In this system, reconstruction of the 3D prostate model requires 200 degree rotation of the 2D end-fire TRUS probe across the surface of the prostate by the tracking robotic arm. Using the volume data of the acquired TRUS images, the computer-software reconstructs the 3D prostate model, after the procedure for segmentation of the prostate. The actual firing of the biopsy needle is monitored and captured by a real-time 2D image, and then, each biopsy trajectory in the real-time 2D image can be registered and overlaid on the initially reconstructed 3D prostate model. Importantly, patient immobility, prostate position, and prostate deformation after the initial reconstruction of the 3D prostate model are critical for the accurate registration of each biopsy trajectory on the 3D model, because of the potential dislocation between the initially reconstructed 3D prostate model and the actual location of the prostate at each biopsy. Once the biopsy trajectories have been recorded in the 3D prostate model, the digital 3D model with the registered biopsy trajectories can be used in the future, after a certain time interval, for revisiting previously recorded sites for repeat biopsy options. In this process, after a certain time interval, the system allows overlaying the previous biopsy sites on the current 3D prostate model, and then, performing resampling from the same site as the previous positive biopsy site, or sampling from different sites from previous negative biopsy sites. In order to achieve precise revisiting using the previous digital 3D model, the challenge is to register precisely the current reconstructed 3D model of the prostate with the previously reconstructed 3D model of the prostate, since the prostate is a mobile soft tissue organ with the deformation or natural growth in size. There is also potential disagreement in the prostate segmentation procedures between previous and current reconstructions of the 3D prostate model.

The recent report on the clinical use of the Artemis^™ system [32] presented an initial experience of 171 men undergoing Artemis system-guided biopsy. Interestingly, the 11 patients who underwent the 3D TRUS-Artemis biopsy were later re-biopsied to evaluate the ability of the Artemis device to guide biopsy needles back precisely to the area at which prior biopsy cores had been taken. The ability to return to prior biopsy sites was mathematically within 1.2 mm on the digital model by geometric analysis. However, the error was determined only by mathematical comparison between the registered digital images of the previous and revisited biopsy trajectories in the computed model, but unfortunately in the study it was not determined by the analysis of the real anatomical sampling sites in the human prostate tissues between the previous and revisited biopsies. Since the reported revisiting accuracy of 1.2 mm is only the outcome of mathematical image analysis in the registered digital model, the actual revisiting sampling accuracy of this system in the real prostate tissues is still pending. There are multiple factors to cause certain errors in 3D localization between the previous and current targeting, which include the processes of segmentation, registration, patient shift, prostate motion, prostate deformation, robotic arm tracking, needle bending, and so on. Even so, the Artemis^™ system offers a powerful promising opportunity of the image-based 3D mapping of prostate cancer, using a digitalized 3D prostate model with real-time tracked and documented biopsy trajectory, potentially improving the current and future practice of prostate biopsy and focal therapy. The emerging generation of the Artemis robotic arm achieves light and easy handling for more physician-friendly manipulation of the TRUS probe (Fig. 9.2).

To preserve the accuracy of a planned intervention, any image guidance system requires the capture of real-time imaging to constantly update the 3D planning. The critical importance of real-time image capture and corresponding real-time 3D planning is supported by the successful practices in some of the most investigated image-guided therapies for prostate cancer, including TRUS-guided focal cryosurgery and high-intensity focused ultrasound (HIFU) [33–35]. The initial 3D planning models for these technologies have been developed to be adjustable at any time during the intervention, based on the comparison of initially referenced images and real-time imaging. This capability is considered a key feature to achieving the precision and efficacy necessary for image-guided targeted focal therapy.

Registration of the reconstructed 3D planning prostate model to the rigid stereotactic navigation frame is the important computing process by which the coordinates of the 3D planning model are transformed to the coordinates of the stereotactic device. This step can be achieved very precisely by rigid registration or digitalized coincidence between the two spaces, the one with the other. Importantly, however, it is well-known that the prostate is a mobile organ, affected by the physiological movements of the pelvic organs such as the rectum and bladder [36, 37]. Actually, cine MRI is able to demonstrate real-time prostate movements related to the rectum, resulting in significant displacements of the prostate gland [38]. Delivery of the biopsy needle or therapeutic ablative probe needs to take into account these organ movements. In other words, the previously reconstructed 3D surgical planning model is becoming useless for real-time guidance in the real space, once prostate motion or deformation has occurred.

It is critical for precise delivery of needle or probe to ensure accurate and reliable localization of the intended target. If unwanted prostatic motion occurred during treatment, this may result in a geographical error impacting on the therapeutic effectiveness in terms of decreased cure rates. In order to avoid such errors, planning of therapeutic fields would have to have an adequate margin to encompass any possible movement of the intended target lesion in the prostate. One implication of possible prostate movement during treatment is that therapeutic contour margins may have to be increased in order to encompass such possible target displacements.

A recent innovative solution in registration of the 3D volume data for the soft tissue organ was the introduction of the “Elastic image-fusion” technique.

The Urostation^™ system (Koelis, France) is an innovative navigation system of prostate biopsy and intervention with the introduction of the “real-time 3D TRUS” image to track every single prostate biopsy with the actual 3D space of the prostate as well as the introduction of “elastic-image fusion” technology [9, 39]. The major advantage of this system is to apply the real-time 3D TRUS probe to acquire real-time 3D volume data of the prostate. At each biopsy procedure in the use of the Urostation^™, the3D prostate volume data are captured in real time to document a hyper-echoic needle trajectory in the real 3D space of the prostate at the time of biopsy firing. On the other hand, since in any other systems 3D prostate volume data are built up only once for the initial planning step by reconstruction out of multiple 2D images, the tracked biopsy location needs to be virtually overlaid on the initially reconstructed model, which may potentially cause registration error due to prostate motion or patient shift. Importantly, since every single prostate biopsy image taken by the Urostation^™ is captured with its digitalized coordinates from the proximal (x1, y1, z1) and the distal end (x2, y2, z2) of the trajectory within the real 3D space of the prostate, the coordinates are available for future revisiting biopsies in active surveillance, or retargeting positive lesions in focal therapy (Fig. 9.3). As the real-time 3D TRUS probe can be manipulated free hand, the use of this system does not change the style of current urological practice at all. However, limitation of this system includes that the freehand manual use of the US probe could create error, if the US probe is not stabilized in the same spatial position, during image acquisition and registration. Therefore, future introduction of robotic handling of the US probe could potentially further improve the targeting accuracy, due to robotic stabilizing orientation of US probe during 3D image acquisition [40].

A further unique ability of this system is the introduction of an elastic (nonrigid) image-fusion technique. For example, in MR/TRUS image fusion technology, the registration and fusion of two imaging modalities are roughly achieved by “rigid fusion,” where the integration of two corresponding images could be achieved by simple rotation and/or magnification. However, since most soft tissue organs such as the prostate do not conform to a rigid 3D shape between the preoperative and intra-operative fields, it has significant limitations when there are deformations to the 3D shape of the prostate between preoperative MR and real-time TRUS during the procedure. In contrast, “nonrigid (elastic) fusion” merges two corresponding images by the accommodation of some real-time local/partial stretching or deflection of the images. Elastic fusion allows real-time “elastic” deformation of the preoperative MR image mirroring changes in real-time 3D TRUS imaging. This feature is available in Urostation^™, and it has great potential to enhance the precision of lesion-targeted biopsy and the intervention of image-visible prostate cancer (Fig. 9.4) [41].