RESUMEN
Intermittent theta burst stimulation (iTBS) is a novel treatment approach for post-traumatic stress disorder (PTSD), and recent neuroimaging work indicates that functional connectivity profiles may be able to identify those most likely to respond. However, prior work has relied on functional magnetic resonance imaging, which is expensive and difficult to scale. Alternatively, electroencephalography (EEG) represents a different approach that may be easier to implement in clinical practice. To this end, we acquired an 8-channel resting-state EEG signal on participants before (n = 47) and after (n = 43) randomized controlled trial of iTBS for PTSD (ten sessions, delivered at 80% of motor threshold, 1,800 pulses, to the right dorsolateral prefrontal cortex). We used a cross-validated support vector machine (SVM) to track changes in EEG functional connectivity after verum iTBS stimulation. We found that an SVM classifier was able to successfully separate patients who received active treatment vs. sham treatment, with statistically significant findings in the Delta band (1-4 Hz, p = 0.002). Using Delta coherence, the classifier was 75.0% accurate in detecting sham vs. active iTBS, and observed changes represented an increase in functional connectivity between midline central/occipital and a decrease between frontal and central regions. The primary limitations of this work are the sparse electrode system and a modest sample size. Our findings raise the possibility that EEG and machine learning may be combined to provide a window into mechanisms of action of TMS, with the potential that these approaches can inform the development of individualized treatment methods.
Asunto(s)
Trastornos por Estrés Postraumático/fisiopatología , Trastornos por Estrés Postraumático/terapia , Máquina de Vectores de Soporte , Ritmo Teta , Estimulación Magnética Transcraneal , Corteza Prefontal Dorsolateral , Electroencefalografía , Humanos , Reproducibilidad de los Resultados , Descanso , Máquina de Vectores de Soporte/normasRESUMEN
Accurate classification of burn severities is of vital importance for proper burn treatments. A recent article reported that using the combination of Raman spectroscopy and optical coherence tomography (OCT) classifies different degrees of burns with an overall accuracy of 85% [1]. In this study, we demonstrate the feasibility of using Raman spectroscopy alone to classify burn severities on ex vivo porcine skin tissues. To create different levels of burns, four burn conditions were designed: (i) 200°F for 10s, (ii) 200°F for 30s, (iii) 450°F for 10s and (iv) 450°F for 30s. Raman spectra from 500-2000cm-1 were collected from samples of the four burn conditions as well as the unburnt condition. Classifications were performed using kernel support vector machine (KSVM) with features extracted from the spectra by principal component analysis (PCA), and partial least-square (PLS). Both techniques yielded an average accuracy of approximately 92%, which was independently evaluated by leave-one-out cross-validation (LOOCV). By comparison, PCA+KSVM provides higher accuracy in classifying severe burns, while PLS performs better in classifying mild burns. Variable importance in the projection (VIP) scores from the PLS models reveal that proteins and lipids, amide III, and amino acids are important indicators in separating unburnt or mild burns (200°F), while amide I has a more pronounced impact in separating severe burns (450°F).
Asunto(s)
Quemaduras/diagnóstico por imagen , Espectrometría Raman/normas , Quemaduras/complicaciones , Humanos , Análisis de Componente Principal , Índice de Severidad de la Enfermedad , Espectrometría Raman/métodos , Máquina de Vectores de Soporte/normas , Máquina de Vectores de Soporte/estadística & datos numéricosRESUMEN
Electroencephalography (EEG) is a complex bioelectrical signal. Analysis of which can provide researchers with useful physiological information. In order to recognize and classify EEG signals, a pattern recognition method for optimizing the support vector machine (SVM) by using improved squirrel search algorithm (ISSA) is proposed. The EEG signal is preprocessed, with its time domain features being extracted and directed to the SVM as feature vectors for classification and identification. In this paper, the method of good point set is used to initialize the population position, chaos and reverse learning mechanism are introduced into the algorithm. The performance test of the improved squirrel algorithm (ISSA) is carried out by using the benchmark function. As can be seen from the statistical analysis of the results, the exploration ability and convergence speed of the algorithm are improved. This is then used to optimize SVM parameters. ISSA-SVM model is established and built for classification of EEG signals, compared with other common SVM parameter optimization models. For data sets, the average classification accuracy of this method is 85.9%. This result is an improvement of 2-5% over the comparison method.
