Lung cancer is one of the tumors with the highest incidence rate and mortality rate in the world. It is also the malignant tumor with the fastest growing number of patients, which seriously threatens human life. How to improve the accuracy of diagnosis and treatment of lung cancer and the survival prognosis is particularly important. Machine learning is a multi-disciplinary interdisciplinary specialty, covering the knowledge of probability theory, statistics, approximate theory and complex algorithm. It uses computer as a tool and is committed to simulating human learning methods, and divides the existing content into knowledge structures to effectively improve learning efficiency and being able to integrate computer science and statistics into medical problems. Through the introduction of algorithm to absorb the input data, and the application of computer analysis to predict the output value within the acceptable accuracy range, identify the patterns and trends in the data, and finally learn from previous experience, the development of this technology brings a new direction for the diagnosis and treatment of lung cancer. This article will review the performance and application prospects of different types of machine learning algorithms in the clinical diagnosis and survival prognosis analysis of lung cancer.
At present the prediction method of epilepsy patients is very time-consuming and vulnerable to subjective factors, so this paper presented an automatic recognition method of epilepsy electroencephalogram (EEG) based on common spatial model (CSP) and support vector machine (SVM). In this method, the CSP algorithm for extracting spatial characteristics was applied to the detection of epileptic EEG signals. However, the algorithm did not consider the nonlinear dynamic characteristics of the signals and ignored the time-frequency information, so the complementary characteristics of standard deviation, entropy and wavelet packet energy were selected for the combination in the feature extraction stage. The classification process adopted a new double classification model based on SVM. First, the normal, interictal and ictal periods were divided into normal and paroxysmal periods (including interictal and ictal periods), and then the samples belonging to the paroxysmal periods were classified into interictal and ictal periods. Finally, three categories of recognition were realized. The experimental data came from the epilepsy study at the University of Bonn in Germany. The average recognition rate was 98.73% in the first category and 99.90% in the second category. The experimental results show that the introduction of spatial characteristics and double classification model can effectively solve the problem of low recognition rate between interictal and ictal periods in many literatures, and improve the identification efficiency of each period, so it provides an effective detecting means for the prediction of epilepsy.
By analyzing the mechanism of pronunciation, traditional acoustic parameters, including fundamental frequency, Mel frequency cepstral coefficients (MFCC), linear prediction cepstrum coefficient (LPCC), frequency perturbation, amplitude perturbation, and nonlinear characteristic parameters, including entropy (sample entropy, fuzzy entropy, multi-scale entropy), box-counting dimension, intercept and Hurst, are extracted as feature vectors for identification of pathological voice. Seventy-eight normal voice samples and 73 pathological voice samples for /a/, and 78 normal samples and 80 pathological samples for /i/ are recognized based on support vector machine (SVM). The results showed that compared with traditional acoustic parameters, nonlinear characteristic parameters could be well used to distinguish between healthy and pathological voices, and the recognition rates for /a/ were all higher than those for /i/ except for multi-scale entropy. That is why the /a/ sound data is used widely in related research at home and abroad for obtaining better identification of pathological voices. Adopting multi-scale entropy for /i/ could obtain higher recognition rate than /a/ between healthy and pathological samples, which may provide some useful inspiration for evaluating vocal compensatory function.
Individual differences of P300 potentials lead to that a large amount of training data must be collected to construct pattern recognition models in P300-based brain-computer interface system, which may cause subjects’ fatigue and degrade the system performance. TrAdaBoost is a method that transfers the knowledge from source area to target area, which improves learning effect in the target area. Our research purposed a TrAdaBoost-based linear discriminant analysis and a TrAdaBoost-based support vector machine to recognize the P300 potentials across multiple subjects. This method first trains two kinds of classifiers separately by using the data deriving from a small amount of data from same subject and a large amount of data from different subjects. Then it combines all the classifiers with different weights. Compared with traditional training methods that use only a small amount of data from same subject or mixed different subjects’ data to directly train, our algorithm improved the accuracies by 19.56% and 22.25% respectively, and improved the information transfer rate of 14.69 bits/min and 15.76 bits/min respectively. The results indicate that the TrAdaBoost-based method has the potential to enhance the generalization ability of brain-computer interface on the individual differences.
