Wooden-Tip Electrospray Ionization Mass Spectrometry Combined With Machine Learning for Differentiating Thyroid...

Area of Science:

• Biomedical Engineering
• Analytical Chemistry
• Oncology

Background:

• Direct mass spectrometry (MS) of human tissues offers diagnostic potential but faces challenges with complex sample preparation.
• Rapid analysis is crucial for clinical applications like intraoperative tissue evaluation.

Purpose of the Study:

• To develop a rapid analytical strategy for differentiating thyroid tumors using ambient ionization MS.
• To integrate machine learning (ML) for classifying thyroid tissue samples.
• To identify potential lipid biomarkers for tumor diagnosis.

Main Methods:

• A disposable wooden tip (WT) was used for direct sample extraction and electrospray ionization (ESI) of thyroid tissues.
• Lipid profiles were obtained within minutes using WT-ESI-MS without extensive sample preparation.
• Machine learning models classified 45 thyroid tissue samples (healthy, benign, malignant).

Main Results:

• The WT-ESI-MS method enabled rapid lipid profiling of thyroid tissues.
• ML classification models successfully differentiated healthy, benign, and malignant thyroid tumors.
• Potential lipid biomarkers for tumor identification were discovered.

Conclusions:

• The integrated WT-ESI-MS and ML approach provides a fast and effective method for thyroid tumor differentiation.
• This technique shows significant potential for clinical diagnosis and intraoperative assessment.
• The method facilitates biomarker discovery for thyroid cancer.