Development of an Algorithm for Cervical High-Grade Squamous Intraepithelial Lesion Based on Breath Print Analysis

Area of Science:

• Gynecologic Oncology
• Biomarker Discovery
• Medical Diagnostics

Background:

• Cervical high-grade squamous intraepithelial lesions (HSIL) require accurate and timely diagnosis.
• Current diagnostic methods may have limitations, necessitating novel approaches.
• Volatile organic compounds (VOCs) in breath may serve as potential biomarkers for cervical cancer precursors.

Purpose of the Study:

• To develop and evaluate an algorithm for diagnosing cervical HSIL using an electronic nose (e-nose).
• To assess the diagnostic performance of the e-nose based on VOC patterns.

Main Methods:

• A pilot study involving 25 patients with confirmed HSIL and 26 controls.
• Breath samples were analyzed using an e-nose to detect VOC patterns.
• Machine learning techniques were employed to build a predictive model for HSIL detection.

Main Results:

• The developed algorithm demonstrated a sensitivity of 0.88 and specificity of 0.92 for HSIL detection.
• Positive and negative predictive values were 0.92 and 0.89, respectively.
• A Cohen κ coefficient of 0.80 indicated substantial agreement.

Conclusions:

• The e-nose effectively distinguishes VOC patterns between HSIL patients and controls.
• The algorithm shows promise for non-invasive HSIL diagnosis.
• Further validation studies are required prior to clinical implementation.