The monitoring of pregnant women is very important. It plays an important role in reducing fetal mortality, ensuring the safety of perinatal mother and fetus, preventing premature delivery and pregnancy accidents. At present, regular examination is the mainstream method for pregnant women's monitoring, but the means of examination out of hospital is scarce, and the equipment of hospital monitoring is expensive and the operation is complex. Using intelligent information technology (such as machine learning algorithm) can analyze the physiological signals of pregnant women, so as to realize the early detection and accident warning for mother and fetus, and achieve the purpose of high-quality monitoring out of hospital. However, at present, there are not enough public research reports related to the intelligent processing methods of out-of-hospital monitoring for pregnant women, so this paper takes the out-of-hospital monitoring for pregnant women as the research background, summarizes the public research reports of intelligent processing methods, analyzes the advantages and disadvantages of the existing research methods, points out the possible problems, and expounds the future development trend, which could provide reference for future related researches.
The diagnosis of pancreatic cancer is very important. The main method of diagnosis is based on pathological analysis of microscopic image of Pap smear slide. The accurate segmentation and classification of images are two important phases of the analysis. In this paper, we proposed a new automatic segmentation and classification method for microscopic images of pancreas. For the segmentation phase, firstly multi-features Mean-shift clustering algorithm (MFMS) was applied to localize regions of nuclei. Then, chain splitting model (CSM) containing flexible mathematical morphology and curvature scale space corner detection method was applied to split overlapped cells for better accuracy and robustness. For classification phase, 4 shape-based features and 138 textural features based on color spaces of cell nuclei were extracted. In order to achieve optimal feature set and classify different cells, chain-like agent genetic algorithm (CAGA) combined with support vector machine (SVM) was proposed. The proposed method was tested on 15 cytology images containing 461 cell nuclei. Experimental results showed that the proposed method could automatically segment and classify different types of microscopic images of pancreatic cell and had effective segmentation and classification results. The mean accuracy of segmentation is 93.46%±7.24%. The classification performance of normal and malignant cells can achieve 96.55%±0.99% for accuracy, 96.10%±3.08% for sensitivity and 96.80%±1.48% for specificity.
With the rapid improvement of the perception and computing capacity of mobile devices such as smart phones, human activity recognition using mobile devices as the carrier has been a new research hot-spot. The inertial information collected by the acceleration sensor in the smart mobile device is used for human activity recognition. Compared with the common computer vision recognition, it has the following advantages: convenience, low cost, and better reflection of the essence of human motion. Based on the WISDM data set collected by smart phones, the inertial navigation information and the deep learning algorithm-convolutional neural network (CNN) were adopted to build a human activity recognition model in this paper. The K nearest neighbor algorithm (KNN) and the random forest algorithm were compared with the CNN network in the recognition accuracy to evaluate the performance of the CNN network. The classification accuracy of CNN model reached 92.73%, which was much higher than KNN and random forest. Experimental results show that the CNN algorithm model can achieve more accurate human activity recognition and has broad application prospects in predicting and promoting human health.
Speech feature learning is the core and key of speech recognition method for mental illness. Deep feature learning can automatically extract speech features, but it is limited by the problem of small samples. Traditional feature extraction (original features) can avoid the impact of small samples, but it relies heavily on experience and is poorly adaptive. To solve this problem, this paper proposes a deep embedded hybrid feature sparse stack autoencoder manifold ensemble algorithm. Firstly, based on the prior knowledge, the psychotic speech features are extracted, and the original features are constructed. Secondly, the original features are embedded in the sparse stack autoencoder (deep network), and the output of the hidden layer is filtered to enhance the complementarity between the deep features and the original features. Third, the L1 regularization feature selection mechanism is designed to compress the dimensions of the mixed feature set composed of deep features and original features. Finally, a weighted local preserving projection algorithm and an ensemble learning mechanism are designed, and a manifold projection classifier ensemble model is constructed, which further improves the classification stability of feature fusion under small samples. In addition, this paper designs a medium-to-large-scale psychotic speech collection program for the first time, collects and constructs a large-scale Chinese psychotic speech database for the verification of psychotic speech recognition algorithms. The experimental results show that the main innovation of the algorithm is effective, and the classification accuracy is better than other representative algorithms, and the maximum improvement is 3.3%. In conclusion, this paper proposes a new method of psychotic speech recognition based on embedded mixed sparse stack autoencoder and manifold ensemble, which effectively improves the recognition rate of psychotic speech.
Amyloid β-protein (Aβ) deposition is an important prevention and treatment target for Alzheimer’s disease (AD), and early detection of Aβ deposition in the brain is the key to early diagnosis of AD. Magnetic resonance imaging (MRI) is the perfect imaging technology for the clinical diagnosis of AD, but it cannot display the plaque deposition directly. In this paper, based on two feature selection modes-filter and wrapper, chain-like agent genetic algorithm (CAGA), principal component analysis (PCA), support vector machine (SVM) and random forest (RF), we designed six kinds of feature learning classification algorithms to detect the information (distribution) of Aβ deposition through magnetic resonance image pixels selection. Firstly, we segmented the brain region from brain MR images. Secondly, we extracted the pixels in the segmented brain region as a feature vector (features) according to rows. Thirdly, we conducted feature learning on the extracted features, and obtained the final optimal feature subset by voting mechanism. Finally, using the final optimal selected features, we could find and mark the corresponding pixels on the MR images to show the information about Aβ plaque deposition by elastic mapping. According to the experimental results, the proposed pixel features learning methods in this paper could extract and reflect Aβ plaque deposition, and the best classification accuracy could be as high as 80%, thereby showing the effectiveness of the methods. The proposed methods can precisely detect the information of the Aβ plaque deposition, thereby being helpful for improving classification accuracy of diagnosis of AD.
Diagnosis of Parkinson’s disease (PD) based on speech data has been proved to be an effective way in recent years. However, current researches just care about the feature extraction and classifier design, and do not consider the instance selection. Former research by authors showed that the instance selection can lead to improvement on classification accuracy. However, no attention is paid on the relationship between speech sample and feature until now. Therefore, a new diagnosis algorithm of PD is proposed in this paper by simultaneously selecting speech sample and feature based on relevant feature weighting algorithm and multiple kernel method, so as to find their synergy effects, thereby improving classification accuracy. Experimental results showed that this proposed algorithm obtained apparent improvement on classification accuracy. It can obtain mean classification accuracy of 82.5%, which was 30.5% higher than the relevant algorithm. Besides, the proposed algorithm detected the synergy effects of speech sample and feature, which is valuable for speech marker extraction.
Methods for achieving diagnosis of Parkinson’s disease (PD) based on speech data mining have been proven effective in recent years. However, due to factors such as the degree of disease of the data collection subjects and the collection equipment and environment, there are different categories of sample aliasing in the sample space of the acquired data set. Samples in the aliased area are difficult to be identified effectively, which seriously affects the classification accuracy of the algorithm. In order to solve this problem, a partition bagging ensemble learning is proposed in this article, which measures the aliasing degree of the sample by designing the the ratio of sample centroid distance metrics and divides the training set into multiple subsets. And then the method of transfer training of misclassified samples is used to adjust the results of subset partitioning. Finally, the optimized weights of each sub-classifier are used to integrate the test results. The experimental results show that the classification accuracy of the proposed method is significantly improved on two public datasets and the increasement of mean accuracy is up to 25.44%. This method not only effectively improves the classification accuracy of PD speech dataset, but also increases the sample utilization rate, providing a new idea for the diagnosis of PD.