Asunto(s)
Electroencefalografía/métodos , Animales , Humanos , Sciuridae , Máquina de Vectores de Soporte/normasRESUMEN
In this article we introduce the idea of Markov resampling for Boosting methods. We first prove that Boosting algorithm with general convex loss function based on uniformly ergodic Markov chain (u.e.M.c.) examples is consistent and establish its fast convergence rate. We apply Boosting algorithm based on Markov resampling to Support Vector Machine (SVM), and introduce two new resampling-based Boosting algorithms: SVM-Boosting based on Markov resampling (SVM-BM) and improved SVM-Boosting based on Markov resampling (ISVM-BM). In contrast with SVM-BM, ISVM-BM uses the support vectors to calculate the weights of base classifiers. The numerical studies based on benchmark datasets show that the proposed two resampling-based SVM Boosting algorithms for linear base classifiers have smaller misclassification rates, less total time of sampling and training compared to three classical AdaBoost algorithms: Gentle AdaBoost, Real AdaBoost, Modest AdaBoost. In addition, we compare the proposed SVM-BM algorithm with the widely used and efficient gradient Boosting algorithm-XGBoost (eXtreme Gradient Boosting), SVM-AdaBoost and present some useful discussions on the technical parameters.
Asunto(s)
Máquina de Vectores de Soporte/normas , Cadenas de MarkovRESUMEN
One of the challenges in brain-computer interface systems is obtaining motor imagery recognition from brain activities. Brain-signal decoding robustness and system performance improvement during the motor imagery process are two of the essential issues in brain-computer interface research. In conventional approaches, ineffective decoding of features and high complexity of algorithms often lead to unsatisfactory performance. A novel method for the recognition of motor imagery tasks is developed based on employing a modified S-transforms for spectro-temporal representation to characterize the behavior of electrocorticogram activities. A classifier is trained by using a support vector machine, and an optimized wrapper approach is applied to guide selection to implement the representation selection obtained. A channel selection algorithm optimizes the wrapper approach by adding a cross-validation step, which effectively improves the classification performance. The modified S-transform can accurately capture event-related desynchronization/event-related synchronization phenomena and can effectively locate sensorimotor rhythm information. The optimized wrapper approach used in this scheme can effectively reduce the feature dimension and improve algorithm efficiency. The method is evaluated on a public electrocorticogram dataset with a recognition accuracy of 98% and an information transfer rate of 0.8586 bit/trial. To verify the effect of the channel selection, both electrocorticogram and electroencephalogram data are experimentally analyzed. Furthermore, the computational efficiency of this scheme demonstrates its potential for online brain-computer interface systems in future cognitive tasks.
Asunto(s)
Interfaces Cerebro-Computador , Corteza Cerebral/fisiología , Electrocorticografía/métodos , Imaginación/fisiología , Actividad Motora/fisiología , Reconocimiento de Normas Patrones Automatizadas/métodos , Procesamiento de Señales Asistido por Computador , Máquina de Vectores de Soporte , Adulto , Conjuntos de Datos como Asunto , Electrocorticografía/normas , Humanos , Reconocimiento de Normas Patrones Automatizadas/normas , Máquina de Vectores de Soporte/normasRESUMEN
Multiview Generalized Eigenvalue Proximal Support Vector Machine (MvGEPSVM) is an effective method for multiview data classification proposed recently. However, it ignores discriminations between different views and the agreement of the same view. Moreover, there is no robustness guarantee. In this paper, we propose an improved multiview GEPSVM (IMvGEPSVM) method, which adds a multi-view regularization that can connect different views of the same class and simultaneously considers the maximization of the samples from different classes in heterogeneous views for promoting discriminations. This makes the classification more effective. In addition, L1-norm rather than squared L2-norm is employed to calculate the distances from each of the sample points to the hyperplane so as to reduce the effect of outliers in the proposed model. To solve the resulting objective, an efficient iterative algorithm is presented. Theoretically, we conduct the proof of the algorithm's convergence. Experimental results show the effectiveness of the proposed method.
