Acute lymphoblastic leukemia diagnosis using machine learning techniques based on selected features.

Area of Science:

• Medical Imaging
• Computational Biology
• Oncology

Background:

• Cancer is a leading cause of death globally, with early detection critical for survival.
• Computer-aided diagnosis (CAD) systems enhance cancer diagnosis accuracy using medical imaging.
• Acute lymphoblastic leukemia (ALL) requires timely and precise diagnosis for effective treatment.

Purpose of the Study:

• To develop a CAD system for the early and accurate diagnosis of ALL from microscopic blood images.
• To evaluate the performance of the developed CAD system in classifying ALL cases.

Main Methods:

• A four-phase CAD system was developed: preprocessing, segmentation, feature extraction/selection, and classification.
• Microscopic blood images were used to train and test the system.
• Naïve Bayes (NB), Support Vector Machine (SVM), and K-nearest Neighbor (K-NN) classifiers were employed with Ant Colony Optimization (ACO) for feature selection.

Main Results:

• The NB classifier, combined with ACO for feature selection, achieved the highest performance.
• The system demonstrated an accuracy of 96.15%, sensitivity of 97.56%, and specificity of 94.59% in classifying ALL.

Conclusions:

• The developed CAD system shows significant potential for accurate and efficient ALL diagnosis.
• This approach can aid in early cancer detection, potentially improving patient prognosis and treatment efficacy.