What question did this study set out to answer?

The aim is to identify radiomic features in CT images and assess the efficacy of various machine learning algorithms.

March 14, 2026Open Access

Radiomics for the Classification of Radiological Patterns in Computed Tomography

Key Points

The aim is to identify radiomic features in CT images and assess the efficacy of various machine learning algorithms.
Analyzed radiomic features from CT scans
Evaluated performance of different machine learning algorithms including KNN, SVM, RF, and MLP
Calculated specificity and sensitivity of each model
SVM achieved the highest performance with mean specificity of 96.25%
Mean sensitivity for SVM was 85.05%
All algorithms showed satisfactory performance with low standard deviation

Abstract

Pulmonary diseases are a major issue today, being one of the leading causes of death worldwide. Because of this, it is essential to develop methods to improve their diagnosis. This study arises from this problem, with the aim of identifying radiomic features in Computed Tomography (CT) and to evaluate the performance of different machine learning algorithms, including KNN, SVM, RF, and MLP. The results obtained show satisfactory values, with specificity standing out. SVM had the best performance with, with a mean specificity of 96.25% and a mean sensitivity of 85.05% both with a small standard deviation.

Demander à l'IA

Bookmark

View Full Paper