Asunto(s)
Máquina de Vectores de Soporte/normas , Reconocimiento de Normas Patrones Automatizadas/métodosRESUMEN
Lane changes made during traffic oscillations on freeways largely affect traffic safety and could increase collision potentials. Predicting the impacts of lane change can help to develop optimal lane change strategies of autonomous vehicles for safety improvement. The study aims at proposing a machine learning method for the short-term prediction of lane-changing impacts (LCI) during the propagation of traffic oscillations. The empirical lane-changing trajectory records were obtained from the Next Generation Simulation (NGSIM) platform. A support vector regression (SVR) model was trained in this study to predict the LCI on the crash risks and flow change using microscopic traffic variables such as individual speed, gap and acceleration on both original lanes and target lanes. Sensitivity analyses were conducted in the SVR to quantify the contributions of correlative lane changing factors. The results showed that the trained SVR model achieved an accuracy of 72.81 % for the risk of crashes and 95.34 % in predicting the flow change. The sensitivity analysis explored the optimal speed and acceleration for the lane changer to achieve the lowest time integrated time-to-collision (TIT) value for safety maximization. Finally, we compared the LCI for motorcycles, automobiles and trucks as well as the LCI for both lane-changing directions (from left to right and from right to left). It was found that motorcycles conducted lane changes with smaller gaps and larger speed differences, which brings the highest crash risks. Passenger cars were found to be the safest when they conduct lane changes. Lane changes to the right had more negative impacts on traffic flow and crash risks.
Asunto(s)
Accidentes de Tránsito/prevención & control , Conducción de Automóvil , Aceleración , Humanos , Seguridad , Máquina de Vectores de Soporte/normasRESUMEN
PURPOSE : Evaluating the quality of surgical procedures is a major concern in minimally invasive surgeries. We propose a bottom-up approach based on the study of Sleeve Gastrectomy procedures, for which we analyze what we assume to be an important indicator of the surgical expertise: the exposure of the surgical scene. We first aim at predicting this indicator with features extracted from the laparoscopic video feed, and second to analyze how the extracted features describing the surgical practice influence this indicator. METHOD : Twenty-nine patients underwent Sleeve Gastrectomy performed by two confirmed surgeons in a monocentric study. Features were extracted from spatial and procedural annotations of the videos, and an expert surgeon evaluated the quality of the surgical exposure at specific instants. The features were used as input of a classifier (linear discriminant analysis followed by a support vector machine) to predict the expertise indicator. Features selected in different configurations of the algorithm were compared to understand their relationships with the surgical exposure and the surgeon's practice. RESULTS : The optimized algorithm giving the best performance used spatial features as input ([Formula: see text]). It also predicted equally the two classes of the indicator, despite their strong imbalance. Analyzing the selection of input features in the algorithm allowed a comparison of different configurations of the algorithm and showed a link between the surgical exposure and the surgeon's practice. CONCLUSION : This preliminary study validates that a prediction of the surgical exposure from spatial features is possible. The analysis of the clusters of feature selected by the algorithm also shows encouraging results and potential clinical interpretations.
Asunto(s)
Algoritmos , Gastrectomía/métodos , Laparoscopía/métodos , Máquina de Vectores de Soporte/normas , Grabación en Video/métodos , HumanosRESUMEN
Triple-negative breast cancer (TNBC), colon adenocarcinoma (COAD), ovarian cancer (OV), and glioblastoma multiforme (GBM) are common malignant tumors, in which significant challenges are still faced in early diagnosis, treatment, and prognosis. Therefore, further identification of genes related to those malignant tumors is of great significance for the improvement of management of the diseases. The database of the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) repository was used as the data source of gene expression profiles in this study. Malignant tumors genes were selected using a feature selection algorithm of maximal relevance and minimal redundancy (mRMR) and the protein-protein interaction (PPI) network. And finally selected 20 genes as potential related genes. Gene Ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were performed on the potential related genes, and different tumor-specific genes and similarities and differences between network modules and pathways were analyzed. Further, using the potential cancer-related genes found above in this study as features, a support vector machine (SVM) model was developed to predict high-risk malignant tumors. As a result, the prediction accuracy reached more than 85%, indicating that such a model can effectively predict the four types of malignant tumors. It is demonstrated that such genes found above in this study indeed play important roles in the differentiation of the four types of malignant tumors, providing basis for future experimental biological validation and shedding some light on the understanding of new molecular mechanisms related to the four types of tumors.
