An efficient intelligent fault diagnosis model was proposed in this paper to timely and accurately offer a dependable basis for identifying the rolling bearing condition in the actual production application. The model is mainly based on an improved butterfly optimizer algorithm‐ (BOA‐) optimized kernel extreme learning machine (KELM) model. Firstly, the roller bearing’s vibration signals in the four states that contain normal state, outer race failure, inner race failure, and rolling ball failure are decomposed into several intrinsic mode functions (IMFs) using the complete ensemble empirical mode decomposition based on adaptive noise (CEEMDAN). Then, the amplitude energy entropies of IMFs are designated as the features of the rolling bearing. In order to eliminate redundant features, a random forest was used to receive the contributions of features to the accuracy of results, and subsets of features were set up by removing one feature in the descending order, using the classification accuracy of the SBOA‐KELM model as the criterion to obtain the optimal feature subset. The salp swarm algorithm (SSA) was introduced to BOA to improve optimization ability, obtain optimal KELM parameters, and avoid the BOA deteriorating into local optimization. Finally, an optimal SBOA‐KELM model was constructed for the identification of rolling bearings. In the experiment, SBOA was validated against ten other competitive optimization algorithms on 30 IEEE CEC2017 benchmark functions. The experimental results validated that the SBOA was evident over existing algorithms for most function problems. SBOA‐KELM employed for diagnosing the fault diagnosis of rolling bearings obtained improved classification performance and higher stability. Therefore, the proposed SBOA‐KELM model can be effectively used to diagnose faults of rolling bearings.

Coronavirus Disease 2019 (COVID-19) is the most severe epidemic that is prevalent all over the world. How quickly and accurately identifying COVID-19 is of great significance to controlling the spread speed of the epidemic. Moreover, it is essential to accurately and rapidly identify COVID-19 lesions by analyzing Chest X-ray images. As we all know, image segmentation is a critical stage in image processing and analysis. To achieve better image segmentation results, this paper proposes to improve the multi-verse optimizer algorithm using the Rosenbrock method and diffusion mechanism named RDMVO. Then utilizes RDMVO to calculate the maximum Kapur’s entropy for multilevel threshold image segmentation. This image segmentation scheme is called RDMVO-MIS. We ran two sets of experiments to test the performance of RDMVO and RDMVO-MIS. First, RDMVO was compared with other excellent peers on IEEE CEC2017 to test the performance of RDMVO on benchmark functions. Second, the image segmentation experiment was carried out using RDMVO-MIS, and some meta-heuristic algorithms were selected as comparisons. The test image dataset includes Berkeley images and COVID-19 Chest X-ray images. The experimental results verify that RDMVO is highly competitive in benchmark functions and image segmentation experiments compared with other meta-heuristic algorithms.

Alzheimer’s disease (AD) is an irreversible neuro-degenerative disease that severely impairs human thinking and memory. The accurate diagnosis of AD and its prodromal stages, such as mild cognitive impairment (MCI), is very important for timely treatment or possible interventions of AD. Recent studies have shown that multiple neuroimaging and biological measures contain supplementary information for diagnosis and prognosis. Most existing methods are proposed to simply integrate the multi-modal data and train the model using all samples once, which do not fully explore the structural information across the different modalities and ignore the significance of sample learning in the training process. In this paper, we propose a multi-modal self-paced locality preserving learning (MSLPL) framework to pre-serve the inherent structural relationships of the original data and realize the sample selection process from "simple" to "complex". Specifically, the model can project the neuroimaging and genetic data into the label space and learn dimensionality reduction man-ners with preserving locality structure. Meanwhile, the contribu-tions of each sample are adaptively evaluated by weighting optimi-zation so that the impact of noises can be reduced during model training by self-paced learning (SPL). Finally, a multi-kernel sup-port vector machine (MK-SVM) is used to fuse the features se-lected from different modalities for the final prediction. We eval-uate MSLPL on 913 subjects from the Alzheimer’s Disease Neu-roimaging Initiative (ADNI) database with imaging and genetic data. The experimental results demonstrate that the proposed method can achieve better classification performances compared with the start-of-the-art multimodality-based methods.

Thyroid nodule is a clinical disorder with a high incidence rate, with large number of cases being detected every year globally. Early analysis of a benign or malignant thyroid nodule using ultrasound imaging is of great importance in the diagnosis of thyroid cancer. Although the b‐mode ultrasound can be used to find the presence of a nodule in the thyroid, there is no existing method for an accurate and automatic diagnosis of the ultrasound image. In this pursuit, the present study envisaged the development of an ultrasound diagnosis method for the accurate and efficient identification of thyroid nodules, based on transfer learning and deep convolutional neural network. Initially, the Total Variation‐ (TV‐) based self‐adaptive image restoration method was adopted to preprocess the thyroid ultrasound image and remove the boarder and marks. With data augmentation as a training set, transfer learning with the trained GoogLeNet convolutional neural network was performed to extract image features. Finally, joint training and secondary transfer learning were performed to improve the classification accuracy, based on the thyroid images from open source data sets and the thyroid images collected from local hospitals. The GoogLeNet model was established for the experiments on thyroid ultrasound image data sets. Compared with the network established with LeNet5, VGG16, GoogLeNet, and GoogLeNet (Improved), the results showed that using GoogLeNet (Improved) model enhanced the accuracy for the nodule classification. The joint training of different data sets and the secondary transfer learning further improved its accuracy. The results of experiments on the medical image data sets of various types of diseased and normal thyroids showed that the accuracy rate of classification and diagnosis of this method was 96.04%, with a significant clinical application value.