The post-marketing clinical safety re-evaluation studies of traditional Chinese medicine injections have obtained safety evidence of various research types such as active monitoring, passive monitoring and literature review. However, there is a lack of comprehensive evaluation methods that can effectively integrate the data of the above research types. So far, it is impossible to further produce more comprehensive and objective high-level evidence-based evidence, which seriously affects the supervision and management of traditional Chinese medicine injections and clinical rational use. The key to establishment of a comprehensive evaluation method is to first establish a comprehensive evaluation of the core indicators of the preferred method, the formation of weighted quantitative scoring model applied to the comprehensive evaluation method. Mixed methods research (MMR) can effectively and deeply integrate different types of research data and scientifically and normatively complete the screening of indicators in the evaluation model through repeated quantitative and qualitative research on data. Secondly, for the most critical index weighting and weight adjustment research in the model construction research, the author innovatively combines the analytic hierarchy process with the invariant weight sub-constraint method, and introduces the quantitative research part of the MMR design. It ensures the accurate weighting of indicators in the process of model construction. Therefore, based on the research on the core outcome set proposed for the core outcome outcomes of the effectiveness test, this paper proposes the use of MMR to carry out index screening and weight adjustment research based on multi-source complex data, and to construct a comprehensive evaluation model of post-marketing clinical safety of traditional Chinese medicine injections that integrates different research types of data. It provides measurement tools and new methods for the comprehensive evaluation of post-marketing clinical safety of traditional Chinese medicine injections.
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 T2KD 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 T2KD 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 (2.2×106 vs. 106.1×106, 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.