Download PDFOpen PDF in browser

Feature Selection and Extraction for Cancer Detection: Different Approaches to Selecting and Extracting Relevant Features from Medical Imaging Data or Other Types of Data for Lung Cancer Detection

EasyChair Preprint 12789

14 pagesDate: March 27, 2024

Abstract

Cancer detection, particularly in the context of lung cancer, is a critical area of research and has significant implications for improving patient outcomes. Feature selection and extraction techniques play a crucial role in enhancing the performance of machine learning models for accurate cancer detection. This topic investigates various approaches employed to identify and extract relevant features from diverse data sources, including medical imaging data.

The primary objective of feature selection and extraction is to reduce the dimensionality of the data while retaining the most informative and discriminative features. Dimensionality reduction techniques, such as principal component analysis (PCA), linear discriminant analysis (LDA), and independent component analysis (ICA), are commonly utilized to transform high-dimensional data into a lower-dimensional representation. These techniques aim to preserve the essential characteristics and patterns of the data while eliminating redundant or correlated features.

Additionally, feature engineering techniques are employed to construct new features that provide enhanced discriminatory power for cancer detection. These techniques involve domain knowledge, statistical analysis, and mathematical transformations to derive meaningful and informative features. Feature engineering approaches can include intensity-based features, texture features, shape features, and wavelet-based features, among others. These techniques are designed to capture distinctive characteristics of cancerous tissues, such as irregularity, texture heterogeneity, and spatial distribution.

Keyphrases: Cancer Detection, LASSO (Least Absolute Shrinkage and Selection Operator), X-ray imaging

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@booklet{EasyChair:12789,
  author    = {Emmanuel Idowu and Lucas Doris},
  title     = {Feature Selection and Extraction for Cancer Detection: Different Approaches to Selecting and Extracting Relevant Features from Medical Imaging Data or Other Types of Data for Lung Cancer Detection},
  howpublished = {EasyChair Preprint 12789},
  year      = {EasyChair, 2024}}
Download PDFOpen PDF in browser