Diagnosis of Breast Cancer Using Computational Intelligence Models and IoT Applications.

Area of Science:

• Medical Imaging
• Machine Learning in Oncology

Background:

• Computer-aided diagnostic (CAD) systems are increasingly utilized for breast cancer detection.
• Accurate classification of breast cancer nodules is crucial for effective treatment planning.

Purpose of the Study:

• To evaluate the performance of multilayer perceptron neural network and nonlinear support vector machine models for breast cancer nodule classification.
• To compare the efficacy of these machine learning algorithms in a diagnostic context.

Main Methods:

• Utilized ten morphological features extracted from the contours of 569 breast cancer samples.
• Trained and tested multilayer perceptron neural network and nonlinear support vector machine classifiers.
• Conducted 50 simulations to assess model performance and robustness.

Main Results:

• Both models demonstrated high accuracy, exceeding 90.0% on the test set.
• The nonlinear support vector machine achieved superior performance with 99% accuracy and a 2% false-negative rate.
• The neural network model exhibited lower performance compared to the nonlinear support vector machine.

Conclusions:

• The evaluated machine learning models show promising results for breast cancer classification.
• Nonlinear support vector machines offer a highly accurate and reliable approach for classifying breast cancer nodules.
• Further application of these CAD models can aid in improving breast cancer diagnosis.