The purpose of this study was to prospectively test the hypothesis that image overlay technology facilitates accurate navigation for magnetic resonance (MR)–guided osseous biopsy.
A prototype augmented reality image overlay system was used in conjunction with a clinical 1.5-T MR imaging system. Osseous biopsy of a total of 16 lesions was planned in 4 human cadavers with osseous metastases. A loadable module of 3D Slicer open-source medical image analysis and visualization software was developed and used for display of MR images, lesion identification, planning of virtual biopsy paths, and navigation of drill placement. The osseous drill biopsy was performed by maneuvering the drill along the displayed MR image containing the virtual biopsy path into the target. The drill placement and the final drill position were monitored by intermittent MR imaging. Outcome variables included successful drill placement, number of intermittent MR imaging control steps, target error, number of performed passes and tissue sampling, time requirements, and pathological analysis of the obtained osseous core specimens including adequacy of specimens, presence of tumor cells, and degree of necrosis.
A total of 16 osseous lesions were sampled with percutaneous osseous drill biopsy. Eight lesions were located in the osseous pelvis (8/16, 50%) and 8 (8/16, 50%) lesions were located in the thoracic and lumbar spine. Lesion size was 2.2 cm (1.1–3.5 cm). Four (2–8) MR imaging control steps were required. MR imaging demonstrated successful drill placement inside 16 of the 16 target lesions (100%). One needle pass was sufficient for accurate targeting of all lesions. One tissue sample was obtained in 8 of the 16 lesions (50%); 2, in 6 of the16 lesions (38%); and 3, in 2 of the 16 lesions (12%). The target error was 4.3 mm (0.8–6.8 mm). Length of time required for biopsy of a single lesion was 38 minutes (20–55 minutes). Specimens of 15 of the 16 lesions (94%) were sufficient for pathological evaluation. Of those 15 diagnostic specimens, 14 (93%) contained neoplastic cells, whereas 1 (7%) specimen demonstrated bone marrow without evidence of neoplastic cells. Of those 14 diagnostic specimens, 11 (79%) were diagnostic for carcinoma or adenocarcinoma, which was concordant with the primary neoplasm, whereas, in 3 of the 14 diagnostic specimens (21%), the neoplastic cells were indeterminate.
Image overlay technology provided accurate navigation for the MR-guided biopsy of osseous lesions of the spine and the pelvis in human cadavers at 1.5 T. The high technical and diagnostic yield supports further evaluation with clinical trials.
From the *Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine; †Department of Mechanical Engineering and Laboratory for Computational Sensing and Robotics, The Johns Hopkins University, Baltimore, MD; ‡Department of Mechanical and Materials Engineering, §School of Computing, Queen’s University, Kingston, Ontario, Canada; ∥Siemens Corporate Research; Center for Applied Medical Imaging; and ¶Department of Pathology, The Johns Hopkins University School ofMedicine, Baltimore, MD.
Received for publication June 5, 2012; and accepted for publication, after revision, October 10, 2012.
Conflicts of interest and sources of funding: Supported by grant 1 R01 CA118371-01A2 from the National Cancer Institute, Bethesda, Maryland.
P.U.T. received money for travel from Natural Sciences and Engineering Research Council of Canada.
A.J.F. has a portion of his retirement funding in Siemens stock.
I.I.I. received grant from NIH, NSF; institution has patent from Sentinelle Medical, Toronto (technology transfer for patent U.S. Provisional Application Serial No. 60/782,705 filed March 14, 2006, Publication No. WO/2007/106558, 09/20/2007; institution receives royalties from Gulmay Medical; institution received funding for research in cochlear electrode insertion from Cochlear Corporation.
J.A.C. is a board member of Vital, consultant to Quality Medical Metics and Medtronic, received money for multiple medical-legal activities; institution received grants from Siemens, Toshiba, Carestream, and Integra; author received payment for multiple grand rounds and invited lectures with honoraria; author has stock in Merge Healthcare.
Reprints: Jan Fritz, MD, The Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, 600 N Wolfe St, Baltimore, MD 21287. E-mail: firstname.lastname@example.org.