Development of clinical hyperspectral imaging system for head and neck tumors

Early-stage cancer detection and classification significantly increase the chances of successful treatment. In this study, we aim to develop a portable clinical hyperspectral imaging (HSI) system for the detection and classification of skin tumors in the head and neck region.

The tripod mounted system is based on a handheld hyperspectral camera (IQ, Specim, Finland) with an additional achromatic doublet lens to achieve sub-millimeter spatial resolution. The imaging regions were illuminated with a halogen lamp coupled to a ring illuminator providing bright and homogenous illumination. Tissue topography was also captured using two time-of-flight cameras (flexx2, PMD, Germany).

Three types of tumors were imaged in the study: benign papillary nevi, and malignant basal and squamous cell carcinomas. Hyperspectral images were analyzed using tissue indices, a metric that estimates various tissue properties, including melanin content, blood content, and oxygenation. Preliminary results show significant differences between tissues: malignant tumors have less melanin than benign ones (p<0.05) and tumors differ from healthy tissue in oxygenation (p<0.02).

Results show that a portable clinical HSI system could assist in early tumor detection, leading to better treatment outcome.