Detection of microcalcification clusters regions in mammograms combining discriminative deep belief networks / 生物医学工程学杂志

In order to overcome the shortcomings of high false positive rate and poor generalization in the detection of microcalcification clusters regions, this paper proposes a method combining discriminative deep belief networks (DDBNs) to automatically and quickly locate the regions of microcalcification clusters in mammograms. Firstly, the breast region was extracted and enhanced, and the enhanced breast region was segmented to overlapped sub-blocks. Then the sub-block was subjected to wavelet filtering. After that, DDBNs model for breast sub-block feature extraction and classification was constructed, and the pre-trained DDBNs was converted to deep neural networks (DNN) using a softmax classifier, and the network is fine-tuned by back propagation. Finally, the undetected mammogram was inputted to complete the location of suspicious lesions. By experimentally verifying 105 mammograms with microcalcifications from the Digital Database for Screening Mammography (DDSM), the method obtained a true positive rate of 99.45% and a false positive rate of 1.89%, and it only took about 16 s to detect a 2 888 × 4 680 image. The experimental results showed that the algorithm of this paper effectively reduced the false positive rate while ensuring a high positive rate. The detection of calcification clusters was highly consistent with expert marks, which provides a new research idea for the automatic detection of microcalcification clusters area in mammograms.

Automatic classification method of arrhythmia based on discriminative deep belief networks / 生物医学工程学杂志

Existing arrhythmia classification methods usually use manual selection of electrocardiogram (ECG) signal features, so that the feature selection is subjective, and the feature extraction is complex, leaving the classification accuracy usually affected. Based on this situation, a new method of arrhythmia automatic classification based on discriminative deep belief networks (DDBNs) is proposed. The morphological features of heart beat signals are automatically extracted from the constructed generative restricted Boltzmann machine (GRBM), then the discriminative restricted Boltzmann machine (DRBM) with feature learning and classification ability is introduced, and arrhythmia classification is performed according to the extracted morphological features and RR interval features. In order to further improve the classification performance of DDBNs, DDBNs are converted to deep neural network (DNN) using the Softmax regression layer for supervised classification in this paper, and the network is fine-tuned by backpropagation. Finally, the Massachusetts Institute of Technology and Beth Israel Hospital Arrhythmia Database (MIT-BIH AR) is used for experimental verification. For training sets and test sets with consistent data sources, the overall classification accuracy of the method is up to 99.84% ± 0.04%. For training sets and test sets with inconsistent data sources, a small number of training sets are extended by the active learning (AL) method, and the overall classification accuracy of the method is up to 99.31% ± 0.23%. The experimental results show the effectiveness of the method in arrhythmia automatic feature extraction and classification. It provides a new solution for the automatic extraction of ECG signal features and classification for deep learning.

Epidemiological characteristics and spatial-temporal clustering analysis on pulmonary tuberculosis in Changsha from 2013 to 2016 / 中南大学学报(医学版)

To analyze the epidemic characteristics and to explore the spatial-temporal clusters of pulmonary tuberculosis (PTB) in Changsha from 2013 to 2016.
 Methods: Descriptive analysis and space-time permutation scan statistic were used to analyze the reported PTB cases in Changsha from 2013 to 2016.
 Results: Between 2013 and 2016, a total of 17 721 PTB cases were reported in Changsha, with annual reported incidence rate at 60.87 per 100 000 population. Males and individuals aged 15 to <60 years accounted for higher proportion of PTB cases compared to females and other age groups. The number of reported PTB cases reached peak from March to May in each year. The space-time permutation scan statistic identified one most likely cluster and two secondary clusters of PTB cases. The most likely cluster covered most areas of Liuyang City and the North-east part of Changsha County from October 1, 2013 to February 28, 2014. The first cluster occupied 12 towns (streets) in Kaifu District and Changsha County in December 2016. The second cluster included four towns (streets) in Yuhua District and Tianxin District from March 1 to September 30, in 2013.
 Conclusion: Between 2013 and 2016, significant space-time clusters of PTB cases were identified in Changsha. These findings could provide a guide for development of regional intervention strategies for PTB control.

Diagnostic value of chest X-ray film of pulmonary emphysema signs for in-fantile bronchitis / 中国感染控制杂志

Objective To evaluate the diagnostic value of chest X-ray film of pulmonary emphysema (PE)signs for infantile bronchitis.Methods Clinical data of 60 infants with bronchitis (case group)in a hospital between Janu-ary 2010 and December 2014 were analyzed retrospectively,and compared with data of 30 infants with non-respira-tory diseases (control group).Results Of 60 infants with bronchitis in case group,95.00%(57/60)showed mani-festations of PE on X-ray,and 18.33% (11/60)of whom were with mild PE(apical or intercostal pneumocele), 76.67%(46/60)were with typical PE (diaphragm descent);one case (3.33%)in control group showed intercostal pneumocele,the other 29 cases (96.67%)were normal X-ray findings and without signs of PE.The sensitivity, specificity and accuracy of PE signs on chest X-ray film for diagnosing infantile bronchitis were 95.00%,96.67%, and 95.56% respectively.Conclusion Signs of PE on chest X-ray film have important diagnosis value for infantile bronchitis.

