Comparison Of Artificial Neural Network and Decision Tree Methods for Predicting the Maternal Outcome in A Tertiary Care Hospital in Odisha, India

Background: This study used an artificial neural network (ANN) and a decision tree to predict maternal outcomes and their major determinants. An artificial neural network (ANN) and a decision tree were used in this study to determine maternal outcomes and their significant determinants. Methods: Data was gathered from 955 pregnant women at a tertiary care hospital in Bhubaneswar, Od-isha. A popular machine learning algorithm, artificial neural networks (ANN), was used to predict mater-nal outcomes and their determinants. Results: In the bivariate analysis, we found gestational age is significantly associated with maternal out-come (p=<0.001). The accuracy of the ANN model and decision tree was 0.882 and 0.823, respectively. Based on the variable importance of ANN, the significant determinants of maternal outcome were birth weight, systolic blood pressure, haemoglobin, gestational age, age of mother, diastolic blood pressure etc. Conclusion: This model can be utilized in future for Proper precautions and medical check-ups required during the maternal period to avoid a negative maternal outcome.

Children's pneumonia diagnosis system based on Mach-Zehnder optical fiber sensing technology / 国际生物医学工程杂志

Objective:To explore a fast and accurate method to diagnose children's pneumonia according to respiratory signals, so as to avoid the cancer induction caused by traditional X-ray examination.Methods:A Mach Zehnder optical fiber sensor was used to build a respiratory signals(RSPs) detection system, and the RSPs of the monitored children were extracted according to the vibration signal generated by the children's lung rales. Preprocessing methods such as the discrete cosine transform(DCT) were used to compress and denoise the RSPs. Multi-feature extraction of RSPs was conducted through signal processing methods such as the Hilbert transform and autoregressive (AR) model spectrum estimation. A support vector machine (SVM) classification model was constructed to classify the collected RSPs.Results:The accuracy rate of the proposed RSP classification of children with or without pneumonia was 94.41%, which was higher than the previous methods.Conclusions:The children's pneumonia diagnosis system based on an optical fiber sensor has a higher detection accuracy, and is expected to be widely used in clinical practice.

Classification of glioma and solitary brain metastasis using magnetic resonance imaging radiomics feature / 解剖学报

Objective To analyze the difference of radiomics features between solitary brain metastasis and glioma using routine 3T TI, T2 and fluid attenuation inversion recovery (FLAIR) magnetic resonance imaging, to explore the significance of texture features constructed in different directions and angles in tumor regions in distinguishing the two kinds of tumors, and to explore a feasible method for high-precision classification of solitary brain metastases and gliomas. Methods Given the multimodal images of 43 patients with glioma and 45 age- and sex- matched patients with solitary brain metastasis, the gray level co-occurrence matrices of different angles of each slice were constructed from the transverse, coronal and sagittal directions of the tumor regions of these images, and the texture spatial relationship features (including contrast, correlation, energy and homogeneity) were calculated. Wilcoxon rank sum test was used to eliminate redundant features and select features with strong distinguishing ability. Finally, SVM linear kernel classifier was used to classify the selected features to achieve the identification of the two kinds of tumors. Results When classifying glioma and solitary brain metastasis, the precision, recall, Fl score and accuracy of multimodal and multidirectional combination features were 0.8857, 0.9114, 0.8944 and 0.8922, respectively. The area under the receiver operating characteristic curve obtained by linear kernel SVM classifier was 0. 9602. Totally 40 of the 45 patients with solitary brain metastases were correctly classified, and 39 of the 43 gliomas were correctly classified. Conclusion The multimodal and multi-directional combination features of tumor areas can be classified by linear kernel SVM classifier to distinguish gliomas from solitary brain metastases, which can be used as a second opinion to effectively assist doctors in making diagnosis.

