metric with increments of over 8% and 1%，respectively. Conclusions:The collaborative diffusion model with a dynamic controller，constructed based on dynamic ultrasound images of thyroid nodules，demonstrates excellent performance in ultrasound image segmentation. It improves the accuracy of thyroid nodule screening，thereby providing a valuable auxiliary diagnostic tool for clinical practice.

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.

OBJECTIVE To investigate the changes of air aerosols in the digestive endoscope cleaning and disinfec-tion room under real working conditions and observe the purification effects of aerosols under different ventilation and disinfection modes.METHODS Under the real working conditions(with both the air conditioner and recircu-lating air disinfection machine under the working mode),the air samples were collected every 1 hour from 8:00 to 16:00 from the digestive endoscope cleaning and disinfection room of Ren Ji Hospital,Shanghai Jiao Tong U-niversity School of Medicine from May 2023 to Oct.2023;the contents of particulate matters(PM)and microor-ganisms in the air were detected.At the busiest moment of the cleaning and disinfection room,the air was respec-tively sampled from the cleaning and disinfection room with the working condition modes of air condition systems,recirculating air disinfection machine or natural ventilation before the starting of systems and after the work for 0.5 hour,1 hour and 2 hours.The contents of PMs and microorganisms were detected.RESULTS The PM0.5 con-centrations during various time periods met the Grade 9 ISO clean room standard(≤35,200 particles/L)under the real working conditions,and the content of airborne viable particles also conformed to the Class Ⅲ environmental requirements(≤500 CFU/m3).All of the three ventilation and disinfection methods showed certain effects on purification of the con-tents of PM0.5 and microorganisms after the ventilation for 2 hours,the contents of PMs and microorganisms of the air disinfection machine group decreased most remarkably,followed by the air conditioner group,the natural ventilation group the least.There were no significant differences in the PM concentration and the content of microorganisms among the three groups at the time points;there were only significant differences in the PM concentration and the content of mi-croorganisms of the air disinfection machine group after the ventilation for 2 hours,0.5 hour and 1 hour(P＜0.05).CONCLUSIONS Both the air conditioning systems and recirculating air disinfection machine under the working con-dition mode have certain effect on purification of the contents of PM and microorganisms,and the specific effect need to be further studied.The implementation of multicenter dynamic surveillance with the help of intelli-gent techniques may provide reference for the optimization of ventilation and disinfection strategies.

Objective:To explore the dynamic effects of venous sinus stenting (VSS) implantation on the optic nerve structure and visual function in patients with idiopathic intracranial hypertension (IIH).Methods:Patients with IIH combined with venous sinus stenosis and optic disc edema who were admitted to the Beijing Tiantan Hospital and received VSS treatment from January 2019 to September 2023 were prospectively included. Before the operation, the general data of the patients were collected and lumbar puncture examination was performed to measure the cerebrospinal fluid pressure. Ophthalmic multimodal imaging examinations such as binocular best corrected visual acuity (BCVA), optic disc stereoscopic photography, visual field, and optical coherence tomography (OCT) were performed respectively before the operation, 3 months and 6 months after the operation. The BCVA of both eyes, the Frisén grade (FPG) of optic disc edema, the mean defect (MD) value of the visual field, the thickness of the retinal nerve fiber layer (RNFL) around the optic disc and the thickness of the optic ganglion cell complex (GCC) were collected respectively. The eye with poor preoperative visual field MD value was defined as the study eye, and the other eye was the contralateral eye. The Friedman rank sum test was used to statistically analyze the differences of various indicators before the operation, 3 months after the operation and 6 months after the operation. Further, the Wilcoxon signed rank test was used to pairwise compare the results at the three time points.Results:A total of 72 patients with IIH who met the criteria were included, among whom 61 were female (84.7%), and the baseline optic disc edema Frisén grade was 3(2-4). The FPG, BCVA, MD values and RNFL thickness of the study eye and the contralateral eye at 3 and 6 months after the operation were significantly improved compared with those before the operation (all P<0.017). At 6 months after the operation, the FPG, BCVA, MD value, RNFL thickness of the study eye and the FPG and RNFL thickness of the contralateral eye were further improved compared with those at 3 months, and the differences were statistically significant (all P<0.017). During the treatment and follow-up period, no serious adverse reactions occurred in all patients. Conclusions:VSS can effectively improve papillary edema and visual function impairment in patients with IIH combined with venous sinus stenosis. Ophthalmic multimodal imaging examination provides an objective basis for preoperative assessment of optic nerve structure and visual function, postoperative efficacy monitoring and visual function recovery, etc. In the diagnosis and treatment process of IIH, the collaborative cooperation between the neurology department and the ophthalmology department is very important.