Dual-modal photoacoustic and ultrasound imaging for smarter biopsy decisions in thyroid nodules

Thyroid nodule diagnosis relies heavily on ultrasound (US) screening, but conventional imaging exhibits poor specificity, resulting in excessive fine-needle aspiration biopsies. This study presents a dual-modality approach integrating photoacoustic (PA) and US technologies to enhance biopsy decisions. We examined 106 thyroid nodules across three pathological categories: benign nodules (n=29), papillary thyroid carcinomas (n=45), and follicular neoplasms (n=32). Three PA biomarkers—spectral gradient, oxygen saturation, and skewness of oxygen saturation distribution asymmetry—were combined via support vector machine. Integration with existing American Thyroid Association protocols yielded the ATA-photoacoustic scoring framework. At clinically optimized thresholds, this combined approach detected 97% of fine needle aspiration biopsy-requiring nodules while correctly reclassifying 38% of benign lesions as low-risk, demonstrating substantial reduction in unnecessary invasive procedures without compromising diagnostic accuracy.