Regression Imputation and Optimized Gaussian Naïve Bayes Algorithm for an Enhanced Diabetes Mellitus Prediction Model

Abstract Diabetes mellitus (DM) is a category of metabolic disorders caused by high blood sugar. The DM affects human metabolism, and this disease causes many complications like Heart disease, Neuropathy, Diabetic retinopathy, kidney problems, skin disorder and slow healing. It is therefore essential to predict the presence of DM using an automated diabetes diagnosis system, which can be implemented using machine learning algorithms. A variety of automated diabetes prediction systems have been proposed in previous studies. Even so, the low prediction accuracy of DM prediction systems is a major issue. This proposed work developed a diabetes mellitus prediction system to improve the diabetes mellitus prediction accuracy using Optimized Gaussian Naive Bayes algorithm. This proposed model using the Pima Indians diabetes dataset as an input to build the DM predictive model. The missing values of an input dataset are imputed using regression imputation method. The sequential backward feature elimination method is used in this proposed model for selecting the relevant risk factors of diabetes disease. The proposed machine learning classifier named Optimized Gaussian Naïve Bayes (OGNB) is applied to the selected risk factors to create an enhanced Diabetes diagnostic system which predicts Diabetes in an individual. The performance analysis of this prediction architecture shows that, over other traditional machine learning classifiers, the Optimized Gaussian Naïve Bayes achieves an 81.85% classifier accuracy. This proposed DM prediction system is effective as compared to other diabetes prediction systems found in the literature. According to our experimental study, the OGNB based diabetes mellitus prediction system is more appropriate for DM disease prediction.

Dynamic Model for Assisted Reproductive Technology Outcome Prediction

Abstract Infertility is becoming a growing issue in almost all countries. Assisted Reproductive Technologies (ART) are recent development in treating infertility that give hope to the infertile couples. However, the pregnancy rates achieved with the aid of ART is considerably low, as success in ART is not only based on the treatment but also on many other controllable and uncontrollable biological, social, and environmental features. High expenditures and painful process of ART cycles are the two major barriers for opting for ART. Moreover, ART treatments are not covered by any health insurance schemes. Computational prediction models could be used to improve the success rate by predicting the treatment outcome, before the start of an ART cycle. This may suggest the couples and the doctors to decide on the next course of action i.e. either to opt for ART or opt for correcting determinants or quit the ART. With the intension to improve the success rate of ART by providing decision support system to the physicians as well to the patients before entering into the treatment this research work proposes a dynamic model for ART outcome prediction using Machine Learning (ML) techniques. The proposed dynamic model is partially implemented with the help of an ensemble of heterogeneous incremental classifier and its performance is compared with state-of-art classifiers such as Naïve Bayes (NB), Random Forest (RF), K-star etc.,using ART dataset. Performance of the model is evaluated with various metrics such as accuracy, Precision Recall Curve (PRC), Receiver Operating Characteristic (ROC), F-Measure etc., However, ROC cure area is taken as the chief metric. Evaluation results shows that the model achieves the performance with the ROC area value of 94.1 %.

Detecting anomalous growth of skin lesion using threshold-based segmentation algorithm and Fuzzy K-Nearest Neighbor classifier

Context: Skin cancer is a complex and life-threatening disease caused primarily by genetic instability and accumulation of multiple molecular alternations. Aim: Currently, there is a great interest in the prospects of image processing to provide quantitative information about a skin lesion, that can be relevance for the clinical images and also used as a stand-alone cautioning tool. Setting and Design: To accomplish a powerful approach to recognize skin cancer without performing any unnecessary skin biopsies, this article presents a new hybrid technique for the classification of skin images using Firefly with K-Nearest Neighbor algorithm (FKNN). Materials and Methods: FKNN classifier is used to predict and classify skin cancer along with threshold-based segmentation and ABCD feature extraction. Image preprocessing and feature extraction techniques are mandatory for any image-based applications. Statistical Analysis Used: Initially, it is essential to eliminate the illumination variation and the other unwanted shadow areas present in the skin image, which is done by homomorphic filtering called preprocessing. Results: The comparison of our proposed method with other existing methods and a comprehensive discussion is explored based on the obtained results. Conclusion: The proposed FKNN provides a quantitative information about a skin lesion through hybrid KNN and firefly optimization that helps for recognizing the skin cancer efficiently than other technique with low computational complexity and time