Asunto(s)
Redes Reguladoras de Genes/genética , Neoplasias/genética , Mapas de Interacción de Proteínas/fisiología , Máquina de Vectores de Soporte/normas , Femenino , Humanos , Masculino , PronósticoRESUMEN
Soil organic matter (SOM) is a major indicator of soil fertility and nutrients. In this study, a soil organic matter measuring method based on an artificial olfactory system (AOS) was designed. An array composed of 10 identical gas sensors controlled at different temperatures was used to collect soil gases. From the response curve of each sensor, four features were extracted (maximum value, mean differential coefficient value, response area value, and the transient value at the 20th second). Then, soil organic matter regression prediction models were built based on back-propagation neural network (BPNN), support vector regression (SVR), and partial least squares regression (PLSR). The prediction performance of each model was evaluated using the coefficient of determination (R2), root-mean-square error (RMSE), and the ratio of performance to deviation (RPD). It was found that the R2 values between prediction (from BPNN, SVR, and PLSR) and observation were 0.880, 0.895, and 0.808. RMSEs were 14.916, 14.094, and 18.890, and RPDs were 2.837, 3.003, and 2.240, respectively. SVR had higher prediction ability than BPNN and PLSR and can be used to accurately predict organic matter contents. Thus, our findings offer brand new methods for predicting SOM.
Asunto(s)
Nariz Electrónica , Suelo/química , Calibración , Gases/química , Análisis de los Mínimos Cuadrados , Redes Neurales de la Computación , Máquina de Vectores de Soporte/normas , Compuestos Orgánicos Volátiles/análisis , Compuestos Orgánicos Volátiles/normasRESUMEN
OBJECTIVES: To investigate the association between proton magnetic resonance spectroscopy (1H-MRS) metabolic features and the grade of gliomas, and to establish a machine-learning model to predict the glioma grade. METHODS: This study included 112 glioma patients who were divided into the training (nâ¯=â¯74) and validation (nâ¯=â¯38) sets based on the time of hospitalization. Twenty-six metabolic features were extracted from the preoperative 1H-MRS image. The Student's t-test was conducted to screen for differentially expressed features between low- and high-grade gliomas (WHO grades II and III/IV, respectively). Next, the minimum Redundancy Maximum Relevance (mRMR) algorithm was performed to further select features for a support vector machine (SVM) classifier building. Performance of the predictive model was evaluated both in the training and validation sets using ROC curve analysis. RESULTS: Among the extracted 1H-MRS metabolic features, thirteen features were differentially expressed. Four features were further selected as grade-predictive imaging signatures using the mRMR algorithm. The predictive performance of the machine-learning model measured by the AUC was 0.825 and 0.820 in the training and validation sets, respectively. This was better than the predictive performances of individual metabolic features, the best of which was 0.812. CONCLUSIONS: 1H-MRS metabolic features could help in predicting the grade of gliomas. The machine-learning model achieved a better prediction performance in grading gliomas than individual features, indicating that it could complement the traditionally used metabolic features.
