scholarly journals Assessment of dysplasia in bone marrow smear with convolutional neural network

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Jinichi Mori ◽  
Shizuo Kaji ◽  
Hiroki Kawai ◽  
Satoshi Kida ◽  
Masaharu Tsubokura ◽  
...  
Keyword(s):  
Neural Network ◽  
Bone Marrow ◽  
Predictive Value ◽  
Blood Cells ◽  
Deep Neural Network ◽  
Cross Validation ◽  
Natural Images ◽  
Point Scale ◽  
Sensitivity Specificity ◽  
Fold Cross Validation

Abstract In this study, we developed the world's first artificial intelligence (AI) system that assesses the dysplasia of blood cells on bone marrow smears and presents the result of AI prediction for one of the most representative dysplasia—decreased granules (DG). We photographed field images from the bone marrow smears from patients with myelodysplastic syndrome (MDS) or non-MDS diseases and cropped each cell using an originally developed cell detector. Two morphologists labelled each cell. The degree of dysplasia was evaluated on a four-point scale: 0–3 (e.g., neutrophil with severely decreased granules were labelled DG3). We then constructed the classifier from the dataset of labelled images. The detector and classifier were based on a deep neural network pre-trained with natural images. We obtained 1797 labelled images, and the morphologists determined 134 DGs (DG1: 46, DG2: 77, DG3: 11). Subsequently, we performed a five-fold cross-validation to evaluate the performance of the classifier. For DG1–3 labelled by morphologists, the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were 91.0%, 97.7%, 76.3%, 99.3%, and 97.2%, respectively. When DG1 was excluded in the process, the sensitivity, specificity, PPV, NPV, and accuracy were 85.2%, 98.9%, 80.6%, and 99.2% and 98.2%, respectively.

scholarly journals Himawari-8 Aerosol Optical Depth (AOD) Retrieval Using a Deep Neural Network Trained Using AERONET Observations

2020 ◽  
Vol 12 (24) ◽  
pp. 4125
Author(s):  
Lu She ◽  
Hankui K. Zhang ◽  
Zhengqiang Li ◽  
Gerrit de Leeuw ◽  
Bo Huang
Keyword(s):  
Neural Network ◽  
Random Forest ◽  
Aerosol Optical Depth ◽  
Optical Depth ◽  
Land Surface ◽  
Deep Neural Network ◽  
Cross Validation ◽  
Japan Meteorological Agency ◽  
Pixel Location ◽  
Fold Cross Validation

Spectral aerosol optical depth (AOD) estimation from satellite-measured top of atmosphere (TOA) reflectances is challenging because of the complicated TOA-AOD relationship and a nexus of land surface and atmospheric state variations. This task is usually undertaken using a physical model to provide a first estimate of the TOA reflectances which are then optimized by comparison with the satellite data. Recently developed deep neural network (DNN) models provide a powerful tool to represent the complicated relationship statistically. This study presents a methodology based on DNN to estimate AOD using Himawari-8 Advanced Himawari Imager (AHI) TOA observations. A year (2017) of AHI TOA observations over the Himawari-8 full disk collocated in space and time with Aerosol Robotic Network (AERONET) AOD data were used to derive a total of 14,154 training and validation samples. The TOA reflectance in all six AHI solar bands, three TOA reflectance ratios derived based on the dark-target assumptions, sun-sensor geometry, and auxiliary data are used as predictors to estimate AOD at 500 nm. The DNN AOD is validated by separating training and validation samples using random k-fold cross-validation and using AERONET site-specific leave-one-station-out validation, and is compared with a random forest regression estimator and Japan Meteorological Agency (JMA) AOD. The DNN AOD shows high accuracy: (1) RMSE = 0.094, R2 = 0.915 for k-fold cross-validation, and (2) RMSE = 0.172, R2 = 0.730 for leave-one-station-out validation. The k-fold cross-validation overestimates the DNN accuracy as the training and validation samples may come from the same AHI pixel location. The leave-one-station-out validation reflects the accuracy for large-area applications where there are no training samples for the pixel location to be estimated. The DNN AOD has better accuracy than the random forest AOD and JMA AOD. In addition, the contribution of the dark-target derived TOA ratio predictors is examined and confirmed, and the sensitivity to the DNN structure is discussed.

scholarly journals Deep Neural Network Approach for Continuous ECG‐Based Automated External Defibrillator Shock Advisory System During Cardiopulmonary Resuscitation