A Bootstrapping Method for Improving the Classification Performance of the P300 Speller

Abstract In this paper, we present a novel approach to training classifiers in a speller based on P300 potentials. The method, based on bootstrapping, is a known strategy for generating new samples, but it is rarely used in neurosciences. The study first demonstrates how the performance of the classification task (detecting P300 and Non-P300 classes) could be sub-optimal in the traditional approach. Then, a new method for taking new samples from the training data is proposed. Each classifier is re-trained using balanced sub-groups of individual P300 and non-P300 samples. Data were collected from 14 healthy subjects, using 16 electroencephalography channels. These were filtered in bandpass and decimated. Subsequently, four linear classifiers were trained using the traditional method followed by the proposed one, with 1000, 2000 and 3000 samples per class. Results indicate an improvement in the accuracy and discrimination capacity of discriminative classifiers with the proposed method, maintaining the same statistical properties between the training and test data. By contrast, for generative classifiers, there is no significant difference in the results. Therefore, the proposed method is highly recommended for training discriminative classifiers in spell-based P300 potentials.

Resumen Este artículo presenta un método novedoso para entrenar clasificadores en un deletreador basado en potenciales P300. El método, basado en bootstrapping, es una estrategia conocida para generar nuevas muestras pero escasamente implementado en neurociencias. El estudio muestra cómo el rendimiento de la detección de P300 (frente a No-P300) puede resultar sub-óptimo usando el método tradicional. Luego, se propone un nuevo método donde se toman nuevas muestras a partir de los datos de entrenamiento. Con ellas, se re-entrena al clasificador usando sub-grupos equilibrados de muestras individuales P300 y No-P300. Los datos se recolectaron de 14 sujetos sanos, usando 16 canales de electroencefalografía. Estos fueron filtrados en pasa-banda y diezmados. Posteriormente, cuatro clasificadores lineales fueron entrenados, usando primero el método tradicional y después el método propuesto, con 1000, 2000 y 3000 muestras por clase. Los resultados muestran una mejoría en la precisión y la capacidad de discriminación de clasificadores discriminativos con el método propuesto, manteniendo las mismas propiedades estadísticas entre los datos de entrenamiento y los de prueba. En contraste, para los clasificadores generativos, no existe una diferencia significativa en los resultados. Por consiguiente, el método propuesto es altamente recomendado para entrenar clasificadores discriminativos en deletreadores basados en potenciales P300.

Single Patient Classifier Assay, Microsatellite Instability, and Epstein-Barr Virus Status Predict Clinical Outcomes in Stage II/III Gastric Cancer: Results from CLASSIC Trial

PURPOSE: Clinical implications of single patient classifier (SPC) and microsatellite instability (MSI) in stage II/III gastric cancer have been reported. We investigated SPC and the status of MSI and Epstein-Barr virus (EBV) as combinatory biomarkers to predict the prognosis and responsiveness of adjuvant chemotherapy for stage II/III gastric cancer. MATERIALS AND METHODS: Tumor specimens and clinical information were collected from patients enrolled in CLASSIC trial, a randomized controlled study of capecitabine plus oxaliplatin-based adjuvant chemotherapy. The results of nine-gene based SPC assay were classified as prognostication (SPC-prognosis) and prediction of chemotherapy benefit (SPC-prediction). Five quasimonomorphic mononucleotide markers were used to assess tumor MSI status. EBV-encoded small RNA in situ hybridization was performed to define EBV status. RESULTS: There were positive associations among SPC, MSI, and EBV statuses among 586 patients. In multivariate analysis of disease-free survival, SPC-prognosis [hazard ratio (HR): 1.879 (1.101–3.205), 2.399 (1.415–4.067), p=0.003] and MSI status (HR: 0.363, 95% confidence interval: 0.161–0.820, p=0.015) were independent prognostic factors along with age, Lauren classification, TNM stage, and chemotherapy. Patient survival of SPC-prognosis was well stratified regardless of EBV status and in microsatellite stable (MSS) group, but not in MSI-high group. Significant survival benefit from adjuvant chemotherapy was observed by SPC-Prediction in MSS and EBV-negative gastric cancer. CONCLUSION: SPC, MSI, and EBV statuses could be used in combination to predict the prognosis and responsiveness of adjuvant chemotherapy for stage II/III gastric cancer.

