Abstract
Interventional MR (IMR) machines have produced unique opportunities for image‐guided surgery. The open configuration design and fast pulse sequences allow intraoperative scanning to monitor procedures. This study was undertaken to assess the potential use of IMR for image‐guided surgery. Benign breast lesion excision was chosen as an uncomplicated surgical model. Ten female patients with known benign tumors underwent excision biopsy under general anesthesia in a Signa SP10 .5‐T IMR machine (General Electric Medical Systems, Milwaukee, WI). Lesions were localized with precontrast and postcontrast (intravenous gadolinium‐diethylenetriamine pentaacetic acid, .2 mmol/kg) fast multiplanar spoiled gradient‐recalled acquisition in the steady state (GRASS) sequences. Preoperative “real‐time” fast gradient‐recalled sequences were also obtained using the Flashpoint (General Electric Medical Systems, Milwaukee, WI) tracker device. The maximum dimensions of each lesion were measured from the resulting images. Excision was performed using titanium instruments and an ultrasonically activated scalpel. Intraoperative real‐time scanning demonstrated the resection margin and confirmed complete excision. The maximum dimensions of the macroscopic specimens were compared with those from the MR images. All tumors were visualized with the Signa scanner and real‐time imaging and the images were enhanced after intravenous contrast. Maximum dimensions on histologic examination were not significantly different from those measured from Signa (P > .17) or real‐time images (P > .4). There was no significant difference between lesion size from Signa and real‐time images (P > .25). All postprocedure scans demonstrated complete excision. There were six fibroadenomas, two foci of sclerosing adenosis, one area of fibrocystic disease, and one schwannoma. Intraoperative MR scanning reliably identifies palpable breast tumors and can accurately guide surgical excision. Further work using MR guidance can now be performed in other general surgical areas.