Enhancing Breast Cancer Detection: A Machine Learning Approach Using Multielectrode Bioimpedance

Area of Science:

• Biomedical Engineering
• Computational Biology
• Oncology

Background:

• Early detection of cancerous tumors is crucial for improving treatment outcomes.
• Bioimpedance measurement of living tissues offers a simple, cost-efficient method for early cancer detection.
• Bioimpedance-based approaches show significant potential for identifying cancerous tumors.

Purpose of the Study:

• To explore a cost-efficient method for early cancer detection using bioimpedance.
• To develop and evaluate machine learning models for breast cancer detection based on bioimpedance data.
• To assess the performance of different machine learning algorithms in identifying cancerous breast tissue.

Main Methods:

• Simulated breast impedance using the Cole-Cole model and finite element modeling for healthy and tumor tissues.
• Collected bioimpedance data using an eight-electrode system around breast models of varying sizes and tumor conditions.
• Developed and trained machine learning models, including Support Vector Machines (SVM), Convolutional Neural Networks (CNN), and Random Forest (RF), on the prepared dataset.

Main Results:

• Demonstrated the feasibility of integrating machine learning with multielectrode bioimpedance for precise, automated breast cancer detection.
• The Random Forest (RF) model exhibited superior accuracy in detecting cancerous tumors compared to SVM and CNN.
• The study successfully utilized simulated and measured bioimpedance data to train effective cancer detection models.

Conclusions:

• Bioimpedance methods combined with machine learning algorithms hold significant potential for early cancer detection.
• Random Forest models show promise for accurate and automated breast cancer detection.
• This approach offers a valuable tool for improving patient outcomes through early diagnosis.