Prediction of rectal toxicity of radiotherapy for prostate cancer based on multi-modality feature and multi-classifiers / 南方医科大学学报

OBJECTIVE@#To evaluate rectal toxicity of radiotherapy for prostate cancer using a novel predictive model based on multi-modality and multi-classifier fusion.@*METHODS@#We retrospectively collected the clinical data from 44 prostate cancer patients receiving external beam radiation (EBRT), including the treatment data, clinical parameters, planning CT data and the treatment plans. The clinical parameter features and dosimetric features were extracted as two different modality features, and a subset of features was selected to train the 5 base classifiers (SVM, Decision Tree, K-nearest-neighbor, Random forests and XGBoost). To establish the multi-modality and multi-classifier fusion model, a multi-criteria decision-making based weight assignment algorithm was used to assign weights for each base classifier under the same modality. A repeat 5-fold cross-validation and the 4 indexes including the area under ROC curve (AUC), accuracy, sensitivity and specificity were used to evaluate the proposed model. In addition, the proposed model was compared quantitatively with different feature selection methods, different weight allocation algorithms, the model based on single mode single classifier, and two integrated models using other fusion methods.@*RESULTS@#Repeated (5 times) 5-fold cross validation of the proposed model showed an accuracy of 0.78 for distinguishing toxicity from non-toxicity with an AUC of 0.83, a specificity of 0.79 and a sensitivity of 0.76.@*CONCLUSIONS@#Compared with the models based on a single mode or a single classifier and other fusion models, the proposed model can more accurately predict rectal toxicity of radiotherapy for prostate cancer.

A TrAdaBoost-based method for detecting multiple subjects' P300 potentials / 生物医学工程学杂志

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.

Early and Accurate Model of Malignant Lung Nodule Detection System with Less False Positives

ABSTRACT The objective of this work is to identify the malignant lung nodules accurately and early with less false positives. 'Nodule' is the 3mm to 30mm diameter size tissue clusters present inside the lung parenchyma region. Segmenting such a small nodules from consecutive CT scan slices are a challenging task. In our work Auto-seed clustering based segmentation technique is used to segment all the possible nodule candidates. Efficient shape and texture features (2D and 3D) were computed to eliminate the false nodule candidates. The change in centroid position of nodule candidates from consecutive slices was used as a measure to remove the vessels. The two-stage classifier is used in this work to classify the malignant and benign nodules. First stage rule-based classifier producing 100 % sensitivity, but with high false positive of 12.5 per patient scan. The BPN based ANN classifier is used as the second-stage classifier which reduces a false positive to 2.26 per patient scan with a reasonable sensitivity of 88.8%. The Rate of Nodule Growth (RNG) was computed in our work to measure the nodules growth between the two scans of the same patient taken at different time interval. Finally, the nodule growth predictive measure was modeled through the features such as compactness (CO), mass deficit (MD), mass excess (ME) and isotropic factor(IF). The developed model results show that the nodules which have low CO, low IF, high MD and high ME values might have the potential to grow in future.