2021 ◽  
Vol 10 (6) ◽  
Author(s):  
Shirin Hajeb‐M ◽  
Alicia Cascella ◽  
Matt Valentine ◽  
K. H. Chon
Keyword(s):  
Neural Network ◽  
Cardiopulmonary Resuscitation ◽  
Sensitivity And Specificity ◽  
Deep Neural Network ◽  
Cross Validation ◽  
Automated External Defibrillator ◽  
Data Set ◽  
Automated External Defibrillators ◽  
Fold Cross Validation ◽  
External Defibrillator

Background Because chest compressions induce artifacts in the ECG, current automated external defibrillators instruct the user to stop cardiopulmonary resuscitation (CPR) while an automated rhythm analysis is performed. It has been shown that minimizing interruptions in CPR increases the chance of survival. Methods and Results The objective of this study was to apply a deep‐learning algorithm using convolutional layers, residual networks, and bidirectional long short‐term memory method to classify shockable versus nonshockable rhythms in the presence and absence of CPR artifact. Forty subjects' data from Physionet with 1131 shockable and 2741 nonshockable samples contaminated with 43 different CPR artifacts that were acquired from a commercial automated external defibrillator during asystole were used. We had separate data as train and test sets. Using our deep neural network model, the sensitivity and specificity of the shock versus no‐shock decision for the entire data set over the 4‐fold cross‐validation sets were 95.21% and 86.03%, respectively. This result was based on the training and testing of the model using ECG data in both the presence and the absence of CPR artifact. For ECG without CPR artifact, the sensitivity was 99.04% and the specificity was 95.2%. A sensitivity of 94.21% and a specificity of 86.14% were obtained for ECG with CPR artifact. In addition to 4‐fold cross‐validation sets, we also examined leave‐one‐subject‐out validation. The sensitivity and specificity for the case of leave‐one‐subject‐out validation were 92.71% and 97.6%, respectively. Conclusions The proposed trained model can make shock versus nonshock decision in automated external defibrillators, regardless of CPR status. The results meet the American Heart Association's sensitivity requirement (>90%).

scholarly journals PREDIKSI TINGKAT KESUKSESAN PROMOSI BANK DENGAN ALGORITMA DNN

2021 ◽  
Vol 21 (1) ◽  
pp. 23-33
Author(s):  
Oscar Oscar ◽  
Nurlaelatul Maulidah ◽  
Annida Purnamawati ◽  
Destiana Putri ◽  
Hilman F Pardede
Keyword(s):  
Neural Network ◽  
Random Forest ◽  
Success Rate ◽  
Deep Neural Network ◽  
Cross Validation ◽  
Nearest Neighbor ◽  
K Nearest Neighbor ◽  
Bank Marketing ◽  
Fold Cross Validation ◽  
Better Than

Telemarketing is one effective way for promoting products. However, it is often difficult to measure the success of telemarketing. Therefore, a way to predict the success rate of telemarketing, and hence strategies could be planned to increase the success rate. In this study, we evaluate several implementations of machine learning for prediction the success of telemarketing. The evaluated methods are Deep Neural Network (DNN), Random Forest, and K-nearest neighbor (K-NN). We validate our experiments using 10-fold cross validation and our experiments show that DNN with 3 hidden layers outperforms other methods. Accuracy of 90% is achieved with the DNN. It is better than Random Forest and KNN that achieve accuracies of algorithm and 88% and 89%.Keywords— Bank Marketing, DNN, KNN, Random Forest.

scholarly journals Atrial Fibrillation Detection by the Combination of Recurrence Complex Network and Convolution Neural Network

2019 ◽  
Vol 2019 ◽  
pp. 1-9 ◽  
Author(s):  
Xiaoling Wei ◽  
Jimin Li ◽  
Chenghao Zhang ◽  
Ming Liu ◽  
Peng Xiong ◽  
...  
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Complex Network ◽  
Individual Variation ◽  
Deep Neural Network ◽  
Detection Algorithm ◽  
Convolution Neural Network ◽  
Electrocardiogram Signal ◽  
Sensitivity Specificity ◽  
Atrial Fibrillation Detection

In this paper, R wave peak interval independent atrial fibrillation detection algorithm is proposed based on the analysis of the synchronization feature of the electrocardiogram signal by a deep neural network. Firstly, the synchronization feature of each heartbeat of the electrocardiogram signal is constructed by a Recurrence Complex Network. Then, a convolution neural network is used to detect atrial fibrillation by analyzing the eigenvalues of the Recurrence Complex Network. Finally, a voting algorithm is developed to improve the performance of the beat-wise atrial fibrillation detection. The MIT-BIH atrial fibrillation database is used to evaluate the performance of the proposed method. Experimental results show that the sensitivity, specificity, and accuracy of the algorithm can achieve 94.28%, 94.91%, and 94.59%, respectively. Remarkably, the proposed method was more effective than the traditional algorithms to the problem of individual variation in the atrial fibrillation detection.

