Optimization of machine learning classification models for tumor cells based on cell elements heterogeneity with...

Area of Science:

• Biomedical Spectroscopy
• Computational Biology
• Oncology

Background:

• Accurate and rapid cancer diagnosis is crucial in clinical medicine.
• Distinguishing between different tumor cell lines is essential for effective treatment strategies.

Purpose of the Study:

• To develop an innovative method for distinguishing and classifying tumor cell lines.
• To evaluate the performance of machine learning algorithms in conjunction with LIBS for cancer cell classification.

Main Methods:

• Acquisition of laser-induced breakdown spectroscopy (LIBS) spectra from cells.
• Spectral pre-processing and optimization to enhance classification accuracy.
• Comparison of convolutional neural network (CNN), support vector machine (SVM), and K-nearest neighbors algorithms.

Main Results:

• Both CNN and SVM algorithms achieved high discrimination performance, with an accuracy of 97.72% for tumor cell classification.
• Elemental heterogeneity within tumor cells was identified as a key factor for accurate cell distinction.
• LIBS demonstrated potential as a rapid classification method for tumor cells.

Conclusions:

• The combined LIBS and machine learning approach offers a powerful tool for rapid and accurate tumor cell classification.
• LIBS spectroscopy provides valuable insights into cellular elemental composition for diagnostic purposes.
• This method holds promise for advancing cancer diagnostics and personalized medicine.