Analysis of Fitting Degree of MRI and PET Images in Simultaneous MRPET Images by Machine Learning Neural Networks

2020 ◽  
Vol 3 (1) ◽  
pp. 43-61
Author(s):  
Giljae Lee ◽  
◽  
Hwunjae Lee ◽  
Gyehwan Jin ◽  
◽  
...  
Keyword(s):  
Neural Network ◽  
High Frequency ◽  
Pulse Sequences ◽  
Discrete Wavelet ◽  
Low Frequencies ◽  
Neural Network Learning ◽  
Network Learning ◽  
Signal Value ◽  
The Neural Network ◽  
Feature Values

Simultaneous MR-PET imaging is a fusion of MRI using various parameters and PET images using various nuclides. In this paper, we performed analysis on the fitting degree between MRI and simultaneous MR-PET images and between PET and simultaneous MR-PET images. For the fitness analysis by neural network learning, feature parameters of experimental images were extracted by discrete wavelet transform (DWT), and the extracted parameters were used as input data to the neural network. In comparing the feature values extracted by DWT for each image, the horizontal and vertical low frequencies showed similar patterns, but the patterns were different in the horizontal and vertical high frequency and diagonal high frequency regions. In particular, the signal value was large in the T1 and T2 weighted images of MRI. Neural network learning results for fitting degree analysis were as follows. 1. T1-weighted MRI and simultaneous MR-PET image fitting degree: Regression (R) values were found to be Training 0.984, Validation 0.844, and Testing 0.886. 2. Dementia-PET image and Simultaneous MR-PET Image fitting degree: R values were found to be Training 0.970, Validation 0.803, and Testing 0.828. 3. T2-weighted MRI and concurrent MR-PET image fitting degree: R values were found to be Training 0.999, Validation 0.908, and Testing 0.766. 4. Brain tumor-PET image and Simultaneous MR-PET image fitting degree: R values were found to be Training 0.999, Validation 0.983, and Testing 0.876. An R value closer to 1 indicates more similarity. Therefore, each image fused in the simultaneous MR-PET images verified in this study was found to be similar. Ongoing study of images acquired with pulse sequences other than the weighted images in the MRI is needed. These studies may establish a useful protocol for the acquisition of simultaneous MR-PET images.

scholarly journals Kernel-blending connection approximated by a neural network for image classification

2020 ◽  
Vol 6 (4) ◽  
pp. 467-476
Author(s):  
Xinxin Liu ◽  
Yunfeng Zhang ◽  
Fangxun Bao ◽  
Kai Shao ◽  
Ziyi Sun ◽  
...  
Keyword(s):  
Neural Network ◽  
Image Classification ◽  
Function Approximation ◽  
Feature Space ◽  
Neural Network Learning ◽  
Network Learning ◽  
Hinge Loss ◽  
The Neural Network ◽  
Good Classification ◽  
Kernel Mapping

AbstractThis paper proposes a kernel-blending connection approximated by a neural network (KBNN) for image classification. A kernel mapping connection structure, guaranteed by the function approximation theorem, is devised to blend feature extraction and feature classification through neural network learning. First, a feature extractor learns features from the raw images. Next, an automatically constructed kernel mapping connection maps the feature vectors into a feature space. Finally, a linear classifier is used as an output layer of the neural network to provide classification results. Furthermore, a novel loss function involving a cross-entropy loss and a hinge loss is proposed to improve the generalizability of the neural network. Experimental results on three well-known image datasets illustrate that the proposed method has good classification accuracy and generalizability.