Desarrollo de una Prótesis para Desarticulado de Muñeca Controlada por Señales de Electromiografía / Development of a Myoelectric-Controlled Prosthesis for Transradial Amputees

RESUMEN En este trabajo se presenta el desarrollo y puesta en operación de una prótesis robótica para pacientes amputados con desarticulado de muñeca. Esta prótesis consiste en un prototipo de impresión 3D que tiene dos grados de libertad que permiten realizar tareas de sujeción de tipo pinza, así como la orientación de objetos mediante los movimientos de pronación y supinación. Para el control de la prótesis se utilizan dos clasificadores de manera independiente: un clasificador bayesiano implementado en la plataforma Arduino y una red neuronal artificial implementada en el software MATLAB®; ambos realizan la clasificación de los movimientos mediante la adquisición, procesamiento y extracción de índices característicos de la señal de electromiografía. El clasificador bayesiano y la red neuronal artificial obtuvieron, respectivamente, una eficiencia de 97% y 100%, lo que muestra que los índices característicos seleccionados son adecuados para realizar la clasificación de señales de electromiografía propuesta. Se logró la creación de una prótesis mioeléctrica completamente funcional que, al ser elaborada con tecnología de impresión 3D, representa una alternativa de bajo costo a aquellas ofrecidas actualmente en el mercado.

ABSTRACT In this paper, the development and operation of a robotic prosthesis for transradial amputees is presented. This prosthesis consists in a 3D-printed prototype with two degrees of freedom, allowing the user to perform grip tasks and to orientate objects through pronation and supination movements. Two classifiers were used independently to control the prosthesis: a bayesian classifier implemented in an Arduino device and an artificial neural network implemented in MATLAB® software; both classify movements through the acquisition, processing and extraction of features from the electromyography signal. The bayesian classifier and the artificial neural network achieved an efficiency of 97% and 100%, respectively, which shows that the extracted features were suitable for the proposed electromyography classification. A completely functional 3D-printed myoelectric prosthesis was achieved, and it represents a low-cost alternative to those existent in the current market.

A fully automatic method for recognizing hand configurations of Brazilian sign language

Abstract Introduction Sign language is a collection of gestures, postures, movements, and facial expressions used by deaf people. The Brazilian sign language is Libras. The use of Libras has been increased among the deaf communities, but is still not disseminated outside this community. Sign language recognition is a field of research, which intends to help the deaf community communication with non-hearing-impaired people. In this context, this paper describes a new method for recognizing hand configurations of Libras - using depth maps obtained with a Kinect® sensor. Methods The proposed method comprises three phases: hand segmentation, feature extraction, and classification. The segmentation phase is independent from the background and depends only on pixel value. The feature extraction process is independent from rotation and translation. The features are extracted employing two techniques: (2D)2LDA and (2D)2PCA. The classification employs two classifiers: a novelty classifier and a KNN classifier. A robust database is constructed for classifier evaluation, with 12,200 images of Libras and 200 gestures of each hand configuration. Results The best accuracy obtained was 96.31%. Conclusion The best gesture recognition accuracy obtained is much higher than the studies previously published. It must be emphasized that this recognition rate is obtained for different conditions of hand rotation and proximity of the depth camera, and with a depth camera resolution of only 640×480 pixels. This performance must be also credited to the feature extraction technique, and to the size standardization and normalization processes used previously to feature extraction step.

A design of automatic classification system for leukocyte / 中国医学装备

Objective:To design a automatic classification system for leukocytes in order to increase detection speed of manual microscope inspection and reduce the inaccurate detected results in clinical laboratory; and to evaluate this system.Methods: In this system, the image processing and algorithm classifying were achieved by MATLAB software consisted of digital image processing module and automatic classification module. Classification decision for discrimination function and simulated detection for this system were achieved by using automatic classification module.Results: In the simulation experiments, the detection results for sample cell demonstrated the recognition accuracy can achieve to 93% and the speed can achieve to 97.8 cells per second for this system.Conclusion: The automatic recognition and classification system for leukocyte not only reduces human consumption, but also improves the detection accuracy and detection speed for leukocyte, and it has some significant in clinical application.

