Spectra-BREAST: Combining Hyperspectral Imaging and Raman Spectroscopy with AI for the Intraoperative Assessment of Breast Tumor Resection Margins

Introduction/Objectives:

Breast-conserving surgery is currently limited by the lack of a fast-intraoperative margin assessment method able to provide surgeon with a near real time feedback on the complete removal of the cancerous lesion. Spectra-BREAST ( https://spectrabreast.eu/ ) is an EIC-Pathfinder funded project aimed at delivering a semi-automatic assessment of tumour resection margins during surgery in less than 5 minutes by a combination of optical spectroscopies.

Materials & Methods:

The project concept is the integration of a fast, hyperspectral imaging (HSI) camera to rapidly scan the entire resected specimen coupled with a fibre-optic Raman spectroscopy (RS) analysis on area at risk identified by HSI. A hybrid AI pipelines merges HSI information with specific RS features to provide a objective assessment of the presence of remaining cancer cells on the freshly excised tissue specimen.

Results/Conclusions:

Spectra-BREAST defines an integrated method for the near-real-time optical spectroscopy-based margin assessment, expected to reduce re-operation rates, improve patient experience, and lower healthcare costs. The approach here developed is designed to be transferable to other solid tumours and to ex vivo biopsy assessment, supporting wider clinical adoption of spectroscopy-enabled decision in oncological surgery.