Quantitative second harmonic generation imaging for differentiating between normal and cancerous thyroid tissue: Proc.SPIE

Thyroid Cancer (TC) accounts for 2.2% of all new cancer cases in the US and is the most common endocrine malignancy. Despite advancements in diagnostic methods, challenges such as high rates of indeterminate results, inadequate sampling, and false positives persist when using fine needle aspiration biopsy or core needle biopsy techniques. To address these limitations, Second Harmonic Generation (SHG) imaging, a label-free nonlinear optical imaging method that selectively uses fibrillar collagen as an endogenous contrast, offers a promising alternative for improved diagnostic accuracy. In this study, we applied SHG-specific feature extraction methods to generate quantitative metrics that reveal critical physical structural properties of imaged thyroid samples. Two key methods investigated: forward-backward ratio (F/B ratio) variance and mean Collagen Area Fraction (CAF), generated significant differences (⪅0.001) between thyroid cancer and normal thyroid samples, highlighting their independent capability to differentiate between cancerous and normal thyroid tissue. Notably, while F/B ratio used both forward and backward detected SHG images, the CAF metric showed statistically significant results exclusively with backward-detected SHG images. As SHG-based imaging evolves towards clinical applications, these quantitative metrics provide a robust foundation for the development of non-invasive diagnostic tools for thyroid cancer.