Establishment and validation of classifiers to predict genotoxic and non-genotoxic carcinogens / 中国药学杂志

OBJECTIVE: To establish classifiers to predict genotoxic and non-genotoxic carcinogens using toxicogenomics methods, explore the effect of exposure time and validated the prediction performance of the classifiers. METHODS: The primary mouse hepatocyte model was treated for 24 and 48 h with two genotoxic carcinogens, aflatoxin Bl (AFB1), benzo(a) pyrene (BAP) and two non-genotoxic carcinogens, thioacetamide (TAA), wyeth-14643 (WY). The differentially expressed genes were input to prediction analysis for microarray (PAM) software to screen out classifiers. The functions and interrelations of genes in classifiers were studied by gene set enrichment analysis (GSEA) and the protein-protein interactions were predicted using STRING database. Two additional carcinogens to validate the prediction performance of the classifiers were used. Finally, the experiment of QuantiGene Multiplex assay (Q-GP) to validate the microarray data was used. RESULTS: Forty-eight h classifiers had a better predicted capability than that of 24 h classifiers. p53 pathway, TNF-α signaling pathway, fatty acid metabolism, PPAR signaling pathway involved in the classifires were enriched by GSEA. Carcinogenic protein-protein interaction network and metabolism-related protein-protein interaction network are obtained using STRING database. The predicted probability of the two additional carcinogens using 48 h classifiers was nearly 100% and data between QuantiGene Multiplex assay and microarray assay had a high conformity. CONCLUSION: The classifiers which could be used to discriminate the potential genotoxic carcinogens and non-genotoxic carcinogens and to predict modes of action for unknown compounds, are successfully established and validated. This might be a promising candidate in vitro method for carcinogenicity study in the field of nonclinical safety evaluation of drugs.

An Efficient Human Identification through MultiModal Biometric System

ABSTRACT Human identification is essential for proper functioning of society. Human identification through multimodal biometrics is becoming an emerging trend, and one of the reasons is to improve recognition accuracy. Unimodal biometric systems are affected by various problemssuch as noisy sensor data,non-universality, lack of individuality, lack of invariant representation and susceptibility to circumvention.A unimodal system has limited accuracy. Hence, Multimodal biometric systems by combining more than one biometric feature in different levels are proposed in order to enhance the performance of the system. A supervisor module combines the different opinions or decisions delivered by each subsystem and then make a final decision. In this paper, a multimodal biometrics authentication is proposed by combining face, iris and finger features. Biometric features are extracted by Local Derivative Ternary Pattern (LDTP) in Contourlet domain and an extensive evaluation of LDTP is done using Support Vector Machine and Nearest Neighborhood Classifier. The experimental evaluations are performed on a public dataset demonstrating the accuracy of the proposed system compared with the existing systems. It is observed that, the combination of face, fingerprint and iris gives better performance in terms of accuracy, False Acceptance Rate, False Rejection Rate with minimum computation time.

Determination of Geographical Location ofGastrodia ElataUsing NIR / 世界科学技术-中医药现代化

Gastrodia elatais graded as top medication in theShen Nong’s Herbal Classic. It was mainly distributed in southwest China. Its quality varied with geographical location. And the quality difference between wild and cultivated sample was extreme. Identifications using traditional methods were unable to accurately distinguish the quality ofG. elata. Therefore, near-infrared (NIR) spectroscopy combined with pattern recognition method was used to distinguish the quality ofG. elata from different geographical locations as well as cultivated or wild. The results demonstrated that using NIR spectroscopy combined with multiclass classification algorithm, the geographical location ofG. elata can be accurately distinguished. The prediction accuracy can reach as high as 94.3% and 96.4% for both applications. Besides, the classification model was built without preprocessing; hence, it can be extended to be applied on-site.

