Remote Controlled Manipulator in combination with Automated Real-time Needle-guide Tracking for fast MR-guided Transrectal Prostate Biopsy

Background

Various strategies have been developed to make in-bore MR-guided biopsies less time-consuming, e.g. robot-assistance and the use of an automatic, real-time needle-guide tracking sequence. The use of a real-time tracking sequence shows a promising reduction in procedure time. However, reaching into the magnet to hold the needle-guide during tracking can be uncomfortable or even impractical for the physician. Robot-assistance might solve this issue.

Purpose

To assess the feasibility of a remote-controlled manipulator (RCM) in combination with an automatic real-time needle-guide tracking sequence for fast in-bore magnetic resonance (MR)–guided transrectal prostate biopsies.

Methods

A phantom study was conducted on a 3-T clinical MRI system (Magnetom Skyra, Siemens, Erlangen, Germany). Three MR-guided biopsy settings were tested: I. manual procedure, II. robot-assisted procedure and III. robot-assisted procedure combined with real-time needle-guide tracking (see figure below).


POCCfantoom.jpg


In the manual procedure an adjustable MR-compatible biopsy device (DynaTrim; Invivo, Schwerin, Germany) was used to direct the needle-guide at a biopsy target. In the robot-assisted procedure an MR-compatible RCM (Soteria Medical BV, Arnhem, the Netherlands) was used, which could be directed at a biopsy target using integrated planning software. For the third procedure the RCM is combined with a phase-only cross correlation (POCC) algorithm–based automatic needle-guide tracking sequence to facilitate targeting under real-time image feedback (see image below).


Pocc.jpg


In each setting, 25 fiducials embedded in agar gel were targeted. Correct targeting was confirmed with short balanced steady-state free precession (bSSFP) imaging. Total procedure times and targeting accuracy were determined. The procedure time was defined as the time from acquisition of the first targeting image to the confirmation image with the biopsy needle in situ. Targeting accuracy was determined by measuring the targeting error, biopsy error and target displacement.

Results

Total procedure time with the manual procedure, robot-assisted procedure and robot-assisted procedure combined with real-time tracking, were respectively 10:49±1:45 min, 6:33±0:46 min and 5:47±0:54 min. All fiducials were successfully targeted and hit on subsequent biopsy. The biopsy error of the manual, robot-assisted and robot-assisted procedure combined with a real-time tracking, were respectively 1.2±0.6 mm, 1.4±0.9mm, and 1.7±0.8 mm.

Conclusion

This study demonstrated the feasibility of MR-guided biopsy by using an RCM in combination with a real-time tracking sequence. Combining robot-assistance with real-time tracking resulted in 46% reduction in procedure time per target compared to the manual procedure.

People