Asunto(s)
Neoplasias Encefálicas/diagnóstico por imagen , Glioma/diagnóstico por imagen , Espectroscopía de Resonancia Magnética/normas , Cuidados Preoperatorios/normas , Máquina de Vectores de Soporte/normas , Adulto , Neoplasias Encefálicas/metabolismo , Femenino , Glioma/metabolismo , Humanos , Espectroscopía de Resonancia Magnética/métodos , Masculino , Persona de Mediana Edad , Clasificación del Tumor/métodos , Clasificación del Tumor/normas , Cuidados Preoperatorios/métodos , Estudios RetrospectivosRESUMEN
OBJECTIVES: This study uses a new approach for classifying the human ethnicity according to the auditory brain responses (electroencephalography [EEG] signals) with a high level of accuracy. Moreover, the study presents three different algorithms used to classify the human ethnicity using auditory brain responses. The algorithms were tested on Malays and Chinese as a case study. MATERIALS AND METHODS: The EEG signal was used as a brain response signal, which was evoked by two auditory stimuli (Tones and Consonant Vowels stimulus). The study was carried out on Malaysians (Malay and Chinese) with normal hearing and with hearing loss. A ranking process for the subjects' EEG data and the nonlinear features was used to obtain the maximum classification accuracy. RESULTS: The study formulated the classification of Normal Hearing Ethnicity Index and Sensorineural Hearing Loss Ethnicity Index. These indices classified the human ethnicity according to brain auditory responses by using numerical values of response signal features. Three classification algorithms were used to verify the human ethnicity. Support Vector Machine (SVM) classified the human ethnicity with an accuracy of 90% in the cases of normal hearing and sensorineural hearing loss (SNHL); the SVM classified with an accuracy of 84%. CONCLUSION: The classification indices categorized or separated the human ethnicity in both hearing cases of normal hearing and SNHL with high accuracy. The SVM classifier provided a good accuracy in the classification of the auditory brain responses. The proposed indices might constitute valuable tools for the classification of the brain responses according to the human ethnicity.
Asunto(s)
Electroencefalografía/instrumentación , Potenciales Evocados Auditivos/fisiología , Pérdida Auditiva Sensorineural/fisiopatología , Pérdida Auditiva/fisiopatología , Estimulación Acústica/métodos , Adulto , Algoritmos , Audiometría de Tonos Puros/métodos , China/epidemiología , China/etnología , Etnicidad/estadística & datos numéricos , Pérdida Auditiva Sensorineural/diagnóstico , Pérdida Auditiva Sensorineural/etnología , Humanos , Lenguaje , Malasia/etnología , Masculino , Persona de Mediana Edad , Ruido/efectos adversos , Percepción del Habla/fisiología , Máquina de Vectores de Soporte/normasRESUMEN
Twin support vector machine (TWSVM) is a classical and effective classifier for binary classification. However, its robustness cannot be guaranteed due to the utilization of squared L2-norm distance that can usually exaggerate the influence of outliers. In this paper, we propose a new robust capped L1-norm twin support vector machine (CTWSVM), which sustains the advantages of TWSVM and promotes the robustness in solving a binary classification problem with outliers. The solution of the proposed method can be achieved by optimizing a pair of capped L1-norm related problems using a newly-designed effective iterative algorithm. Also, we present some theoretical analysis on existence of local optimum and convergence of the algorithm. Extensive experiments on an artificial dataset and several UCI datasets demonstrate the robustness and feasibility of our proposed CTWSVM.
Asunto(s)
Máquina de Vectores de Soporte/normas , Algoritmos , Clasificación/métodosRESUMEN
OBJECTIVE: The goal of this research is to develop a machine learning supervised classification model to automatically code clinical encounter transcripts using a behavioral code scheme. METHODS: We first evaluated the efficacy of eight state-of-the-art machine learning classification models to recognize patient-provider communication behaviors operationalized by the motivational interviewing framework. Data were collected during the course of a single weight loss intervention session with 37 African American adolescents and their caregivers. We then tested the transferability of the model to a novel treatment context, 80 patient-provider interactions during routine human immunodeficiency virus (HIV) clinic visits. RESULTS: Of the eight models tested, the support vector machine model demonstrated the best performance, achieving a .680 F1-score (a function of model precision and recall) in adolescent and .639 in caregiver sessions. Adding semantic and contextual features improved accuracy with 75.1% of utterances in adolescent and 73.8% in caregiver sessions correctly coded. With no modification, the model correctly classified 72.0% of patient-provider utterances in HIV clinical encounters with reliability comparable to human coders (k = .639). CONCLUSIONS: The development of a validated approach for automatic behavioral coding offers an efficient alternative to traditional, resource-intensive methods with the potential to dramatically accelerate the pace of outcomes-oriented behavioral research. The knowledge gained from computer-driven behavioral research can inform clinical practice by providing clinicians with empirically supported communication strategies to tailor their conversations with patients. Lastly, automatic behavioral coding is a critical first step toward fully automated eHealth/mHealth (electronic/mobile Health) behavioral interventions.
