An algorithm for tracking microcatheters in fluoroscopy

Unique migration of a dental needle into the parapharyngeal space: successful removal by an intraoral approach and simulation for tracking visibility in X-ray fluoroscopy

OBJECTIVES

The first objective was to describe a novel case of migration of a broken dental needle into the parapharyngeal space. The second was to address the importance of simulation elucidating visualization of such a thin needle under X-ray fluoroscopy.

METHODS

Clinical case records (including computed tomography [CT] and surgical approaches) were reviewed, and a simulation experiment using a head phantom was conducted using the same settings applied intraoperatively.

RESULTS

A 36-year-old man was referred after failure to locate a broken 31-G dental needle. Computed tomography revealed migration of the needle into the parapharyngeal space. Intraoperative X-ray fluoroscopy failed to identify the needle, so a steel wire was applied as a reference during X-ray to locate the foreign body. The needle was successfully removed using an intraoral approach with tonsillectomy under surgical microscopy. The simulation showed that the dental needle was able to be identified only after applying an appropriate compensating filter, contrasting with the steel wire.

CONCLUSION

Meticulous preoperative simulation regarding visual identification of dental needle foreign bodies is mandatory. Intraoperative radiography and an intraoral approach with tonsillectomy under surgical microscopy offer benefits for accessing the parapharyngeal space, specifically for cases medial to the great vessels.

Real-time 3D position reconstruction of guidewire for monoplane X-ray

We present a novel real-time method for the 3D reconstruction of the guidewire using a monoplane X-ray. The method consists of two steps: (1) the backprojection step to reconstruct a 3D surface that contains the guidewire and (2) the optimization step to select a curve on the surface that is the best match under the pre-specified constraints. The proposed method utilizes a priori knowledge in the form of a volume that indicates positions of the blood vessels and thus restricts the reconstruction. The reconstruction precision is limited by the local thickness of the vessels. The method is quantitatively evaluated on five phantom datasets and qualitatively on two patient datasets. For the phantom datasets the average reconstruction error is resolution limited to 1-2 voxels and is biased in the depth direction. The worst-case reconstruction error for any point, including the guidewire tip, is not larger than the local vessel thickness. A visual inspection of results for the patient datasets shows the guidewire is always placed in the proper vessel and is aligned with the 2D image, which is sufficient for the guidewire navigation. The developed implementation achieves the processing speed of 12 fps using Core™i7 CPU 920 at 2.67 GHz.