Whole‐body bone segmentation from MRI for PET/MRI attenuation correction using shape‐based averaging

Abstract

Purpose: The authors evaluate the performance of shape-based averaging (SBA) technique for whole-body bone segmentation from MRI in the context of MRI-guided attenuation correction (MRAC) in hybrid PET/MRI. To enhance the performance of the SBA scheme, the authors propose to combine it with statistical atlas fusion techniques. Moreover, a fast and efficient shape comparisonbased atlas selection scheme was developed and incorporated into the SBA method. Methods: Clinical studies consisting of PET/CT and MR images of 21 patients were used to assess the performance of the SBA method. In addition, the authors assessed the performance of simultaneous truth and performance level estimation (STAPLE) and the selective and iterative method for performance level estimation (SIMPLE) combined with SBA. In addition, a local shape comparison scheme (L-Shp) was proposed to improve the performance of SBA. The SIMPLE method was applied globally (G-SIMPLE) while STAPLE method was employed at both global (G-STAPLE) and local (L-STAPLE) levels. The evaluation was performed based on the accuracy of extracted whole-body bones, fragmentation, and computation time achieved by the different methods. The majority voting (MV) atlas fusion scheme was also evaluated as a conventional and commonly used method. MRI-guided attenuation maps were generated using the different segmentation methods. Thereafter, quantitative analysis of PET attenuation correction was performed using CT-based attenuation correction as reference. Results: The SBA and MV methods resulted in considerable underestimation of bone identification (Dice ≈ 0.62) and high factious fragmentation error of contiguous structures. Applying global atlas selection or regularization (G-STAPLE and G-SIMPLE) to the SBA method enhanced bone segmentation accuracy up to a Dice = 0.66. The best results were achieved when applying the L-STAPLE method with a Dice of 0.76 and the L-Shp method with a Dice of 0.75. However, L-STAPLE required up to five-fold increased computation time compared to the L-Shp method. Moreover, both L-STAPLE and L-Shp methods resulted in less than 3% SUV mean relative error and 6% SUV mean absolute error in bony structures owing to superior bone identification accuracy. The quantitative analysis using joint histograms revealed good correlation between PET-MRAC images using the proposed L-Shp algorithm and the corresponding reference PET-CT images. Conclusions: The performance of SBA was enhanced through application of local atlas weighting or regularization schemes (L-STAPLE and L-Shp). Bone recognition, fragmentation of the contiguous structures, and quantitative PET uptake recovery improved dramatically using these methods while the proposed L-Shp method significantly reduced the computation time. C 2016 American Association of Physicists in Medicine. [http://dx.doi.org/10.1118/1.4963809]

Hybrid Imaging: Instrumentation and Data Processing

“…In general, these approaches perform reasonably well, although for specific application their accuracy may not be sufficient [222]. Thus, several advanced AC methods have been explored in conjunction with various types of PET/MR examinations [217,223,224]. Most of the developed AC methods are tailored to brain examinations and provide acceptable accuracies (with deviations below 5% respect to CT-based AC) [217].…”

MRI‐guided attenuation correction in torso PET/MRI: Assessment of segmentation‐, atlas‐, and deep learning‐based approaches in the presence of outliers

“…The three‐tissue class segmentation approach investigated in this work 28 exhibited poor performance in the presence of body truncation and metal artifacts. However, the latest (advanced) versions of the segmentation‐based synthetic CT generation methods tend to accounted for void signals in MR images (due to metal artifacts or body truncation) through exploiting prior knowledge in the form body atlas and/or template 26 . In this regard, Lindemann et al 50 proposed B0 homogenization using gradient enhancement to provide an extension of the transaxial FOV in MR imaging.…”

Truncation compensation and metallic dental implant artefact reduction in PET/MRI attenuation correction using deep learning-based object completion

“…However, since the original MR images contain plenty of anatomical details and information associated with seminal structures in the missing regions, it is expected that the object completion method performs better when applied on the original MR images, though this claim should be verified. Moreover, MR images, in the context of PET/MR imaging and MRI-only radiation therapy, are also utilized by atlas-based methods (Arabi and Zaidi 2016a), MRI-guided joint activity and attenuation reconstruction algorithms (Mehranian et al 2016a, Fuin et al 2017, Hwang et al 2019, dose estimation in MRI-only radiation planning (Arabi et al 2018), low-dose PET imaging (Chen et al 2019), deep learning-based synthetic CT generation (Arabi et al 2019), synergistic PET and MR image reconstruction (Mehranian et al 2019), and segmentation of anatomical structures (Arabi and Zaidi 2016b). The proposed method would also be applicable to continuous-valued AC maps generation from Dixon fat/water sequences (Wollenweber et al 2013) to accurately estimate the ratio/fraction of fat and water tissues in the missing regions.…”