scholarly journals CT-Based Radiomics Analysis for Preoperative Diagnosis of Pancreatic Mucinous Cystic Neoplasm and Atypical Serous Cystadenomas

2021 ◽  
Vol 11 ◽  
Author(s):  
Tiansong Xie ◽  
Xuanyi Wang ◽  
Zehua Zhang ◽  
Zhengrong Zhou
Keyword(s):  
Predictive Value ◽  
Cross Validation ◽  
Clinical Decision Making ◽  
Area Under The Curve ◽  
Clinical Decision ◽  
Cystic Neoplasm ◽  
Cystic Neoplasms ◽  
Serous Cystadenomas ◽  
Radiological Model ◽  
Fold Cross Validation

ObjectivesTo investigate the value of CT-based radiomics analysis in preoperatively discriminating pancreatic mucinous cystic neoplasms (MCN) and atypical serous cystadenomas (ASCN).MethodsA total of 103 MCN and 113 ASCN patients who underwent surgery were retrospectively enrolled. A total of 764 radiomics features were extracted from preoperative CT images. The optimal features were selected by Mann-Whitney U test and minimum redundancy and maximum relevance method. The radiomics score (Rad-score) was then built using random forest algorithm. Radiological/clinical features were also assessed for each patient. Multivariable logistic regression was used to construct a radiological model. The performance of the Rad-score and the radiological model was evaluated using 10-fold cross-validation for area under the curve (AUC), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy.ResultsTen screened optimal features were identified and the Rad-score was then built based on them. The radiological model was built based on four radiological/clinical factors. In the 10-fold cross-validation, the Rad-score was proved to be robust and reliable (average AUC: 0.784, sensitivity: 0.847, specificity: 0.745, PPV: 0.767, NPV: 0.849, accuracy: 0.793). The radiological model performed slightly less well in classification (average AUC: average AUC: 0.734 sensitivity: 0.748, specificity: 0.705, PPV: 0.732, NPV: 0.798, accuracy: 0.728.ConclusionsThe CT-based radiomics analysis provided promising performance for preoperatively discriminating MCN from ASCN and showed good potential in improving diagnostic power, which may serve as a novel tool for guiding clinical decision-making for these patients.

scholarly journals Coordinate Transformation between Global and Local Datums Based on Artificial Neural Network with K-Fold Cross-Validation: A Case Study, Ghana

2019 ◽  
Vol 23 (1) ◽  
pp. 67-77 ◽  
Author(s):  
Yao Yevenyo Ziggah ◽  
Hu Youjian ◽  
Alfonso Rodrigo Tierra ◽  
Prosper Basommi Laari
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Coordinate Transformation ◽  
Cross Validation ◽  
Data Partitioning ◽  
Transformation Model ◽  
Data Partition ◽  
Data Set ◽  
Artificial Neural ◽  
Fold Cross Validation

The popularity of Artificial Neural Network (ANN) methodology has been growing in a wide variety of areas in geodesy and geospatial sciences. Its ability to perform coordinate transformation between different datums has been well documented in literature. In the application of the ANN methods for the coordinate transformation, only the train-test (hold-out cross-validation) approach has usually been used to evaluate their performance. Here, the data set is divided into two disjoint subsets thus, training (model building) and testing (model validation) respectively. However, one major drawback in the hold-out cross-validation procedure is inappropriate data partitioning. Improper split of the data could lead to a high variance and bias in the results generated. Besides, in a sparse dataset situation, the hold-out cross-validation is not suitable. For these reasons, the K-fold cross-validation approach has been recommended. Consequently, this study, for the first time, explored the potential of using K-fold cross-validation method in the performance assessment of radial basis function neural network and Bursa-Wolf model under data-insufficient situation in Ghana geodetic reference network. The statistical analysis of the results revealed that incorrect data partition could lead to a false reportage on the predictive performance of the transformation model. The findings revealed that the RBFNN and Bursa-Wolf model produced a transformation accuracy of 0.229 m and 0.469 m, respectively. It was also realised that a maximum horizontal error of 0.881 m and 2.131 m was given by the RBFNN and Bursa-Wolf. The obtained results per the cadastral surveying and plan production requirement set by the Ghana Survey and Mapping Division are applicable. This study will contribute to the usage of K-fold cross-validation approach in developing countries having the same sparse dataset situation like Ghana as well as in the geodetic sciences where ANN users seldom apply the statistical resampling technique.

