Automated fiducial marker detection and localization in volumetric computed tomography images: a three-step hybrid approach with deep learning

Abstract

Purpose: Automating fiducial detection and localization in the patient's pre-operative images can lead to better registration accuracy, reduced human errors, and shorter intervention time. Most current approaches are optimized for a single marker type, mainly spherical adhesive markers. A fully automated algorithm is proposed and evaluated for screw and spherical titanium fiducials, typically used in high-accurate frameless surgical navigation.Approach: The algorithm builds on previous approaches with morphological functions and pose estimation algorithms. A 3D convolutional neural network (CNN) is proposed for the fiducial classification task and evaluated for both traditional closed-set and emerging open-set classifiers. A proposed digital ground-truth experiment, with cone-beam computed tomography (CBCT) imaging software, is performed to determine the localization accuracy of the algorithm. The localized fiducial positions in the CBCT images by the presented algorithm were compared to the actual known positions in the virtual phantom models. The difference represents the fiducial localization error (FLE).Results: A total of 241 screws, 151 spherical fiducials, and 1550 other structures are identified with the best true positive rate 95.9% for screw and 99.3% for spherical fiducials at 8.7% and 3.4% false positive rate, respectively. The best achieved FLE mean and its standard deviation for a screw and spherical marker are 58 ( 14) and 14 ð6Þ μm, respectively.Conclusions: Accurate marker detection and localization were achieved, with spherical fiducials being superior to screws. Large marker volume and smaller voxel size yield significantly smaller FLEs. Attenuating noise by mesh smoothing has a minor effect on FLE. Future work will focus on expanding the CNN for image segmentation.

Fiducial marker recovery and detection from severely truncated data in navigation‐assisted spine surgery

“…Although the abovementioned deep learning approaches are investigated for markers without distortion, they have the risk of FP detection. For example, the hybrid approach 32 has FP rates of 8.7% and 3.4% for screws and spherical fiducial markers, respectively. In this work, the fiducial markers reconstructed from severely truncated data suffer from shape distortion and intensity decrease, making it difficult for the above approaches to detect them accurately.…”

“…With marker recovery, various existing marker detection algorithms 13,14,[30][31][32] can be applied. In this work,…”

Novel microscope-based visual display and nasopharyngeal registration for auditory brainstem implantation: a feasibility study in an ex vivo model

“…2). An algorithm was developed and verified to automatically localize fiducials in preoperative imagery and match these positions with the intraoperative positions provided by integrated magnetic sensors [13].…”

“…2c-e). A landmark-based registration [21] served to pair positions from magnetic sensors with the centroids of the detected spherical fiducials in images [9,13]. This process was automated and was repeated multiple times during the intervention.…”

“…The predicted TRE at the surgical point of interest in the foramen of Luschka was 1.75 mm and 1.76 mm for the left and the right ear, respectively. The TRE was modeled using [17] with the fiducial configuration obtained from the CT scan; and input FLE (fiducial localization error) [16] resulted as in a combination of 0.1 mm error in image space [13] and…”