Neural Network-Based Speed Control of A Two-Mass-Model System

1999 ◽  
Vol 3 (5) ◽  
pp. 427-430 ◽  
Author(s):  
Rached Dhaouadi ◽  
◽  
Khaled Nouri
Keyword(s):  
Neural Network ◽  
Inverse Dynamics ◽  
Mass Model ◽  
Neural Network Learning ◽  
Inverse Control ◽  
Network Learning ◽  
The Neural Network ◽  
Control Scheme ◽  
On Line ◽  
Neural Network Structure

We present an application of artificial neural networks to the problem of controlling the speed of an elastic drive system. We derive a neural network structure to simulate the inverse dynamics of the system, then implement the direct inverse control scheme in a closed loop. The neural network learning is done on-line to adaptively control the speed to follow a stepwise changing reference. The experimental results with a two-mass-model analog board confirm the effectiveness of the proposed neurocontrol scheme.

scholarly journals ANALYSIS OF NEURAL NETWORK CLASSIFICATION ALGORITHM TO KNOW THE SUCCESS LEVEL OF IMMUNOTHERAPY

2020 ◽  
Vol 17 (1) ◽  
pp. 57-62
Author(s):  
Agung Fazriansyah ◽  
Mochammad Abdul Azis ◽  
Yudhistira Yudhistira
Keyword(s):  
Neural Network ◽  
High Accuracy ◽  
Large Datasets ◽  
Neural Network Learning ◽  
Network Learning ◽  
Neural Network Classification ◽  
The Neural Network ◽  
Neural Network Algorithm ◽  
Auc Value ◽  
Learning Machine

Cancer is a disease that is feared by humans at this stage, the genetic term of most diseases that have the characteristics of abnormal cell growth and beyond the normal cell limits so that they can attack cells that cover and are able to spread to other organs. For cancer recovery therapy is immunization therapy. Of course in this alternative treatment still needs to be done research to determine the level of success with existing conditions and parameters. Increasingly sophisticated, developing technology that helps human work. The neural network algorithm is used to analyze large datasets, the purpose of this study is to find the accuracy and immunotherapy methods of the dataset using a neural network learning machine with 200 data training cycles, 0.9 momentum and 0.01 learning levels that produce quite high accuracy 80 % and AUC value of 0.738

scholarly journals Tutorial on Using Machine Learning for Activity Recognition Via a Smartphone

Telecom IT ◽  
2020 ◽  
Vol 8 (2) ◽  
pp. 92-108
Author(s):  
I. Zelichenok ◽  
R. Pirmagomedov
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Data Collection ◽  
Activity Recognition ◽  
Learning Process ◽  
Daily Activity ◽  
Application Development ◽  
Neural Network Learning ◽  
Network Learning ◽  
The Neural Network

This article provides a tutorial for developing a simple machine learning application in Python. More spe-cifically, the paper considers daily activity recognition using sensors of a smartphone. For development, we used TensorFlow, Skikit learn, NumPy, Pandas, and Matplotlib. The paper explains in detail the main steps of the application development, including data collection and pre-processing, design of the neural network, learning process, and use of a trained model. The overall accuracy of the developed application when recognizing the activity is about 95 %. This paper can be useful for students and specialists who want to start work on machine learning.

scholarly journals Robot Manipulators Control with Guaranteed Stability Using Feedback Error Learning Neural Networks

1996 ◽  
Vol 8 (4) ◽  
pp. 383-391
Author(s):  
Ju-Jang Lee ◽  
◽  
Sung-Woo Kim ◽  
Kang-Bark Park
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Learning Rule ◽  
Neural Network Learning ◽  
Network Learning ◽  
The Neural Network ◽  
Control Schemes ◽  
Feedback Error Learning ◽  
Feedback Error ◽  
Error Learning

Among various neural network learning control schemes, feedback error learning(FEL)8),9) has been known that it has advantages over other schemes. However, such advantages are founded on the assumption that the systems is linearly parameterized and stable. Thus, FEL has difficulties in coping with uncertain and unstable systems. Furthermore, it is not clear how the learning rule of FEL is obtained in the minimization sense. Therefore, to overcome such problems, we propose neural network control schemes using FEL with guaranteed performance. The proposed strategy is to use multi-layer neural networks, to design a stabilityguaranteeing controller(SGC), and to derive a learning rule to obtain the tracking performance. Using multilayer neural networks we can fully utilize the learning capability no matter how the system is linearly parameterized or not. The SGC makes it possible for the neural network to learn without fear of instability. As a result, the more the neural network learning proceeds, the better the tracking performance becomes.

Sign in / Sign up

Export Citation Format

Share Document