295 - Toward Augmented Reality-Driven MR-Guided Interstitial Brachytherapy

E. S. Paulson, T. J. Gildemeister, J. J. Rownd, and B. A. Erickson; Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, WI

Purpose/Objective(s): High dose rate interstitial brachytherapy (IB) is an effective treatment regimen for GYN malignancies and has been shown to improve survival. However, providing real-time information to physicians performing transperineal, free-hand or applicator-based catheter insertions can be challenging. Augmented reality (AR) is a non-immersive technology that superimposes digital information onto the physical environment perceived by a user. We demonstrate here the feasibility of utilizing AR-driven MR guidance for transperineal catheter insertion during IB for cervix cancer.

Materials/

Methods: An industrial AR headset was modified for use in a 3T interventional MRI suite. MRI safety tests were performed to evaluate projectile motion, heating, torque, and RF interference induced by the AR headset. A wearable wireless communication and power pack was developed to receive real-time information in the AR headset while also enabling physicians to roam out of the magnet room for surgical scrubbing. Real-time, interleaved 2D GRE sagittal/coronal images were streamed to a raspberry pi placed on the local MRI network. A python image viewer was developed to display real-time images in the AR headset. The accuracy and latency of the overall system was evaluated in a biopsy phantom. A dedicated GYN phantom was used to practice using AR+MR guidance to implant needles for IB. Finally, AR+MR navigation was used during transperineal catheter insertion in four GYN cancer patients under an IRB protocol (PRO00036793). Anesthetized patients were positioned foot-first supine in the MRI. An MR-compatible leg riser was developed to immobilize lower extremities. Custom 3D printed brachytherapy applicators were utilized during the procedures. Catheter utilization was planned online using pre-insertion images. Fast template localizers, acquired prior to each insertion, were used to account for template movement that occurred during prior insertions when planning real-time imaging planes. Finally, T2 SPACE and PETRA images were acquired for treatment planning.

Results: Projectile force induced by the AR headset was <0.3 Nm at the bore entrance, heating was <1 deg F, no RF interference or torque were observed. Battery life for the power pack was >2 hours. Overall system latency was about 1 second. No signal loss occurred when operating the AR headset in the MRI bore during imaging. Using AR+MR, needles were able to be guided to each target consistently in the biopsy phantom. The foot-first patient orientation provided increased physician access at the rear of the MRI bore. Updated receive coil sensitivity profiles were required prior to each catheter insertion to account for physician hands during image reconstruction. All needles were successfully implanted in the four GYN cancer patients.

Conclusion: AR-driven, MR-guided IB is feasible. Interstitial catheters can be placed successfully prior to leaving the procedure room to assure an acceptable implant with little need for post implant modifications.