A P-wave detection method based on multi-feature / 生物医学工程学杂志

Generally, P-wave is the wave of low-frequency and low-amplitude, and it could be affected by baseline drift, electromyography (EMG) interference and other noises easily. Not every heart beat contains the P-wave, and it is also a major problem to determine the P-wave exist or not in a heart beat. In order to solve the limitation of suiting the diverse morphological P-wave using wavelet-amplitude-transform algorithm and the limitation of selecting the pseudo-P-wave sample using the wavelet transform and neural network, we presented new P-wave detecting method based on wave-amplitude threshold and using the multi-feature as the input of neural networks. Firstly, we removed the noise of ECG through the wavelet transform, then determined the position of the candidate P-wave by calculating modulus maxima of the wavelet transform, and then determine the P-wave exist or not by wave-amplitude threshold method initially. Finally we determined whether the P-wave existed or not by the neural networks. The method is validated based on the QT database which is supplied with manual labels made by physicians. We compared the detection effect of ECG P-waves, which was obtained with the method developed in the study, with the algorithm of wavelet threshold value and the method based on "wavelet-amplitude-slope", and verified the feasibility of the proposed algorithm. The detected ECG signal, which is recorded in the hospital ECG division, was consistent with the doctor's labels. Furthermore, after detecting the 13 sets of ECG which were 15 min long, the detection rate for the correct P-wave is 99.911%.

Construction of learning teaching faculty according to the idea of ‘ school-based’ training / 中华医学教育探索杂志

We took measures to construct sound learning mechanism and a set of effective learning system,which were connected tightly with various learning ways according to the idea of ‘ school-based ’ training.Goals of learning mechanism and the set of effective learning system were to improve teacher's learning ability,enhance knowledge and competence of medical college teachers and to construct learning teaching faculty.

An annotation approach for masto-calcifications based on semantic model / 生物医学工程学杂志

To realize the medical semantic annotation of mammogram, a semantic modeling approach for micro-calcifications in mammogram based on hierarchical Bayesian network (BN) was proposed. Firstly, support vector machines (SVM) were used to map low-level image feature into feature semantics, then high-level semantic was captured through fusing the feature semantics using BN. Finally semantic model was established. To validate the method, the model was applied to annotate the semantic information of mammograms. In this experiment, 142 images were chosen as training set and 50 images as testing set. The results showed that the accuracy of malignant samples was 81.48%, and that of benign samples was 73.91%.

QRS complex detection using empirical mode decomposition and windowing technique / 生物医学工程学杂志

A QRS complex detection algorithm based on empirical mode decomposition (EMD) and adaptive windowing technique is proposed in this paper. In this algorithm we mainly used Hilbert-Huang transform to propose EMD method suitable for QRS complex detection, with which the 30th signal in sddb database and the 208th signal in mit-db database could be processed, and then obtained R wave detection results. At the same time, Q and S points' detection technique was analyzed with adaptive windowing technique. The detection results, through proceeding R wave detection on part data of MIT/BIT arrhythmia database, showed that the proposed algorithm in this paper had a very good detection effect, and that its average correct detection rate of R wave reached 99.62%, its average sensitivity of QRS complex was 98.91%, and the corresponding average specificity was 99.35%.

Research on algorithm of independent component analysis with two channel electrocardiograms / 生物医学工程学杂志

According to the independent component analysis (ICA) theory, in order to seperate the fetal electrocardiograms (FECG) from the observed data, we must use the larger number of observed signals than that of independent components. This requires that the number of the sentors on the electrocardiograph must be larger than a particular value, which is in practice hard to be satisfied. So we proposed another algorithm using fewer channels of abdominal electrocardiograms, which combines ICA and adaptive noise cancellation to extract the FECG from two leads of abdominal electrocardiograms. The experiment results showed that we could obtain clear FECG by the method we proposed.

A technology research for diagnosis of mammographic masses based on content-based image retrieval / 生物医学工程学杂志

In order to assist doctors in making the diagnosis of mammographic masses, a method is proposed in this paper. Twenty-two features are extracted from each queried region of interest (ROI). A k-nearest neighbor (KNN) algorithm is used to retrieve similar images from database, and further calculate the mutual information (MI) between the queried image and the images which are in the retrieval results, so as to improve the retrieval performance. Finally, the scheme takes the first nine images with the highest MI scores as the final retrieval results. With the purpose of providing available decision-making information of diagnostic aids, we compare and analyze three calculating methods of decision index. The experiment results show that this method is better than the method using KNN only, and this method improves the accuracy of diagnosis effectively.

An Idea on the Reform of Computer Courses in Regional Medical College / 中华医学教育探索杂志

Based on the analysis of the status quo of computer teaching,the present paper discusses the reform of teaching contents in the computer education in our college.