Permanent interstitial brachytherapy is a widely used treatment for localized prostate cancer, in which radioactive seeds are implanted directly into the prostate through needles inserted via the perineum. The procedure relies on a template grid with holes spaced 5 mm apart to guide needle placement, but this fixed spacing limits how precisely seeds can be positioned.
A major challenge during the procedure is tissue deflection -- as a needle travels through tissue, it can veer from its intended path, causing seeds to land in the wrong location. This misplacement can alter the radiation dose distribution, potentially under-treating cancer or over-exposing healthy tissue.
Surgeons currently address deflection by manually re-steering or bending the needle, or by using deflecting devices at the template entry point. These workarounds require additional needle insertions, increasing tissue trauma and worsening side effects such as urinary symptoms.
Another problem is pubic arch interference, where the pelvic bone blocks needle access to certain parts of the prostate. Standard repositioning techniques like extended lithotomy positioning do not work for all patients and may increase the risk of rectal injury.
Researchers at Johns Hopkins University developed a compact robotic needle guide that mounts directly onto a standard ultrasound stepper device, replacing the conventional template. The device consists of two motorized stages -- an xy translational stage and an angulation stage -- allowing the needle guide tube to be repositioned or angled precisely.
The robot uses computer-controlled motors with integrated encoders and high-resolution linear encoders with 10-micrometer precision. The needle guide tube is an aluminum tube held by ball-and-socket joints, accommodating a standard 18-gauge implant needle. Movement in the x and y directions of up to 40 mm is possible across the perineum.
A graphical user interface running on the same laptop as the dosimetry planning software allowed the operator to direct robot movements. The system was calibrated to align with the ultrasound coordinate frame using a standard water bath procedure outlined in the AAPM Task Group 128 guidelines.
For the clinical pilot study, approved by the institutional review board, translational movement only was tested -- the angulation capability was reserved for future studies. A Polaris infrared tracking system was used during the first three patients to independently verify robotic positioning accuracy before each needle insertion, serving as a safety measure.
Five patients with localized prostate cancer were enrolled in this prospective pilot study. Each patient was anesthetized and placed in the lithotomy position, and transrectal ultrasound images were acquired at 5 mm intervals to create a treatment plan. Organs were contoured and a dosimetric plan was generated by the physician.
During the implant, the Interplant dosimetry software was modified so that once the physician identified which needle to place, the robot would automatically move to the target coordinates. The physician then inserted the needle manually through the robotic guide, confirming position on real-time ultrasound.
If the needle tip on ultrasound was 3 mm or more off the planned location, or if anatomy had changed due to edema, the physicist could adjust the robot position using the GUI. If a systematic directional error was noticed early in the procedure, a global calibration offset was applied to all subsequent needle positions.
Real-time intraoperative dosimetry was used throughout to estimate dose coverage from needle trajectories visible on ultrasound, allowing the team to make last-minute adjustments to seed placement if dose gaps were identified.
The robotic system functioned successfully in all five patients. The Polaris tracking system confirmed that robotic movements were highly accurate, with a mean positioning error of 0.49 mm (standard deviation 0.33 mm) -- well within clinically acceptable limits. After the first three cases showed no significant errors, use of the Polaris tracker was discontinued.
Despite precise robot movements, tissue deflection of needles was still observed. In total, 54 out of 179 needles (30.2%) required positional adjustment after the robot had placed the guide, and 36 of those adjustments (20.1%) were greater than 2 mm. A systematic directional correction was applied in three of the five patients.
Statistical analysis found no significant correlation between needle depth and the magnitude of deflection (p = 0.353), and only about 4% of the variability in deflection was explained by differences between patients, suggesting needle-to-needle variation was the dominant source of unpredictability.
Twenty-seven needle insertions were deliberately placed at positions between the standard 5 mm template grid points -- something impossible with a fixed template. This capability was used to improve dose coverage, avoid blood vessels, navigate around pubic arch interference, and prevent seeds from ending up too close to the rectal wall.
The duration of the robotic implant procedures ranged from 104 to 159 minutes, with a median of 133 minutes. This was not statistically different from a comparison group of 21 consecutive patients treated with the standard template during the same period (range 70-191 minutes, median 121 minutes; p = 0.58).
Two minor technical issues were encountered. In the first patient, the needle guide tube was too long, preventing needle reach to the prostate base; the robot had to be temporarily swapped for a standard template mid-procedure. The tube was subsequently shortened by 2 cm for all remaining cases, resolving the problem.
The robot arms also had the potential to collide with the ultrasound probe when moving laterally across the posterior grid. Once the order of needle placement was modified to avoid this path, no further collisions occurred. These were described as minor issues that did not compromise clinical outcomes.
This study represents the first reported clinical trial of a robotic needle positioner for prostate brachytherapy. Prior work had described pre-clinical robotic systems, but none had been tested on human patients. The system demonstrated that robotic guidance could correct tissue deflection, enable off-grid needle placement, and help avoid critical structures.
One practical advantage was that the robot and the standard template were interchangeable -- both mount in the same location and are calibrated to the same coordinate frame. This allowed seamless switching mid-procedure if needed, without major disruption, providing a safety fallback for the surgeon.
Ergonomic challenges were noted. The reliance on verbal communication between physician and physicist to direct robot movements was described as cumbersome. The absence of direct physician control also introduced the risk of collision between the needle guide and the ultrasound probe, since the robot moved to preset X and Y coordinates rather than under live guidance.
The study was not designed to measure dosimetric outcomes or toxicity, so conclusions about whether robotic guidance improves treatment quality or reduces side effects await future trials with larger patient populations. The angulation capability of the device, which could help bypass the pubic arch, was also not tested in this pilot.
The authors plan to add an interface for direct physician control of robot movement and software collision-avoidance algorithms to prevent the needle guide from striking the ultrasound probe. An optical encoder may also be incorporated to monitor needle insertion depth automatically.
Future studies will test the angulation capability of the robot, which could allow needles to be aimed around the pubic arch or urethra without requiring patient repositioning or hormone therapy to shrink the prostate. This feature could expand the eligibility of patients for brachytherapy.
Beyond correcting deflections, robotic systems may enable continuum spacing with computerized inverse optimization, potentially creating more conformal dose distributions and reducing the total number of needles required for a given target volume. These theoretical advantages require dedicated clinical validation.
Overall, this pilot study establishes proof-of-concept that robotic needle guidance is feasible and safe in the clinical setting, opening the door to larger trials that could redefine standards of precision in prostate brachytherapy.