PURPOSE: To develop an MR-compatible vacuum-assisted core needle breast biopsy system. MATERIALS AND METHODS: A vacuum-assisted breast biopsy system (Mammotome Hand Held; Ethicon Endo-Surgery, USA) was modified for freehand MRI-guided biopsy in an open, interventional 0.5-T scanner (Signa SP; GE, USA). Probes (11 gauge [G]) were fabricated without significant susceptibility artifact. These mate with an electromechanical hand piece and control system that were modified for use within the MRI scanner. A total of 62 breast lesions were simulated in the mammary tissues of six recently lactating sows by injecting between 0.1 and 1.0 mL of an aqueous gel containing dilute gadopentetate dimeglumine (Gd-DTPA) that formed a bright mass on T1-weighted imaging. RESULTS: Mechanical performance was satisfactory. Magnetic susceptibility and radiofrequency (RF) artifacts from the 11-G probe were negligible. T1-weighted fast spin echo (FSE) was used to guide biopsy. Up to eight samples were removed per lesion. Overall, 461 samples were obtained in 493 attempts (94%). Sample weights averaged 54 mg (N = 493) compared to 4.6 mg (N = 24) from 14-G titanium MRI-compatible needles. On average, 59% of the attempted samples yielded target lesion material. CONCLUSION: Preliminary results demonstrate the feasibility of a vacuum-assisted breast biopsy system in the MRI environment. Small 0.1-mL lesions can be biopsied without needle artifacts obscuring the target.
PURPOSE: To develop an MR-compatible vacuum-assisted core needle breast biopsy system. MATERIALS AND METHODS: A vacuum-assisted breast biopsy system (Mammotome Hand Held; Ethicon Endo-Surgery, USA) was modified for freehand MRI-guided biopsy in an open, interventional 0.5-T scanner (Signa SP; GE, USA). Probes (11 gauge [G]) were fabricated without significant susceptibility artifact. These mate with an electromechanical hand piece and control system that were modified for use within the MRI scanner. A total of 62 breast lesions were simulated in the mammary tissues of six recently lactating sows by injecting between 0.1 and 1.0 mL of an aqueous gel containing dilute gadopentetate dimeglumine (Gd-DTPA) that formed a bright mass on T1-weighted imaging. RESULTS: Mechanical performance was satisfactory. Magnetic susceptibility and radiofrequency (RF) artifacts from the 11-G probe were negligible. T1-weighted fast spin echo (FSE) was used to guide biopsy. Up to eight samples were removed per lesion. Overall, 461 samples were obtained in 493 attempts (94%). Sample weights averaged 54 mg (N = 493) compared to 4.6 mg (N = 24) from 14-G titanium MRI-compatible needles. On average, 59% of the attempted samples yielded target lesion material. CONCLUSION: Preliminary results demonstrate the feasibility of a vacuum-assisted breast biopsy system in the MRI environment. Small 0.1-mL lesions can be biopsied without needle artifacts obscuring the target.