PURPOSE: Needle biopsy of the prostate is guided by Transrectal Ultrasound (TRUS) imaging. The TRUS images do not provide proper spatial localization of malignant tissues due to the poor sensitivity of TRUS to visualize early malignancy. Magnetic Resonance Imaging (MRI) has been shown to be sensitive for the detection of early stage malignancy, and therefore, a novel 2D deformable registration method that overlays pre-biopsy MRI onto TRUS images has been proposed. METHOD: The registration method involves B-spline deformations with Normalized Mutual Information (NMI) as the similarity measure computed from the texture images obtained from the amplitude responses of the directional quadrature filter pairs. Registration accuracy of the proposed method is evaluated by computing the Dice Similarity coefficient (DSC) and 95% Hausdorff Distance (HD) values for 20 patients prostate mid-gland slices and Target Registration Error (TRE) for 18 patients only where homologous structures are visible in both the TRUS and transformed MR images. RESULTS: The proposed method and B-splines using NMI computed from intensities provide average TRE values of 2.64 ± 1.37 and 4.43 ± 2.77 mm respectively. Our method shows statistically significant improvement in TRE when compared with B-spline using NMI computed from intensities with Student's t test p = 0.02. The proposed method shows 1.18 times improvement over thin-plate splines registration with average TRE of 3.11 ± 2.18 mm. The mean DSC and the mean 95% HD values obtained with the proposed method of B-spline with NMI computed from texture are 0.943 ± 0.039 and 4.75 ± 2.40 mm respectively. CONCLUSIONS: The texture energy computed from the quadrature filter pairs provides better registration accuracy for multimodal images than raw intensities. Low TRE values of the proposed registration method add to the feasibility of it being used during TRUS-guided biopsy.
PURPOSE: Needle biopsy of the prostate is guided by Transrectal Ultrasound (TRUS) imaging. The TRUS images do not provide proper spatial localization of malignant tissues due to the poor sensitivity of TRUS to visualize early malignancy. Magnetic Resonance Imaging (MRI) has been shown to be sensitive for the detection of early stage malignancy, and therefore, a novel 2D deformable registration method that overlays pre-biopsy MRI onto TRUS images has been proposed. METHOD: The registration method involves B-spline deformations with Normalized Mutual Information (NMI) as the similarity measure computed from the texture images obtained from the amplitude responses of the directional quadrature filter pairs. Registration accuracy of the proposed method is evaluated by computing the Dice Similarity coefficient (DSC) and 95% Hausdorff Distance (HD) values for 20 patients prostate mid-gland slices and Target Registration Error (TRE) for 18 patients only where homologous structures are visible in both the TRUS and transformed MR images. RESULTS: The proposed method and B-splines using NMI computed from intensities provide average TRE values of 2.64 ± 1.37 and 4.43 ± 2.77 mm respectively. Our method shows statistically significant improvement in TRE when compared with B-spline using NMI computed from intensities with Student's t test p = 0.02. The proposed method shows 1.18 times improvement over thin-plate splines registration with average TRE of 3.11 ± 2.18 mm. The mean DSC and the mean 95% HD values obtained with the proposed method of B-spline with NMI computed from texture are 0.943 ± 0.039 and 4.75 ± 2.40 mm respectively. CONCLUSIONS: The texture energy computed from the quadrature filter pairs provides better registration accuracy for multimodal images than raw intensities. Low TRE values of the proposed registration method add to the feasibility of it being used during TRUS-guided biopsy.
Authors: Irving Kaplan; Nicklas E Oldenburg; Paul Meskell; Michael Blake; Paul Church; Edward J Holupka Journal: Magn Reson Imaging Date: 2002-04 Impact factor: 2.546
Authors: Baowei Fei; Jeffrey L Duerk; Daniel T Boll; Jonathan S Lewin; David L Wilson Journal: IEEE Trans Med Imaging Date: 2003-04 Impact factor: 10.048
Authors: Anurag K Singh; Jochen Kruecker; Sheng Xu; Neil Glossop; Peter Guion; Karen Ullman; Peter L Choyke; Bradford J Wood Journal: BJU Int Date: 2007-12-05 Impact factor: 5.588
Authors: Sota Oguro; Junichi Tokuda; Haytham Elhawary; Steven Haker; Ron Kikinis; Clare M C Tempany; Nobuhiko Hata Journal: J Magn Reson Imaging Date: 2009-11 Impact factor: 4.813
Authors: Jayashree Kalpathy-Cramer; Musaddiq Awan; Steven Bedrick; Coen R N Rasch; David I Rosenthal; Clifton D Fuller Journal: J Digit Imaging Date: 2014-02 Impact factor: 4.056
Authors: John A Onofrey; Lawrence H Staib; Saradwata Sarkar; Rajesh Venkataraman; Cayce B Nawaf; Preston C Sprenkle; Xenophon Papademetris Journal: Med Image Anal Date: 2017-04-12 Impact factor: 8.545
Authors: John A Onofrey; Lawrence H Staib; Saradwata Sarkar; Rajesh Venkataraman; Xenophon Papademetris Journal: Proc IEEE Int Symp Biomed Imaging Date: 2015-04
Authors: Yipeng Hu; Eli Gibson; Hashim Uddin Ahmed; Caroline M Moore; Mark Emberton; Dean C Barratt Journal: Med Image Anal Date: 2015-10-31 Impact factor: 8.545