The accurate segmentation of breast ultrasound images is an important precondition for the lesion determination. The existing segmentation approaches embrace massive parameters, sluggish inference speed, and huge memory consumption. To tackle this problem, we propose T ²KD Attention U-Net (dual-Teacher Knowledge Distillation Attention U-Net), a lightweight semantic segmentation method combined double-path joint distillation in breast ultrasound images. Primarily, we designed two teacher models to learn the fine-grained features from each class of images according to different feature representation and semantic information of benign and malignant breast lesions. Then we leveraged the joint distillation to train a lightweight student model. Finally, we constructed a novel weight balance loss to focus on the semantic feature of small objection, solving the unbalance problem of tumor and background. Specifically, the extensive experiments conducted on Dataset BUSI and Dataset B demonstrated that the T ²KD Attention U-Net outperformed various knowledge distillation counterparts. Concretely, the accuracy, recall, precision, Dice, and mIoU of proposed method were 95.26%, 86.23%, 85.09%, 83.59%and 77.78% on Dataset BUSI, respectively. And these performance indexes were 97.95%, 92.80%, 88.33%, 88.40% and 82.42% on Dataset B, respectively. Compared with other models, the performance of this model was significantly improved. Meanwhile, compared with the teacher model, the number, size, and complexity of student model were significantly reduced (2.2×10 ⁶ vs. 106.1×10 ⁶, 8.4 MB vs. 414 MB, 16.59 GFLOPs vs. 205.98 GFLOPs, respectively). Indeedy, the proposed model guarantees the performances while greatly decreasing the amount of computation, which provides a new method for the deployment of clinical medical scenarios.
Humans ; Breast Neoplasms/diagnostic imaging* ; Female ; Ultrasonography, Mammary/methods* ; Image Processing, Computer-Assisted/methods* ; Algorithms ; Neural Networks, Computer ; Breast/diagnostic imaging*

Objective:To assess the radiation dose and clinical value of "one-stop" whole-brain CT perfusion (CTP) imaging in the evaluation of collateral circulation for patients with acute ischemic stroke (AIS), regarding the digital subtraction angiography (DSA) as the reference.Methods:This retrospective study included 32 AIS patients, for whom both CTP and DSA were obtained <24 h since onset. All CTP scans were acquired in whole-brain volume perfusion mode using a 320-row CT with the phase-specific settings of tube currents to optimize the image quality of CTA images, where multiple-phase (mp) CTA images were extracted from the CTP data in post-processing. The volume CT dose index (CTDI vol), dose length product (DLP), and effective dose were compared to those reported in previous studies. The perfusion parameters of the infarct lesions and their contralateral regions were compared using the paired t-tests. One radiologist scored the collateral circulation with only the CTP and with the CTP plus mp-CTA using a 5-point scale. Another radiologist performed the same evaluation on the DSA. The diagnostic accuracy was calculated referring to the result based on DSA. The scores were analyzed using the Pearson correlation coefficient. The agreement of scores was quantified with the Kappa test. Results:The mean CTDI vol was 184.18 mGy, which was comparable to the result of a previous study (184.19 mGy), and the mean effective dose was reduced 39% compared to that reported in the literature for combined CTP and CTA scanning (6.1 vs 10 mSv). There were statistically significant differences in cerebral blood volume (CBV), cerebral blood flow (CBF), mean transit time (MTT), transit time to peak (TTP), and time-to-maximum (Tmax) between the infarct lesions and their contralateral regions ( P<0.01). The scores between CTP and DSA were significantly correlated ( r=0.95, P<0.01), as well as the scores between CTP plus mp-CTA and DSA ( r=0.98, P<0.01). The Kappa value was 0.64 ( t=7.53, P<0.01) between CTP and DSA, while it increased to 0.88 ( t=9.99, P<0.01) for CTP plus mp-CTA. With the result of DSA as a reference, the diagnostic accuracy was 71.9% and 90.6% for CTP and CTP plus mp-CTA, respectively. Conclusions:The "one-stop" whole-brain CTP imaging with phase-specific settings of tube currents can provide reliable CTP and multiple-phase CTA images simultaneously, which could reasonably reduce the radiation dose. Combined use of multi-phase CTA and CT perfusion improves the diagnostic accuracy of collateral circulation in AIS patients.

Objective:To construct and implement a multidisciplinary pain management model during the perioperative period based on the project-achieving quality control circle, so as to improve the quality of patient pain management during the perioperative period.Methods:Using the convenient sampling, 310 surgical patients from the Department of Gastrointestinal Surgery, Hepatobiliary Surgery, Thoracic Surgery, Urology Surgery and Joint Surgery of Liaocheng People's Hospital from June to July 2020 were taken as the pre-improvement group, and the routine perioperative pain management model was implemented. Starting from August 2020, a project-achieving quality control circle was carried out, following the steps of theme selection, topic clarification, goal setting, formulation of strategies, investigation of the best strategies, implementation of strategies, and confirmation of effectiveness, to implement a multidisciplinary pain management model during the perioperative period. A total of 310 surgical patients admitted to 5 departments from February to March 2021 were included in the improvement group.Results:The implementation rate of multidisciplinary pain management plan, the rate of out-of-bed activity within 24 hours after surgery, the rate of excellent postoperative rehabilitation compliance, and the average sleep score of patients in the improvement group all increased, with statistical differences ( P<0.05). After improvement, the awareness rate of pain knowledge among medical and nursing staff, the accuracy rate of nurses' rest and active pain assessment records, and the score of nurse pain knowledge all increased, and the differences were statistically significant ( P<0.05) . Conclusions:The construction and implementation of a multidisciplinary pain management model during the perioperative period based on the project-achieving quality control circle can effectively improve the quality of pain management for surgical patients, accelerate patient recovery, and improve the pain management of medical and nursing staff.

Objective:To construct and validate prognostic model for breast cancer based on metabolic pathway-related genes.Methods:Gene expression data and clinical information of breast cancer patients were downloaded from The Cancer Genome Atlas (TCGA) website. Then all metabolic pathway-related genes were extracted from the Gene Set Enrichment Analysis (GSEA) website for differential analysis to obtain differentially expressed genes between tumor and normal tissues, and then differential metabolic genes associated with prognosis for constructing a prognostic risk score were screened by univariate Cox and Least Absolute Shrinkage and Selection Operator (LASSO) regression analysis. Patients were divided into high-risk group and low-risk group based on the median risk scores, and the efficacy of the prognostic model was evaluated using Kaplan-Meier survival analysis and receiver operating characteristic (ROC) curve analysis. The nomogram was constructed by combining this model with other clinical factors to predict the survival rate of breast cancer patients. Finally, the model was validated using the Gene Expression Omnibus (GEO) database.Results:A total of six metabolism-related genes ( NT5 E, PAICS, PFKL, PLA2 G2 D, QPRT and SHMT2) were finally screened by univariate Cox and LASSO regression for prognosis model. The prognostic risk score was an independent risk factor for breast cancer in both the training set and validating set, and the results of the Kaplan-Meier survival analysis suggested that the overall survival of patients in the high-risk group was significantly lower than that in the low-risk group, the difference was statistically significant ( P<0.001). The results of the ROC curve indicated that the nomogram model had higher predictive accuracy than other clinicopathological features, with an area under the curve value of 0.794 for both. Calibration curve showed good agreement between predicted and actual values. Based on GSEA, it was determined that the model could reveal metabolic features while monitoring the status of the tumor microenvironment (TME). Conclusion:The metabolism-related gene prognostic model constructed in this study may serve as a promising independent prognostic marker for breast cancer patients and may indicate the status of TME.