Minimally invasive percutaneous interventions, such as needle biopsies, benefit from imaging guidance during e.g. instrument positioning, monitoring of therapeutic progress, and assessment of therapeutic outcome. MRI with its unparalleled soft tissue contrast and ability for functional imaging would be ideally suited for interventional guidance¹, and closed-bore high-field MR systems would be highly preferable to achieve optimal image quality. However, the magnet size and the small bore diameters severely limit patient access.

Here, we present the first accuracy evaluation of a novel small assistance system which we combined with a passive tracking method for instrument positioning and alignment. The system was studied in a phantom experiment to assess the theoretically achievable accuracy in needle procedures.

The assistance system (prototype, iSYS Medizintechnik GmbH, Kitzbuehel, Austria; Fig. 1) consists of a gross-positioning platform with a rectangular opening for instrument insertion and features a micro-manipulator with a distal instrument holder. All components are constructed from MR-safe plastic materials. The manipulator comprises two plates which can be manually positioned via mechanical adjustment threads that are accessible through extension rods from outside the magnet. Each plate can be moved in two translational degrees of freedom (forward-backward – FB, left-right – LR). Manipulation of the lower plate moves the entire unit in FB/LR direction relatively to the gross-positioning platform (approximate translational range: 40×40 mm²). The upper plate can only be moved in FB/LR direction relatively to the lower plate translating into rotations of the instrument holder (approximate angular range ±30° in both directions). The distal holder can be either equipped with a customized cylindrical marker or a needle sleeve so that the marker’s symmetry axis is aligned with the planned needle pathway. The passive marker (outer diameter: 15 mm) is filled with a contrast agent solution (Gd-DTPA/H₂O 1:150) and automatically detected with a phase-only cross correlation (POCC) tracking sequence^2-4.

All experiments were carried out in a 1.5T whole body system (bore size: 60 cm; Siemens Symphony) using the system’s loop coil for signal reception. The setup was tested during needle procedures (Fig. 2) in an agar phantom with embedded fiducial targets (target number: 13, mean diameter: 8.1 mm, distributed over approximately 50×50×40 mm³) which was placed into a plastic casing mimicking the shape of a human torso (Fig. 1c). For each target the following needle procedure was done:
(i) attachment of the marker to the instrument holder at the manipulator’s center position,
(ii) instrument positioning under online guidance with the POCC tracking sequence (tracking time: 1.9 s) until the planned needle pathway was aligned with the target,
(iii) exchange of marker and needle sleeve,
(iv) insertion of a 16G needle (Somatex GmbH, Teltow, Germany) into the target under online monitoring with a HASTE sequence (acquisition time: 2.5 s),
(v) withdrawal of the needle and acquisition of a high-resolution (0.5×0.5×0.5 mm³) 3D bSSFP data set to evaluate the needle pathway.
The lateral distance of the needle pathway to the geometric center of each target was measured using reformatted multi-planar views of the bSSFP data (Fig. 3).

Without special training, all targets were successfully punctured without unintentional perforating surrounding targets within a mean procedure time (duration for instrument positioning, exchange of marker and needle sleeve, and needle insertion) of about 6.5 minutes. Figure 2 exemplarily shows images during instrument tracking, at the end of needle insertion, and after needle withdrawal. The needle channels are clearly depicted in the post-insertion images. A mean lateral distance of the needle channels to the target centers of 1.7±1.0 mm was found (cf. Table 1).

The novel assistance system together with the passive tracking technique provided an accurate and versatile setup for MR-guided needle interventions. Mean procedural duration (6.5 min) and accuracy (1.7 mm) were comparable to previous studies^2-4, which was sufficient to reliably puncture target structures with sizes of about 8 mm.

Compared to other assistance systems, the setup in its current implementation appeals as simple and accurate but could be expanded in future e.g. with more advanced robotic technologies.^5-7 With its flexible design, it is not restricted to single interventional scenarios and could be used in many abdominal applications, e.g. transgluteal prostate biopsies⁸. The main components are reusable and only small parts (distal holder, marker, instruments) are intended as sterile disposables which might be important for clinical applicability. Besides parameter optimization, e.g. temporal resolution of the POCC sequence, future work will focus on in vivo testing and integration of a dedicated RF coil. Based on our preliminary work, the system demonstrates its potential as an accurate and versatile in MR-guided interventions.