A study to test whether images taken by a new handheld medical imaging device can be used to detect the microscopic presence of cancer during breast cancer surgery.

Handheld MEprobe scans of the surgical breast cavity will be performed to determine whether residual tumour can be visualised in the breast cavity of patients with positive margins, as determined by post-operative histological assessment of the corresponding excised cavity shavings. The MEprobe will be located in the operating theatre at Fiona Stanley Hospital (FSH). Two engineers will be present during each scan, one to operate the imaging scanner and the other to guide and assist the breast surgeon to use the handheld MEprobe. TGA-approved sterile sheaths routinely used in surgery will be used to cover the sterilized MEprobe and associated cables to ensure the scanning is performed in a sterile environment. After completing the wide local excision of breast tissue, the surgeon will use their clinical judgement (which may include intra-operative imaging of the specimen), as is standard practice, to determine which area(s) of the resection bed requires additional removal of breast tissue (i.e. cavity shavings). On suitable areas, the surgeon will insert the tip of the MEprobe, clad in its sterile sheath, into the wound and scan the cavity before excising the cavity shaving. All specimens will be orientated with suture markings and radio-opaque clips/markers as per standard operating procedures before being placed in a labelled specimen container for pathology. This procedure will be repeated until all clinically relevant cavity shavings have been excised. Multiple images may be acquired from each patient and each image will be 7×7 mm in size and take less than 5 seconds to acquire. The wide local excision specimen and all cavity shavings will be transported by the research team to PathWest at FSH for validation purposes. Ex vivo micro-elastograms of excised cavity shavings +/- wide local excision specimens transported from surgery to PathWest will be acquired using the micro-elastography benchtop system previously validated by the research team. Following ex vivo scanning, all excised tissue will be fixed, embedded in paraffin, and hematoxylin and eosin (H&E) stained. Histological sections will be prepared, as per standard practice for all excised breast tissue. Initially, one surgeon (the CPI) will operate and test the new probe on up to 10 study participants. Detailed feedback from the surgeon will be used to generate operator training guidelines and a protocol for study use during in vivo scanning of the resection bed during breast-conserving surgery. Up to four additional surgeons operating at FSH will then be trained in the use of the probe. They will provide feedback to refine the operator training guidelines so that the technique can be standardised for the accuracy arm of the study which aims to determine the accuracy (including sensitivity and specificity) of the MEprobe for use in classifying in vivo breast tissue as either benign or cancerous. To determine the accuracy of the MEprobe a blinded reader study will be done. Blinded readers, including surgeons, radiologists, engineers and pathologists, will be trained to read breast micro-elastograms using a training data set. The training data set will include histology-validated micro-elastograms from a previous benchtop micro-elastography study of ex vivo breast tissue in addition to MEprobe study images acquired before recruitment into the accuracy arm of the study. The training data set findings will establish the evaluation methodology and image feature criteria for identifying cancer. As part of the training procedure, readers will be taught how to use ImageJ, a widely-used software that is needed to visualise the 3-dimensional micro-elastograms.