Prediction time of breast cancer tumor recurrence using Machine Learning

Area of Science:

• Oncology
• Medical Informatics
• Biostatistics

Background:

• Breast cancer is a leading cause of mortality globally.
• Early detection and prediction of recurrence are crucial for patient outcomes.
• Existing models predict tumor nature or recurrence but lack time-based predictions.

Purpose of the Study:

• To develop and evaluate machine learning models for predicting breast cancer tumor recurrence time.
• To provide patients with accurate estimates of recovery time or the need for timely medical consultation.

Main Methods:

• Utilized a database from GLOBOCAN, CDC, and WHO.
• Applied machine learning classification models: Spectral clustering, DBSCAN, k-means.
• Employed prediction models: Support Vector Machines (SVM), Decision Trees, Random Forest.
• Analyzed data from 198 patients.

Main Results:

• Machine learning models accurately predict tumor recurrence time, with forecasts up to 1 year.
• 40% of analyzed patients were predicted to have recurrence within the first year.
• Support Vector Machines (SVM) achieved the highest prediction accuracy of 78.7%.

Conclusions:

• Machine learning models can effectively predict breast cancer recurrence time.
• The study provides a valuable tool for estimating patient recovery and guiding medical consultations.
• This population-based study offers a reasonable time-based prognosis for breast cancer patients.