The purpose of using brain-computer interface (BCI) is to build a bridge between brain and computer for the disable persons, in order to help them to communicate with the outside world. Electroencephalography (EEG) has low signal to noise ratio (SNR), and there exist some problems in the traditional methods for the feature extraction of EEG, such as low classification accuracy, lack of spatial information and huge amounts of features. To solve these problems, we proposed a new method based on time domain, frequency domain and space domain. In this study, independent component analysis (ICA) and wavelet transform were used to extract the temporal, spectral and spatial features from the original EEG signals, and then the extracted features were classified with the method combined support vector machine (SVM) with genetic algorithm (GA). The proposed method displayed a better classification performance, and made the mean accuracy of the Graz datasets in the BCI Competitions of 2003 reach 96%. The classification results showed that the proposed method with the three domains could effectively overcome the drawbacks of the traditional methods based solely on time-frequency domain when the EEG signals were used to describe the characteristics of the brain electrical signals.
Alzheimer’ s disease is the most common kind of dementia without effective treatment. Via early diagnosis, early intervention after diagnosis is the most effective way to handle this disease. However, the early diagnosis method remains to be studied. Neuroimaging data can provide a convenient measurement for the brain function and structure. Brain structure network is a good reflection of the fiber structural connectivity patterns between different brain cortical regions, which is the basis of brain’s normal psychology function. In the paper, a brain structure network based on pattern recognition analysis was provided to realize an automatic diagnosis research of Alzheimer’s disease and gray matter based on structure information. With the feature selection in pattern recognition, this method can provide the abnormal regions of brain structural network. The research in this paper analyzed the patterns of abnormal structural network in Alzheimer’s disease from the aspects of connectivity and node, which was expected to provide updated information for the research about the pathological mechanism of Alzheimer’s disease.
ObjectiveTo establish a machine learning model based on computed tomography (CT) radiomics for preoperatively predicting invasive degree of lung ground-glass nodules (GGNs). MethodsWe retrospectively analyzed the clinical data of GGNs patients whose solid component less than 3 cm in the Department of Thoracic Surgery of Shanghai Pulmonary Hospital from March 2021 to July 2021 and the First Hospital of Lanzhou University from January 2019 to May 2022. The lesions were divided into pre-invasiveness and invasiveness according to postoperative pathological results, and the patients were randomly divided into a training set and a test set in a ratio of 7∶3. Radiomic features (1 317) were extracted from CT images of each patient, the max-relevance and min-redundancy (mRMR) was used to screen the top 100 features with the most relevant categories, least absolute shrinkage and selection operator (LASSO) was used to select radiomic features, and the support vector machine (SVM) classifier was used to establish the prediction model. We calculated the area under the curve (AUC), sensitivity, specificity, accuracy, negative predictive value, positive predictive value to evaluate the performance of the model, drawing calibration and decision curves of the prediction model to evaluate the accuracy and clinical benefit of the model, analyzed the performance in the training set and subgroups with different nodule diameters, and compared the prediction performance of this model with Mayo and Brock models. Two primary thoracic surgeons were required to evaluate the invasiveness of GGNs to investigate the clinical utility of the model. ResultsA total of 400 patients were divided into the training set (n=280) and the test set (n=120) according to the admission criteria. There were 267 females and 133 males with an average age of 52.4±12.7 years. Finally, 8 radiomic features were screened out from the training set data to build SVM model. The AUC, sensitivity and specificity of the model in the training and test sets were 0.91, 0.89, 0.75 and 0.86, 0.92, 0.60, respectively. The model showed good prediction performance in the training set 0-10 mm, 10-20 mm and the test set 0-10 mm, 10-20 mm subgroups, with AUC values of 0.82, 0.88, 0.84, 0.72, respectively. The AUC of SVM model was significantly better than that of Mayo model (0.73) and Brock model (0.73). With the help of this model, the AUC value, sensitivity, specificity and accuracy of thoracic surgeons A and B in distinguishing invasive or non-invasive adenocarcinoma were significantly improved. ConclusionThe SVM model based on radiomics is helpful to distinguish non-invasive lesions from invasive lesions, and has stable predictive performance for GGNs of different sizes and has better prediction performance than Mayo and Brock models. It can help clinicians to more accurately judge the invasiveness of GGNs, to make more appropriate diagnosis and treatment decisions, and achieve accurate treatment.
