Eine prospektive klinische Machbarkeitsstudie zur Bewertung der In-vivo-Visualisierung von Kopf-Hals-Tumoren während der Operation mittels hyperspektraler Bildgebung

Introduction:

Successful surgical treatment of head and neck cancer (HNC) is highly dependent on complete tumor resection. Currently, resection of malignant tumors of the head and neck region depends primarily on the expertise of the surgeon and postoperative histological report, which can occasionally lead to incomplete resection. Hyperspectral imaging (HSI) has proven to be a promising tool to identify accurately tumor margins.

Methods:

The ongoing HSI-Kopf-Hals trial (DRKS00033009) assesses the feasibility of HSI for intraoperative identification of tumor boundaries compared to histopathology as the gold standard. First, a workflow was established not disturbing the standard tumor surgery. For this study, the surgeon used an endoscopic CE-certified HSI device to acquire images of the tumor areas in 15 patients prior to resection. These HSI images were then processed through an automated pipeline and validated by comparison with annotations from an independent expert surgeon.

Results:

Our workflow requires approximately 2-3 minutes from positioning the endoscope to acquiring the images, with each HSI acquisition taking 6 seconds. The automated analysis pipeline achieved a sensitivity of 53% and a specificity of 86% compared to the expert surgeon's annotations. This HSI-based workflow for detecting tumor margins during surgery is easy to use and quick to execute, which can enhance the precision of tumor resection and improve patient outcomes in cancer treatment.

Conclusions:

HSI performed in compliance with stringent local and global medical regulatory standards seem to be able to differentiate between healthy and tumor issues in vivo.

nein