| It is generally observed that the mitochondrial diseases are caused by the inherited or spontaneous mutations in mtDNA or nDNA,which affect the original functionality of the proteins or RNA molecules that normally reside in mitochondria.The mitochondrial cell diseases have potential to disturb the normal functionality of living organisms and sometimes such diseases may be a cause of the death of a living organism.Therefore,it is essential to detect the mitochondrial cell diseases to figure out the precautionary measures.The microscopy images can be used to detect the morphology of mitochondrial cells and comprehensive analysis of the images to detect the diseased cells.However,it is trivial to analyze these images with human eye,artificial intelligence can play an important role to detect intrinsic patterns hidden inside these images.Furthermore,the classification of the mitochondrial cell diseases aims to classify the future behavior of mitochondrial diseases.Based on the analysis of microscopy image data,it can be used to identify disease in advance,thus providing a solution for the above-mentioned clinical and disease concerns.The normal and affected mitochondrial cells have different morphological properties.The diseases change the morphology of the mitochondrial cell and it is biological meaningful to detect the change occurred in the morphology of the cell and time associated with this change.Our main research is carried out to analyze the microscopy images of mitochondrial cells.The main research results and innovations are as follows:1.A normal and drug treated cell image correlation analysis algorithm is proposed which is called Improve Correlation(IC).The experimental results show the effective correlation ability of the proposed method.2.A normal and drug treated cell image recognition algorithm is proposed which is based on convolutional neural networks.The accuracy of the classification is verified by a set of experiments followed by comparisons with related traditional methods.3.A normal and decease cell classification algorithm is proposed which is based on convolutional neural networks.The variations between normal and disease cells are analyzed.Experiments are conducted to show the effectiveness of the algorithm.4.The mitochondrial cell dynamic is analyzed and a mitochondria organelle movement classification algorithm(MOMC)is proposed.The behavior of a cell can be described through tissue morphogenesis,which involves the migration,division or death of tissue,and is regulated with the molecular scale.Automated cell detection from microscopy image has become an important step in cell-based experiments.We have developed a method to detect changings occurred in the morphology of mitochondrial cell through real-time images.Our method consists of pixel classification using K-Means and Bayesian classifier as well as combination of gray level thresholding.The following steps are involved in this method,convert the RGB image into grayscale,improve the image adjustment by using unsharp filter,apply a global threshold to obtain a binary image of cytoplasmic candidate and calculate the cytoplasmic features and cropping cytoplasm.The key challenges for affected cells,evolutionary biology,and precision medicine are the effect of drug,viscosity and the intensity of drug-treated cells.However,this is extremely difficult because of the enormous cells are affected by the drug.We developed a deep learning based framework for drug and normal cell image classification(DNCIC)that can accurately predict normal mitochondria and drug-affected cells that are rarely observed.For optimization,we used a convolutional neural network and trained it using a dataset of mitochondrial images,which were collected through the confocal microscope.The obtained algorithm was validated on the normal and affected cell images.We have trained CNN that can classify(normal and affected cells)and Two-Photon Excited Fluorescence Probes images(TPEF).The proposed model has classified images and videos with 98%accuracy.Our results provided a foundation for drug-affected cell diagnosis.Cell classification refers to detecting normal and diseased cells from mitochondrial images.Sometimes it is difficult to classify cells because some cells seem to fall into different categories/classes.Current state-of-the-art cell classification methods have been developed based on tumor cell classification but are not suitable for classifying diseased or normal cells.This study investigated the performance of two classification methods that are used to classify and differentiate broad categories of normal and diseased cells.Millions of normal cells undergo controlled growth and uncontrolled growth may be involved in disease causation but their clinical applications remain limited due to difficulties in distinguishing normal and diseased cells.Previously published studies are limited to systematically identify the normal and diseased cells.Our method consists of deep classification network normal and diseased cell classification(NDCC)for gathering information on diseased and normal cells,and verification network for accurate cell classification and removal of false positives.We used machine learning methods,including logistic regression(LR),support vector machine(SVM),and convolutional neural network(CNN).We found that CNN performed better for the normal and diseased cells.Using two types of images,normal and diseased cells,we trained a convolutional neural network that identified diseasthe clinical utility of human diseased cells.Mitochondria are highly dynamic cellular organelles,with the ability to change size,shape,and position over the course of a few seconds.Mitochondrial organelle movement refers to the problem of finding fission and fusion and generates energy for the cell.To address this problem,we proposed a deep learning method mitochondrial organelle movement classification(MOMC)for mitochondrial movement classification using a convolutional neural network.We present a three-step feature description strategies:1)local descriptions,which are first extracted via the GoogLeNet,followed by the production of mid-level features by ResNet-50,2)global descriptor features by Inception-V3 model,and 3)final classification of the position of mitochondrial organelle movement. |
PDF Full Text Request