Asunto(s)
Investigación Conductal/métodos , Comunicación , Aprendizaje Automático/normas , Entrevista Motivacional , Relaciones Profesional-Paciente , Adolescente , Femenino , Humanos , Masculino , Investigación Cualitativa , Reproducibilidad de los Resultados , Máquina de Vectores de Soporte/normasRESUMEN
The aim of this study was to develop methods for detecting the nonstationary periodic characteristics of neonatal electroencephalographic (EEG) seizures by adapting estimates of the correlation both in the time (spike correlation; SC) and time-frequency domain (time-frequency correlation; TFC). These measures were incorporated into a seizure detection algorithm (SDA) based on a support vector machine to detect periods of seizure and nonseizure. The performance of these nonstationary correlation measures was evaluated using EEG recordings from 79 term neonates annotated by three human experts. The proposed measures were highly discriminative for seizure detection (median AUCSC : 0.933 IQR: 0.821-0.975, median AUCTFC : 0.883 IQR: 0.707-0.931). The resultant SDA applied to multi-channel recordings had a median AUC of 0.988 (IQR: 0.931-0.998) when compared to consensus annotations, outperformed two state-of-the-art SDAs (p < 0.001) and was noninferior to the human expert for 73/79 of neonates.
Asunto(s)
Algoritmos , Electroencefalografía/normas , Convulsiones/diagnóstico , Máquina de Vectores de Soporte/normas , Electroencefalografía/métodos , Humanos , Recién Nacido , Convulsiones/fisiopatología , Factores de TiempoRESUMEN
BACKGROUND: Lysine succinylation is a new kind of post-translational modification which plays a key role in protein conformation regulation and cellular function control. To understand the mechanism of succinylation profoundly, it is necessary to identify succinylation sites in proteins accurately. However, traditional methods, experimental approaches, are labor-intensive and time-consuming. Computational prediction methods have been proposed recent years, and they are popular because of their convenience and high speed. In this study, we developed a new method to predict succinylation sites in protein combining multiple features, including amino acid composition, binary encoding, physicochemical property and grey pseudo amino acid composition, with a feature selection scheme (information gain). And then, it was trained using SVM (Support Vector Machine) and an ensemble learning algorithm. RESULTS: The performance of this method was measured with an accuracy of 89.14% and a MCC (Matthew Correlation Coefficient) of 0.79 using 10-fold cross validation on training dataset and an accuracy of 84.5% and a MCC of 0.2 on independent dataset. CONCLUSIONS: The conclusions made from this study can help to understand more of the succinylation mechanism. These results suggest that our method was very promising for predicting succinylation sites. The source code and data of this paper are freely available at https://github.com/ningq669/PSuccE .
Asunto(s)
Biología Computacional/métodos , Máquina de Vectores de Soporte/normas , AlgoritmosRESUMEN
Imbalance problem occurs when the majority class instances outnumber the minority class instances. Conventional extreme learning machine (ELM) treats all instances with same importance leading to the prediction accuracy biased towards the majority class. To overcome this inherent drawback, many variants of ELM have been proposed like Weighted ELM, class-specific cost regulation ELM (CCR-ELM) etc. to handle the class imbalance problem effectively. This work proposes class-specific extreme learning machine (CS-ELM), a variant of ELM for handling binary class imbalance problem more effectively. This work differs from weighted ELM as it does not require assigning weights to the training instances. The proposed work also has lower computational complexity compared to weighted ELM. This work uses class-specific regularization parameters. CCR-ELM also uses class-specific regularization parameters. In CCR-ELM the computation of regularization parameters does not consider class distribution and class overlap. This work uses class-specific regularization parameters which are computed using class distribution. This work also differ from CCR-ELM in the computation of the output weight, ß. The proposed work has lower computational overhead compared to CCR-ELM. The proposed work is evaluated using benchmark real world imbalanced datasets downloaded from the KEEL dataset repository. The results show that the proposed work has better performance than weighted ELM, CCR-ELM , EFSVM, FSVM, SVM for class imbalance learning.
