A co-registration method to validate in vivo optical coherence tomography in the breast surgical cavity

Breast-conserving surgery accompanied by adjuvant radiotherapy is the standard of care for patients with early-stage breast cancer. However, re-excision is reported in 20-30 % of cases, largely because of close or involved tumor margins in the specimen. Several intraoperative tumor margin assessment techniques have been proposed to overcome this issue, however, none have been widely adopted. Furthermore, tumor margin assessment of the excised specimen provides only an indirect indication of residual cancer in the patient following excision of the primary tumor. Handheld optical coherence tomography (OCT) probes and their functional extensions have the potential to detect residual cancer <i>in vivo</i> in the surgical cavity. Until now, validation of <i>in vivo</i> OCT has been achieved through correlation with <i>ex vivo</i> histology performed on the specimen removed during surgery that is adjacent to the tissue scanned <i>in vivo</i>. However, this indirect approach cannot accurately validate <i>in vivo</i> imaging performance. To address this, we present a method for robust co-registration of <i>in vivo</i> OCT scans with histology performed, not on the main specimen, but on cavity shavings corresponding directly to the tissue scanned <i>in vivo</i>. In this approach, we use <i>ex vivo</i> OCT scans as an intermediary, surgical sutures as fiducial markers, and extend the <i>in vivo</i> field-of-view to 15 × 15 mm² by acquiring partially overlapping scans. We achieved successful co-registration of 78 % of 139 <i>in vivo</i> OCT scans from 16 patients. We present a detailed analysis of three cases, including a case where a functional extension of OCT, quantitative micro-elastography, was performed.