Search

Deep learning based medical image analysis has played an important role in the computer-aided diagnosis and treatment for diseases. The accuracy of classification has always been the primary goal pursued by researchers. However, the transmission process of images also faces the problems of limited bandwidth in WAN and increased risks of data security. Additionally, individual privacy is vulnerable when user data are exposed to an unauthorized user. To address these problems, this paper constructs a system model for collaborative analysis of diagnosis of diabetic retinopathy (DR). This model consists of two stages: data cleaning and lesion classification. In the data cleaning phase, the private cloud writes the trained model into the blockchain, other private clouds can use the best-performing model shared by private clouds on the blockchain to identify the image quality and transfer high-quality image to the lesion classification model for use. In the classification stage of lesions, each private cloud uses high-quality images for classification and uploads its model parameters to the public cloud for aggregation to obtain a global model. Then, the public cloud sends the global model to each private cloud to achieve collaborative learning, reduce the amount of data transferred, and protect personal privacy. The access control scheme includes the improved role-based access control (RAC) used within the private cloud and the blockchain-based access control scheme (BAC) used during the interaction between the private cloud and the public cloud. RAC can grant both functional and data access permissions to roles, and consider object attributes to realize fine-grained control. BAC is based on certificateless public key cryptography technology and blockchain technology, which can realize identity authentication and permission identification of private cloud while requesting to transfer model parameters from private cloud to public cloud, protect the identity, permission and model parameters of private cloud, and achieve lightweight access control. Two retinal datasets are utilized for the classification of DR. Experimental results demonstrate that data cleaning can efficiently remove low quality images and improve the accuracy of the classifica-tion for early lesions of DR. The accuracy is up to 90.2%.

Disassembly lines are widely utilized to disassemble end-of-life products. Most of the research focuses on the complete disassembly of obsolete products. However, there is a lack of studies on profit and on carbon emission saved. Hence, this study considers the multi-objective partial disassembly line balancing problem with AND/OR precedence relations to optimize profit, saved carbon emission and line balance simultaneously. Firstly, a multi-objective mixed-integer programming model is formulated, which could optimally solve the small number of instances with a single objective. Meanwhile, an improved multi-objective artificial bee colony algorithm is developed to generate a set of high-quality Pareto solutions. This algorithm utilizes two-layer encoding of the task permutation vector and the number of selected parts, and develops two-phase decoding to handle the precedence relation constraint and cycle time constraint. In addition, the modified employed bee phase utilizes the neighborhood operation, and the onlooker phase utilizes the crossover operator to achieve a diverse population. The modified scout phase selects a solution from the Pareto front to replace the abandoned individual to obtain a new high-quality solution. To test the performance of the proposed algorithm, the algorithm is compared with the multi-objective simulated annealing algorithm, the original multi-objective artificial bee colony algorithm and the well-known fast non-dominated genetic algorithm. The comparative study demonstrates that the proposed improvements enhance the performance of the method presented, and the proposed methodology outperforms all the compared algorithms. [ABSTRACT FROM AUTHOR]

Objective To investigate whether shikonin can regulate autophagy and promote wound healing in diabetic rats by upregulating C-X-C chemokine receptor type 4 (CXCR4). Methods A total of 24 Sprague-Dawley (SD) rats were assigned to control, diabetic, diabetic+ Shikonin, and diabetic+Shikonin+sh-CXCR4 groups (6 rats in each group) using random number table method. A diabetic rat model was established by intraperitoneal injection of streptozotocin, and a 1 cm-diameter skin punch was used to create an ulcer model on the back of the rat. Wound healing rates were calculated on days 0, 7, 14, and 21 after wounding. On day 21 after wounding, skin tissues were collected for hematoxylin & eosin and Masson trichrome staining to analyze the pathological changes of the wound. CD31 immunohistochemistry was performed to analyze the microvascular density of the wound and Western blotting was used to detect the expression of CXCR4 and cell autophagy-related proteins microtubule-related proteins light chain 3 (LC3) II/ LC3 I, Beclin1, and soluble p62 in the wound tissue. Autophagosomes were observed by transmission electron microscopy. One-way analysis of variance was used for comparison between groups. Results Compared with the control group, the diabetic group had reduced wound healing rate (on day 21: 72.44%±6.21% vs. 93.71%±6.76%), neovascularization, collagen deposition, microvascular density, levels of CXCR4, LC3 II /LC3 I, and Beclin, and increased p62 levels in the wound tissue, as well as decreased autophagosomes (P<0.05). Compared with the diabetic group, wound healing rate (on day 21: 86.74%±4.86% vs. 72.44%±6.21%), neovascularization, collagen deposition, microvascular density, levels of CXCR4, LC3 II/LC3 I, and Beclin, and autophagosomes were increased, but p62 levels were decreased in the diabetic+Shikonin group (P<0.05). Compared with the diabetic+Shikonin group, the diabetic+Shikonin group had decreased wound healing rate (on day 21: 76.15%±3.85% vs. 86.74%±4.86%), neovascularization, collagen deposition, microvascular density, levels of CXCR4, LC3 II/LC3 I, and Beclin, and autophagosomes in the wound tissue, as well as increased p62 levels (P<0.05). Conclusion Shikonin promotes the healing of diabetic ulcer wounds by upregulating CXCR4 expression and activating cell autophagy. [ABSTRACT FROM AUTHOR]

Given that a U-shaped assembly line is usually regarded as a serial flow production system, preventive maintenance (PM) on machines in any station of the line may lead to stoppage of the whole line and sequentially cause great loss of production cost. To improve production continuity during PM, this paper proposes multiple alternative assignment plans with interchangeable abilities to ascertain the production recovery from changes of line structure. A mixed-integer mathematical model is further formulated to make decisions on alternative assignment plans. In this model, cycle time and total assignment plan alteration cost are minimised simultaneously. And an enhanced JAYA algorithm is developed to obtain well-distributed Pareto frontier solutions effectively and efficiently. The proposed algorithm suggests a similarity-based selection and various crossover/neighbourhood search operators to enhance its search ability for global optima. Numerical experiment results prove the effectiveness of the proposed model and enhanced JAYA. This novel methodology of integration optimisation bridges assembly line balancing and PM, and thus promotes production continuity and achieves economic benefits under maintenance within large-scaled assembly lines in real situations. Furthermore, a real-world case study is conducted to illustrate the significance of this novel methodology. [ABSTRACT FROM AUTHOR]