A Fast Nonrigid Image Registration With Constraints on the Jacobian Using Large Scale Constrained Optimization

AbstractTo determine whether temporal subtraction (TS) CT obtained with non-rigid image registration improves detection of various bone metastases during serial clinical follow-up examinations by numerous radiologists. Six board-certified radiologists retrospectively scrutinized CT images for patients with history of malignancy sequentially. These radiologists selected 50 positive and 50 negative subjects with and without bone metastases, respectively. Furthermore, for each subject, they selected a pair of previous and current CT images satisfying predefined criteria by consensus. Previous images were non-rigidly transformed to match current images and subtracted from current images to automatically generate TS images. Subsequently, 18 radiologists independently interpreted the 100 CT image pairs to identify bone metastases, both without and with TS images, with each interpretation separated from the other by an interval of at least 30 days. Jackknife free-response receiver operating characteristics (JAFROC) analysis was conducted to assess observer performance. Compared with interpretation without TS images, interpretation with TS images was associated with a significantly higher mean figure of merit (0.710 vs. 0.658; JAFROC analysis, P = 0.0027). Mean sensitivity at lesion-based was significantly higher for interpretation with TS compared with that without TS (46.1% vs. 33.9%; P = 0.003). Mean false positive count per subject was also significantly higher for interpretation with TS than for that without TS (0.28 vs. 0.15; P < 0.001). At the subject-based, mean sensitivity was significantly higher for interpretation with TS images than that without TS images (73.2% vs. 65.4%; P = 0.003). There was no significant difference in mean specificity (0.93 vs. 0.95; P = 0.083). TS significantly improved overall performance in the detection of various bone metastases.

Download Full-text

Fully Automated Targeting Using Nonrigid Image Registration Matches Accuracy and Exceeds Precision of Best Manual Approaches to Subthalamic Deep Brain Stimulation Targeting in Parkinson Disease

Neurosurgery ◽

10.1227/neu.0000000000000714 ◽

2015 ◽

Vol 76 (6) ◽

pp. 756-765 ◽

Cited By ~ 21

Author(s):

Srivatsan Pallavaram ◽

Pierre-François D'Haese ◽

Wendell Lake ◽

Peter E. Konrad ◽

Benoit M. Dawant ◽

...

Keyword(s):

Deep Brain Stimulation ◽

Image Registration ◽

Parkinson Disease ◽

Brain Stimulation ◽

Red Nucleus ◽

Nonrigid Image Registration ◽

Direct Targeting ◽

Fully Automatic ◽

Stn Dbs ◽

Deep Brain

Abstract BACKGROUND: Finding the optimal location for the implantation of the electrode in deep brain stimulation (DBS) surgery is crucial for maximizing the therapeutic benefit to the patient. Such targeting is challenging for several reasons, including anatomic variability between patients as well as the lack of consensus about the location of the optimal target. OBJECTIVE: To compare the performance of popular manual targeting methods against a fully automatic nonrigid image registration-based approach. METHODS: In 71 Parkinson disease subthalamic nucleus (STN)-DBS implantations, an experienced functional neurosurgeon selected the target manually using 3 different approaches: indirect targeting using standard stereotactic coordinates, direct targeting based on the patient magnetic resonance imaging, and indirect targeting relative to the red nucleus. Targets were also automatically predicted by using a leave-one-out approach to populate the CranialVault atlas with the use of nonrigid image registration. The different targeting methods were compared against the location of the final active contact, determined through iterative clinical programming in each individual patient. RESULTS: Targeting by using standard stereotactic coordinates corresponding to the center of the motor territory of the STN had the largest targeting error (3.69 mm), followed by direct targeting (3.44 mm), average stereotactic coordinates of active contacts from this study (3.02 mm), red nucleus-based targeting (2.75 mm), and nonrigid image registration-based automatic predictions using the CranialVault atlas (2.70 mm). The CranialVault atlas method had statistically smaller variance than all manual approaches. CONCLUSION: Fully automatic targeting based on nonrigid image registration with the use of the CranialVault atlas is as accurate and more precise than popular manual methods for STN-DBS.

Download Full-text