Hyperspectral based discrimination of thyroid and parathyroid during surgery

Abstract

Unintended injuring of anatomical structures during endocrine neck operations can have severe consequences for patient. Especially the nerves and the parathyroid gland can be hard to identify visually. Therefore, intraoperative methods are needed to support the surgeon in this task. Hyperspectral imaging (HSI) is a new approach in the medical area which combines a camera with a spectrometer. It showed promising results for the discrimination of tissue. In this work, HSI-data of seven patients were acquired during thyroid and parathyroid operations. The mean absorbance spectra of both glands showed differences in the range between 600 and 700 nm and at 760 and 960 nm. This means that thyroid and parathyroid have different oxygenation states and different contents of deoxygenated hemoglobin and water. From these observations, it is possible to define spectral signatures to characterize both glands. We showed on one patient how spectral signatures can be used in classification algorithms to automatically identify the thyroid and parathyroid from other structures.

Keywords

• intraoperative imaging
• gastrointestinal operation
• patient data
• spectral signature
• classification algorithms