scholarly journals Classification of Brain Tumors from MRI Images Using a Convolutional Neural Network

2020 ◽  
Vol 10 (6) ◽  
pp. 1999 ◽  
Author(s):  
Milica M. Badža ◽  
Marko Č. Barjaktarović
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Brain Tumors ◽  
Convolutional Neural Network ◽  
Cross Validation ◽  
Magnetic Resonance Images ◽  
Generalization Capability ◽  
Data Set ◽  
Fold Cross Validation

The classification of brain tumors is performed by biopsy, which is not usually conducted before definitive brain surgery. The improvement of technology and machine learning can help radiologists in tumor diagnostics without invasive measures. A machine-learning algorithm that has achieved substantial results in image segmentation and classification is the convolutional neural network (CNN). We present a new CNN architecture for brain tumor classification of three tumor types. The developed network is simpler than already-existing pre-trained networks, and it was tested on T1-weighted contrast-enhanced magnetic resonance images. The performance of the network was evaluated using four approaches: combinations of two 10-fold cross-validation methods and two databases. The generalization capability of the network was tested with one of the 10-fold methods, subject-wise cross-validation, and the improvement was tested by using an augmented image database. The best result for the 10-fold cross-validation method was obtained for the record-wise cross-validation for the augmented data set, and, in that case, the accuracy was 96.56%. With good generalization capability and good execution speed, the new developed CNN architecture could be used as an effective decision-support tool for radiologists in medical diagnostics.

Rule Extraction from Privacy Preserving Neural Network: Application to Banking

2011 ◽  
Vol 403-408 ◽  
pp. 920-928 ◽  
Author(s):  
Nekuri Naveen ◽  
V. Ravi ◽  
C. Raghavendra Rao
Keyword(s):  
Neural Network ◽  
Data Mining ◽  
Privacy Preservation ◽  
Cross Validation ◽  
Hybrid Approach ◽  
Rule Extraction ◽  
Privacy Preserving ◽  
Preservation Method ◽  
Network Application ◽  
Fold Cross Validation

In the last two decades in areas like banking, finance and medical research privacy policies restrict the data owners to share the data for data mining purpose. This issue throws up a new area of research namely privacy preserving data mining. In this paper, we proposed a privacy preservation method by employing Particle Swarm Optimization (PSO) trained Auto Associative Neural Network (PSOAANN). The modified (privacy preserved) input values are fed to a decision tree (DT) and a rule induction algorithm viz., Ripper for rule extraction purpose. The performance of the hybrid is tested on four benchmark and bankruptcy datasets using 10-fold cross validation. The results are compared with those obtained using the original datasets where privacy is not preserved. The proposed hybrid approach achieved good results in all datasets.

scholarly journals Evolving super stimuli for real neurons using deep generative networks

2019 ◽  
Author(s):  
Carlos R. Ponce ◽  
Will Xiao ◽  
Peter F. Schade ◽  
Till S. Hartmann ◽  
Gabriel Kreiman ◽  
...  
Keyword(s):  
Neural Network ◽  
Genetic Algorithm ◽  
Visual Cortex ◽  
Neural Coding ◽  
Deep Neural Network ◽  
Natural Images ◽  
List Type ◽  
Neuronal Responses ◽  
Diagnostic Features ◽  
Neural Selectivity

AbstractFinding the best stimulus for a neuron is challenging because it is impossible to test all possible stimuli. Here we used a vast, unbiased, and diverse hypothesis space encoded by a generative deep neural network model to investigate neuronal selectivity in inferotemporal cortex without making any assumptions about natural features or categories. A genetic algorithm, guided by neuronal responses, searched this space for optimal stimuli. Evolved synthetic images evoked higher firing rates than even the best natural images and revealed diagnostic features, independently of category or feature selection. This approach provides a way to investigate neural selectivity in any modality that can be represented by a neural network and challenges our understanding of neural coding in visual cortex.HighlightsA generative deep neural network interacted with a genetic algorithm to evolve stimuli that maximized the firing of neurons in alert macaque inferotemporal and primary visual cortex.The evolved images activated neurons more strongly than did thousands of natural images.Distance in image space from the evolved images predicted responses of neurons to novel images.

Sign in / Sign up

Export Citation Format

Share Document