Fast 5DOF needle tracking in iOCT.

Area of Science:

• Ophthalmic Microsurgery
• Medical Imaging
• Surgical Robotics

Background:

• Intraoperative optical coherence tomography (iOCT) offers high-resolution imaging for ophthalmic microsurgery.
• Accurate tracking of surgical instruments is crucial for precision and safety in microsurgery.
• Traditional instrument tracking often requires complex multimodal calibration.

Purpose of the Study:

• To develop and present a method for tracking the orientation and location of a surgical needle using iOCT.
• To enable direct analysis of instrument-tissue interaction within the iOCT space.
• To eliminate the need for complex multimodal calibration required by traditional methods.

Main Methods:

• Detection of the needle's intersection with the iOCT scan using a multistep ellipse fitting algorithm.
• Utilizing the directionality of the iOCT modality for accurate needle detection.
• Employing geometric modeling and a latency-aware estimator to infer the 5-degrees of freedom (DOF) pose during needle movement.

Main Results:

• The algorithm demonstrated high angular precision, particularly during lateral needle movements.
• Experiments on phantom data and ex vivo porcine eyes showed robust and consistent pose estimation.
• The proposed method outperformed baseline tracking techniques in terms of accuracy and consistency.

Conclusions:

• A 5-DOF pose of the surgical instrument can be successfully and robustly estimated using only cross-sectional iOCT information.
• The estimation process is highly efficient, completing in under 5.4 milliseconds on a standard CPU.
• This technique offers a simplified and effective approach to instrument tracking in iOCT-guided ophthalmic surgery.