Please enable JavaScript.
Coggle requires JavaScript to display documents.
Lung Cancer, Note: lacks further explanations on why certain methods were…
Lung Cancer
Leading cause of millions of deaths
Survivability is high if detected earlier
Difficult to diagnose early
Use of CT images for diagnosis
Non-obstructive image extraction
Current methods detect lung cancer nodules at later stages
There is a need to detect it early
Contributions
First phase
Selection of features using
Linear Discriminative Analysis
Second Phase
Classification using
Modified Gravitational Search Algorithm
Classification Methods
Deep Belief Networks
For parameter learning
Restricted Boltzmann Machine
Helps in learning parameters
Stochastic method
Modified Gravitational Search Algorithm
For weight optimization
Weight initialization
Force evaluation
Fitness evaluation
Use of maximum specificity
Selection of random value
Mass and force updates
Updates values based on fit, best, and worst values
Optimal solution and termination
Fine-tuning
Use of backpropagation
Enhancement and Extraction Methods
Filtering and Enhancement
Used for noise removal
Prevalent in medical images
Use of filters if neighboring pixels are near 0 or 255
Use of histogram equalization to improve contrast
Feature Extraction
Texture
Spatial relationship of gray level pixels
Outputs 22 features
Homogeneity
Determines values of minor constrasts
Energy
Maximum values as maximum shape
Entropy
Quantity of data
For compression
Contrast
Variance of pixels based on neighbors
Correlations
Linear dependence of gray levels
Wavelet-based
Transformation to frequencies through linear transformation
Histograms
Plotting through histograms and determine gray levels
Variance
Determines number of gray level fluctuations
Mean
Average gray level in regions
Standard Deviation
Skewness
Based on tails
Negative
Positive
Kurtosis
For shape distribution and determines anomalies in images
Feature Reduction
LDA
Combines bad features into one
Results and Discussion
High accuracy and Specificity
LDA is a better than PCA and ICA
Note: lacks further explanations on why certain methods were used
Note: Comparison of performance should be in tabular form than bar charts
S.K., L., Mohanty, S. N., K., S., N., A., and Ramirez, G. Optimal
deep learning model for classification of lung cancer on ct images. Future
Generation Computer Systems 92 (2019), 374–382.