Optical imaging modalities for the detection of head and neck cancer margins

Abstract

Positive surgical margins in head and neck cancers (HNC) are associated with poor survival. Standard frozen section analysis for intraoperative assessment of margins is limited by processing time and sampling error. Optical imaging technologies may address these limitations. We identified the following techniques in a literature search of optical imaging modalities for the detection of head and neck tumor margins: autofluorescence imaging, dynamic optical contrast imaging, optical coherence tomography, narrow band imaging, hyperspectral imaging, Raman spectroscopy, near-infrared fluorescence imaging, confocal laser endomicroscopy, and high-resolution microendoscopy. Penetration depths range from surface level to 6 mm, image acquisition times range from real-time to several minutes, and 3/9 require exogenous contrast agents. Reported sensitivity and specificity range from 71-100% and 43-100%, respectively. Each reviewed modality lends unique strengths such as fast image acquisition times, wide field of view, high native contrast, or seamless integration with existing endoscopes. However, none have yet to replace palpation and frozen section analysis in the operating room.