A decision support system to recommend appropriate therapy protocol for AML patients.

Area of Science:

• Hematology
• Oncology
• Medical Informatics

Background:

• Acute Myeloid Leukemia (AML) is an aggressive cancer requiring precise treatment planning.
• Current risk stratification for AML has challenges, especially in the intermediate-risk group, potentially delaying care.
• Accurate prognostic predictions are crucial for effective AML therapy decisions and patient outcomes.

Purpose of the Study:

• To develop and validate a machine learning-based decision support system for personalized AML treatment.
• To improve the accuracy of outcome prediction for AML patients, particularly those in the intermediate-risk category.

Main Methods:

• Implementation of a decision support system utilizing advanced machine learning algorithms.
• Automatic recommendation of tailored oncology therapy protocols based on predicted patient outcomes.
• Utilizing gene expression data for model generation to enhance predictive performance.

Main Results:

• The developed system achieved high performance metrics, with an F1-Score and AUC close to 0.9.
• Models trained on gene expression data demonstrated superior predictive accuracy.
• The system effectively supports specialists in selecting optimal and safe AML therapies.

Conclusions:

• The proposed decision support system streamlines treatment initiation for AML patients.
• This approach has the potential to enhance patient survival and quality of life.
• The system represents a significant advancement in optimizing therapeutic interventions for AML.