Asunto(s)
Máquina de Vectores de Soporte/normasRESUMEN
BACKGROUND: Diagnostic guidelines recommend using a variety of methods to assess and diagnose ADHD. Applying subjective measures always incorporates risks such as informant biases or large differences between ratings obtained from diverse sources. Furthermore, it has been demonstrated that ratings and tests seem to assess somewhat different constructs. The use of objective measures might thus yield valuable information for diagnosing ADHD. This study aims at evaluating the role of objective measures when trying to distinguish between individuals with ADHD and controls. Our sample consisted of children (n = 60) and adults (n = 76) diagnosed with ADHD and matched controls who completed self- and observer ratings as well as objective tasks. Diagnosis was primarily based on clinical interviews. A popular pattern recognition approach, support vector machines, was used to predict the diagnosis. RESULTS: We observed relatively high accuracy of 79% (adults) and 78% (children) applying solely objective measures. Predicting an ADHD diagnosis using both subjective and objective measures exceeded the accuracy of objective measures for both adults (89.5%) and children (86.7%), with the subjective variables proving to be the most relevant. CONCLUSIONS: We argue that objective measures are more robust against rater bias and errors inherent in subjective measures and may be more replicable. Considering the high accuracy of objective measures only, we found in our study, we think that they should be incorporated in diagnostic procedures for assessing ADHD.
Asunto(s)
Trastorno por Déficit de Atención con Hiperactividad/diagnóstico , Pruebas Neuropsicológicas/normas , Máquina de Vectores de Soporte/normas , Evaluación de Síntomas/métodos , Evaluación de Síntomas/normas , Adulto , Trastorno por Déficit de Atención con Hiperactividad/psicología , Niño , Humanos , Persona de Mediana EdadRESUMEN
Multi-voxel pattern analysis often necessitates feature selection due to the high dimensional nature of neuroimaging data. In this context, feature selection techniques serve the dual purpose of potentially increasing classification accuracy and revealing sets of features that best discriminate between classes. However, feature selection techniques in current, widespread use in the literature suffer from a number of deficits, including the need for extended computational time, lack of consistency in selecting features relevant to classification, and only marginal increases in classifier accuracy. In this paper we present a novel method for feature selection based on a single-layer neural network which incorporates cross-validation during feature selection and stability selection through iterative subsampling. Comparing our approach to popular alternative feature selection methods, we find increased classifier accuracy, reduced computational cost and greater consistency with which relevant features are selected. Furthermore, we demonstrate that importance mapping, a technique used to identify voxels relevant to classification, can lead to the selection of irrelevant voxels due to shared activation patterns across categories. Our method, owing to its relatively simple architecture, flexibility and speed, can provide a viable alternative for researchers to identify sets of features that best discriminate classes.
Asunto(s)
Bases de Datos Factuales/normas , Redes Neurales de la Computación , Máquina de Vectores de Soporte/normas , HumanosRESUMEN
Specific biomarker reflecting neurobiological substrates of schizophrenia (SZ) is required for its diagnosis and treatment selection of SZ. Evidence from neuroimaging has implicated disrupted functional connectivity in the pathophysiology. We aimed to develop and validate a method of disease definition for SZ by resting-state functional connectivity using radiomics strategy. This study included 2 data sets collected with different scanners. A total of 108 first-episode SZ patients and 121 healthy controls (HCs) participated in the current study, among which 80% patients and HCs (n = 183) and 20% (n = 46) were selected for training and testing in intra-data set validation and 1 of the 2 data sets was selected for training and the other for testing in inter-data set validation, respectively. Functional connectivity was calculated for both groups, features were selected by Least Absolute Shrinkage and Selection Operator (LASSO) method, and the clinical utility of its features and the generalizability of effects across samples were assessed using machine learning by training and validating multivariate classifiers in the independent samples. We found that the accuracy of intra-data set training was 87.09% for diagnosing SZ patients by applying functional connectivity features, with a validation in the independent replication data set (accuracy = 82.61%). The inter-data set validation further confirmed the disease definition by functional connectivity features (accuracy = 83.15% for training and 80.07% for testing). Our findings demonstrate a valid radiomics approach by functional connectivity to diagnose SZ, which is helpful to facilitate objective SZ individualized diagnosis using quantitative and specific functional connectivity biomarker.