Proposing the novelty classifier for face recognition

INTRODUCTION: Face recognition, one of the most explored themes in biometry, is used in a wide range of applications: access control, forensic detection, surveillance and monitoring systems, and robotic and human machine interactions. In this paper, a new classifier is proposed for face recognition: the novelty classifier. METHODS: The performance of a novelty classifier is compared with the performance of the nearest neighbor classifier. The ORL face image database was used. Three methods were employed for characteristic extraction: principal component analysis, bi-dimensional principal component analysis with dimension reduction in one dimension and bi-dimensional principal component analysis with dimension reduction in two directions. RESULTS: In identification mode, the best recognition rate with the leave-one-out strategy is equal to 100%. In the verification mode, the best recognition rate was also 100%. For the half-half strategy, the best recognition rate in the identification mode is equal to 98.5%, and in the verification mode, 88%. CONCLUSION: For face recognition, the novelty classifier performs comparable to the best results already published in the literature, which further confirms the novelty classifier as an important pattern recognition method in biometry.

Leukocytes Detection, Classification and Counting in Smears of Peripheral Blood / Detección, Clasificación y Conteo de Leucocitos en Frotis de Sangre Periférica

Using the k-NN classifier in combination with the first Minkowski metric, in addition to techniques of digital image processing, we developed a computational system platform-independent, which is able to identify, to classify and to count five normal types of leukocytes: neutrophils, eosinophils, basophils, monocytes and lymphocytes. It is important to emphasize that this work does not attempt to diferentiate between smears of leukocytes coming from healthy and sick people; this is because most diseases produce a change in the differential count of leukocytes rather than in theirs forms. In the other side, the system could be used in emerging areas such as the topographic hematology and the chronobiology.

Mediante un clasificador k-NN en combinación con la primera métrica de Minkowski y técnicas de procesamiento digital de imágenes, se desarrolló un sistema computacional independiente de la plataforma, capaz de identificar, clasificar y contar cinco formas normales de leucocitos: neutrófilos, eosinófilos, basófilos, monocitos y linfocitos. Es importante enfatizar que este trabajo no intenta diferenciar entre muestras de leucocitos provenientes de gente sana y enferma, debido a que la mayoría de las enfermedades se detectan principalmente por un cambio en el conteo diferencial de leucocitos más que por cambios en su forma. Finalmente, el contador de leucocitos puede ser usado en áreas emergentes como la hematología topográfica y la cronobiología.

Rapid Identification of Traditional Chinese Medicine Using Electronic Nose Based on RBF-RF Cascade Classifier / 世界科学技术-中医药现代化

This study was aimed to apply the electronic nose (E-nose) in the research of traditional Chinese medicine (TCM). The discussion was made on difficulties of using E-nose. The solution plan was proposed and the discrimination model was established. It provided a simple, rapid and effective analysi method in the identification of TCM. It also provided new ideas for the research and application of gas sensor arrays. E-nose was used in the ex-traction of TCM scent characteristics. Based on ion mobility spectrometry of MOS sensor, the fingerprint of TCM scent was established. The maximum response value of the sensor was used as analysis index. According to the diffi-culties of identification, two solution plans were proposed. Firstly, different detectors were employed to complete the classification. Secondly, radial basis function (RBF) and random forests (RF) were combined and then a cascade classifier was constructed in order to achieve the maximum of information obtained in conditions where the number of measurements, metal oxide semiconductor sensors in E-nose was limited. The results showed that both plans were accurate and practical with relatively high upper correct judge rate and better cross-validation (The highest upper correct judge rates were 95% and 100%, 96% and 80%, respectively). It was concluded that this study firstly ap-plied cascade classifier in the establishment of TCM identification by E-nose. With limited amount of sensors, the maximum information was received through data mining. Using E-nose in the identification of TCM was rapid and accurate. The established pattern recognition method was maneuverable with accurate identification rate and stability compared to conventional sensory identification method. It provided a simple and rapid analysis method for the iden-tification of TCM.