Computer assisted electromagnetic navigation improves accuracy in computed tomography guided interventions: A prospective randomized clinical trial

Abstract

Purpose: To assess the accuracy and usability of an electromagnetic navigation system designed to assist Computed Tomography (CT) guided interventions.

Materials and methods: 120 patients requiring a percutaneous CT intervention (drainage, biopsy, tumor ablation, infiltration, sympathicolysis) were included in this prospective randomized trial. Nineteen radiologists participated. Conventional procedures (CT group) were compared with procedures assisted by a navigation system prototype using an electromagnetic localizer to track the position and orientation of a needle holder (NAV group). The navigation system displays the needle path in real-time on 2D reconstructed CT images extracted from the 3D CT volume. The regional ethics committee approved this study and all patients gave written informed consent. The main outcome was the distance between the planned trajectory and the achieved needle trajectory calculated from the initial needle placement.

Results: 120 patients were analyzable in intention-to-treat (NAV: 60; CT: 60). Accuracy improved when the navigation system was used: distance error (in millimeters: median[P25%; P75%]) with NAV = 4.1[2.7; 9.1], vs. with CT = 8.9[4.9; 15.1] (p<0.001). After the initial needle placement and first control CT, fewer subsequent CT acquisitions were necessary to reach the target using the navigation system: NAV = 2[2; 3]; CT = 3[2; 4] (p = 0.01).

Conclusion: The tested system was usable in a standard clinical setting and provided significant improvement in accuracy; furthermore, with the help of navigation, targets could be reached with fewer CT control acquisitions.

Fig 1. Patient flow.

Fig 2. Main outcome computation.

The accuracy of the initial needle placement is computed as the maximum distance between the planned trajectory and the achieved trajectory of the needle: Distance error = max(d,D), with d = distance between the achieved trajectory and the Planned Target Point, D = distance between the achieved trajectory and the Planned Entry Point. The achieved trajectory is defined as the straight line which passes through the actual position of the needle extracted from the first control CT acquisition, i.e. the straight line passing through the needle skin entry point and the needle tip (cf. Fig 3). This maximum distance error is computed along the entire real needle path, as if no iterative needle trajectory adjustments were made after the initial needle placement and as if the needle was pushed straight ahead up to the target depth. Thus it is related to the initial risk that neighboring critical tissue would be punctured by the needle, or that the target would not be reached by the needle tip. The angle θ between the planned trajectory and the needle is also computed.

Fig 3. Examples of image data processing during main outcome computation.

Distance and angle measurements are performed in two steps using a dedicated software: 1: The planned trajectory (i.e. planned entry point and planned target) is extracted from images saved by the operator before the needle is placed, showing the chosen trajectory that the operator will try to reproduce when placing the needle. The operator saves the planned trajectory either on the CT console (CT group) or on the navigation system (NAV group). 2: Once the needle has been positioned, the achieved trajectory is extracted from the first control CT acquisition showing the actual position of the needle. Multiplanar image reconstructions allow a fine selection of the points that define the entry point and the tip of the needle. The distance and angle between the planned and achieved trajectories (cf. Fig 2) are then computed using the software and saved for statistical analysis.

Fig 4. CTNAV trial protocol.

The operator tries to reproduce as accurately as possible the planned trajectory using a navigated procedure (NAV group) or the standard procedure (CT group). The main outcome is the difference in distance and angle between the achieved and the planned trajectories.

Fig 5. Angle and distance between the planned and the achieved trajectories according to NAV/CT group and intervention difficulty.

Fig 6. Illustration case (NAV).

This 65 year old male patient had a record of lung carcinoma with recurrence after chemotherapy. There was a collegial indication for a percutaneous CT-guided biopsy for immunohistochemical (IHC) characterization. The patient gave written informed consent. The operator (senior) predicted an easy intervention. The patient was randomized to the NAV group. Planning was performed using the navigation system directly on the patient, with a double angulated intercostal trajectory (a, b). The first control acquisition showed the achieved trajectory (c, d) to be very similar to the planned trajectory (step Δ₁). Two more control acquisitions were necessary to reach the target (step Δ₂), and perform the biopsy. No adverse event was reported. Pathologic findings on the biopsy sample showed an Anaplastic Lymphoma Kinase (ALK) negative acinar adenocarcinoma of the lung.

Fig 7. Illustration case (CT).

This 54 year old female patient had a growing right adrenal gland mass (white arrow) and an indication for a percutaneous CT-guided biopsy. She gave written informed consent. The operator (resident) in agreement with the supervising senior operator predicted a difficult intervention, due to the necessity of a double angulated trajectory. The patient was randomized to the CT group. The planned trajectory was saved on the CT console. The control CT acquired after the initial needle placement (step Δ₁) showed a large difference between the planned (a, b) and the achieved trajectories (multiplanar reconstructions) (c, d). Twelve more attempts and iterative CT control acquisitions were required to reach the target (step Δ₂). After biopsy, the last control acquisition showed a small periadrenal hematoma without the need for additional treatment. Pathologic findings on the biopsy sample showed adrenal tissues with metastasis from a malignant lung tumor.