Three-dimensional CT-Guided Bronchoscopy With a Real-Time Electromagnetic Position Sensor

Feuerstein

Deguchi

et al. 2012

Medical Image Analysis

This paper presents a new hybrid camera motion tracking method for bronchoscopic navigation combining SIFT, epipolar geometry analysis, Kalman filtering, and image registration. In a thorough evaluation, we compare it to state-of-the-art tracking methods. Our hybrid algorithm for predicting bronchoscope motion uses SIFT features and epipolar constraints to obtain an estimate for interframe pose displacements and Kalman filtering to find an estimate for the magnitude of the motion. We then execute bronchoscope tracking by performing image registration initialized by these estimates. This procedure registers the actual bronchoscopic video and the virtual camera images generated from 3D chest CT data taken prior to bronchoscopic examination for continuous bronchoscopic navigation. A comparative assessment of our new method and the state-of-the-art methods is performed on actual patient data and phantom data. Experimental results from both datasets demonstrate a significant performance boost of navigation using our new method. Our hybrid method is a promising means for bronchoscope tracking, and outperforms other methods based solely on Kalman filtering or image features and image registration.

Section: Phantom Studymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Development and comparison of new hybrid motion tracking for bronchoscopic navigation

Feuerstein

Deguchi

et al. 2012

Medical Image Analysis

“…We compare the proposed method to our own implementations of four previously published approaches, which have already been applied to similar trajectories: bronchoscope tracking by EMT only [1], intensity-based registration (IBR) with direct initialization from EMT [2], IBR with dynamic initialization from EMT [3], and IBR with a Sequential Monte Carlo sampler based on EMT [4]. Quantitative results for the accuracy are given in Table 1.…”

Section: Evaluation and Resultsmentioning

confidence: 99%

“…Since bronchoscopy is an inherently monitor-based procedure, augmentation of video images with guidance or targeting information is straightforward and promises high clinical acceptance due to smooth integration into the clinical workflow. The combination of flexible bronchoscopy with electromagnetic tracking (EMT) was first reported by Solomon et al [1], and hybrid image-based and EM tracking was proposed by Mori et al [2], and was improved by Luo et al [3,4] and Soper et al [5]. Hybrid tracking complements the advantages of image-based tracking and EMT, since imagebased tracking is inaccurate for frames with little to no information due to low contrast.…”

Section: Introductionmentioning

confidence: 99%

Deformable Registration of Bronchoscopic Video Sequences to CT Volumes with Guaranteed Smooth Output

Reichl

Lecture Notes in Computer Science

Menzel

et al. 2011

Abstract. We present a novel approach to tracking of flexible bronchoscopes by modeling the output as spatially continuous over time. Bronchoscopy is a widespread clinical procedure for diagnosis and treatment of lung diseases and navigation systems are highly needed. Tracking of the bronchoscope can be regarded as a deformable registration problem. In our approach we use hybrid image-based and electromagnetic tracking, and the bronchoscope pose relative to CT data is interpolated using Catmull-Rom splines for position and SLERP for orientation.We evaluate the method using ground truth poses manually selected by experts, where mean inter-expert agreement was determined as 1.26 mm. For four dynamic phantom data sets, the accuracy of our method is between 4.13 and 5.93 mm and shown to be equivalent to previous methods. We significantly improve inter-frame smoothness from 2.35-3.08 mm to 1.08-1.51 mm. Our method provides a more realistic and physically plausible solution with significantly less jitter. This quantitative result is confirmed by video output, which is much more consistent and robust, with fewer occasions of tracking loss or unexpected movement.

“…So far, two main approaches (or their combination) for bronchoscope tracking have been proposed in the literature: (a) sensor-based and (b) vision-based tracking. The former uses an electromagnetic (EM) tracking system (e.g., the superDimension navigation system [12]) to locate an electromagnetic sensor that is usually fastened at the bronchoscope tip to directly measure the bronchoscope camera position and orientation. The latter analyzes the bronchoscopic video images obtained from the bronchoscope camera to continuously track the bronchoscope tip on the basis of image registration methods [10,6].…”

Section: Introductionmentioning

confidence: 99%

An Application Driven Comparison of Several Feature Extraction Algorithms in Bronchoscope Tracking During Navigated Bronchoscopy

Lecture Notes in Computer Science

Feuerstein

Reichl

et al. 2010

Abstract. This paper compares Kanade-Lucas-Tomasi (KLT), speeded up robust feature (SURF), and scale invariant feature transformation (SIFT) features applied to bronchoscope tracking. In our study, we first use KLT, SURF, or SIFT features and epipolar constraints to obtaininterframe translation (up to scale) and orientation displacements and Kalman filtering to recover an estimate for the magnitude of the motion (scale factor determination), and then multiply inter-frame motion parameters onto the previous pose of the bronchoscope camera to achieve the predicted pose, which is used to initialize intensity-based image registration to refine the current pose of the bronchoscope camera. We evaluate the KLT-, SURF-, and SIFT-based bronchoscope camera motion tracking methods on patient datasets. According to experimental results, we may conclude that SIFT features are more robust than KLT and SURF features at predicting the bronchoscope motion, and all methods for predicting the bronchoscope camera motion show a significant performance boost compared to sole intensity-based image registration without an additional position sensor.