Traditional clustering algorithms for medical image segmentation can only achieve satisfactory clustering performance under relatively ideal conditions, in which there is adequate data from the same distribution, and the data is rarely disturbed by noise or outliers. However, a sufficient amount of medical images with representative manual labels are often not available, because medical images are frequently acquired with different scanners (or different scan protocols) or polluted by various noises. Transfer learning improves learning in the target domain by leveraging knowledge from related domains. Given some target data, the performance of transfer learning is determined by the degree of relevance between the source and target domains. To achieve positive transfer and avoid negative transfer, a negative-transfer-resistant mechanism is proposed by computing the weight of transferred knowledge. Extracting a n egative- t ransfer- r esistant f uzzy c lustering model with a s hared c ross-domain t ransfer latent space (called NTR-FC-SCT) is proposed by integrating negative-transfer-resistant and maximum mean discrepancy (MMD) into the framework of fuzzy c-means clustering. Experimental results show that the proposed NTR-FC-SCT model outperformed several traditional non-transfer and related transfer clustering algorithms.

A Novel Negative-Transfer-Resistant Fuzzy Clustering Model With a Shared Cross-Domain Transfer Latent Space and its Application to Brain CT Image Segmentation

Multi-agent systems enable the division of complicated tasks into individual objects that can cooperate Such architecture can be useful in building solutions in the Internet of Medical Things (IoMT) In this paper, we propose an architecture of such a system that ensures the security of private data, as well as allows the addition and/or modification of the used classification methods The main advantages of the proposed system are based on the implementation of blockchain technology elements and threaded federated learning The individual elements are located on the agents who exchange information Additionally, we propose building an agent with a consortium mechanism for classification results from many machine learning solutions This proposal offers a new model of agents that can be implemented as a system for processing medical data in real-time Our proposition was described and tested to present advantages over other, existing state-of-the-art methods We show, that this proposition can improve the Internet of Medical Thing solutions by presenting a new idea of a multi-agent system that can separate different tasks like security, or classification and as a result minimize operation time and increase accuracy

Agent architecture of an intelligent medical system based on federated learning and blockchain technology

Novel fuzzy clustering algorithm with variable multi-pixel fitting spatial information for image segmentation

Due to the narrow row spacing of corn, the lack of light in the field caused by the blocking of branches, leaves and weeds in the middle and late stages of corn growth, it is generally difficult for machinery to move between rows and also impossible to observe the corn growth in real time. To solve the problem, a robot for corn interlines information collection thus is designed. First, the mathematical model of the robot is established using the designed control system. Second, an improved convolutional neural network model is proposed for training and learning, and the driving path is fitted by detecting and identifying corn rhizomes. Next, a multi-body dynamics simulation software, RecurDyn/track, is used to establish a dynamic model of the robot movement in soft soil conditions, and a control system is developed in MATLAB/SIMULINK for joint simulation experiments. Simulation results show that the method for controlling a sliding-mode variable structure can achieve better control results. Finally, experiments on the ground and in a simulated field environment show that the robot for field information collection based on the method developed runs stably and shows little deviation. The robot can be well applied for field plant protection, the control of corn diseases and insect pests, and the realization of human–machine separation.

An Feng-Ping

Papers

Medical image segmentation algorithm based on feedback mechanism convolutional neural network