The clinical manifestations of patients with schizophrenia and patients with depression not only have a certain similarity, but also change with the patient's mood, and thus lead to misdiagnosis in clinical diagnosis. Electroencephalogram (EEG) analysis provides an important reference and objective basis for accurate differentiation and diagnosis between patients with schizophrenia and patients with depression. In order to solve the problem of misdiagnosis between patients with schizophrenia and patients with depression, and to improve the accuracy of the classification and diagnosis of these two diseases, in this study we extracted the resting-state EEG features from 100 patients with depression and 100 patients with schizophrenia, including information entropy, sample entropy and approximate entropy, statistical properties feature and relative power spectral density (rPSD) of each EEG rhythm (δ, θ, α, β). Then feature vectors were formed to classify these two types of patients using the support vector machine (SVM) and the naive Bayes (NB) classifier. Experimental results indicate that: ① The rPSD feature vector P performs the best in classification, achieving an average accuracy of 84.2% and a highest accuracy of 86.3%; ② The accuracy of SVM is obviously better than that of NB; ③ For the rPSD of each rhythm, the β rhythm performs the best with the highest accuracy of 76%; ④ Electrodes with large feature weight are mainly concentrated in the frontal lobe and parietal lobe. The results of this study indicate that the rPSD feature vector P in conjunction with SVM can effectively distinguish depression and schizophrenia, and can also play an auxiliary role in the relevant clinical diagnosis.
The study was intended to introduce a novel method for aided diagnosis of cardiovascular diseases based on photoplethysmography (PPG). For this purpose, 40 volunteers were recruited in this study, of whom the physiological and pathological information was collected, including blood pressure and simultaneous PPG data on fingertips, by using a sphygmomanometer and a smart fingertip sensor. According to the PPG signal and its first and second derivatives, 52 features were defined and acquired. The Relief feature selection algorithm was performed to calculate the contribution of each feature to cardiovascular diseases. And then 10 core features which had the greatest contribution were selected as an optimal feature subset. Finally, the efficiency of the Relief feature selection algorithm was demonstrated by the results of k-nearest neighbor (kNN) and support vector machine (SVM) classifier applications of the features. The prediction accuracy of kNN model and SVM reached 66.67% and 83.33% respectively, indicating that: ① Age was the foremost feature for aided diagnosis of cardiovascular diseases; ② The optimal feature subset provided an important evaluation of cardiovascular health status. The obtained results showed that the application of the Relief feature selection algorithm provided high accuracy in aided diagnosis of cardiovascular diseases.
This paper explores a methodology used to discriminate the electroencephalograph (EEG) signals of patients with vegetative state (VS) and those with minimally conscious state (MCS). The model was derived from the EEG data of 33 patients in a calling name stimulation paradigm. The preprocessing algorithm was applied to remove the noises in the EEG data. Two types of features including sample entropy and multiscale entropy were chosen. Multiple kernel support vector machine was investigated to perform the training and classification. The experimental results showed that the alpha rhythm features of EEG signals in severe disorders of consciousness were significant. We achieved the average classification accuracy of 88.24%. It was concluded that the proposed method for the EEG signal classification for VS and MCS patients was effective. The approach in this study may eventually lead to a reliable tool for identifying severe disorder states of consciousness quantitatively. It would also provide the auxiliary basis of clinical assessment for the consciousness disorder degree.