Rigid and deformable corrections in real-time using deep learning for prostate fusion biopsy

Bhardwaj, Aditya; Park, Jun-Sung; Mukhopadhyay, Soumik; Sharda, Sikander; Son, Yuri; Ajani, Bhavya; Kudavelly, Srinivas Rao

doi:10.1117/12.2548589

Cited by 5 publications

(13 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Trans-registration error is a measure of Euclidian distance between cross-marked anatomically homologous points on US and MRI, such as urethra, calcification and, in this case, PCa. It measures the distance between the landmark on MRI and the corresponding point mapped to the MRI from US, in 3D MRI space [10]. Our results showed that TRE decreased with the transition from rigid to nonrigid registration.…”

Section: Discussionmentioning

confidence: 73%

“…TRE was re-measured for nonrigid registered images. A more detailed description of the technique was published in a previous study [10].…”

Section: Image Fusion Registration and Error Measurementsmentioning

confidence: 99%

“…However, many US vendors currently either provide only rigid registration or a sophisticated elastic fusion system with retrospective nonrigid registration [8,9]. We developed a solution offering both rigid and real-time corrected nonrigid registration during the live biopsy without requiring an intermediate 3D US [10]. With this system, we show error reduction through real-time nonrigid registration technique for prostate fusion biopsies.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Comparison of Accuracies between Real-Time Nonrigid and Rigid Registration in the MRI–US Fusion Biopsy of the Prostate

et al. 2021

Self Cite

View full text Add to dashboard Cite

Magnetic resonance imaging (MRI) is increasingly important in the detection and localization of prostate cancer. Regarding suspicious lesions on MRI, a targeted biopsy using MRI fused with ultrasound (US) is widely used. To achieve a successful targeted biopsy, a precise registration between MRI and US is essential. The purpose of our study was to show any decrease in errors using a real-time nonrigid registration technique for prostate biopsy. Nineteen patients with suspected prostate cancer were prospectively enrolled in this study. Registration accuracy was calculated by the measuring distance of corresponding points by rigid and nonrigid registration between MRI and US, and compared for rigid and nonrigid registration methods. Overall cancer detection rates were also evaluated by patient and by core. Prostate volume was measured automatically from MRI and manually from US, and compared to each other. Mean distances between the corresponding points in MRI and US were 5.32 ± 2.61 mm for rigid registration and 2.11 ± 1.37 mm for nonrigid registration (p < 0.05). Cancer was diagnosed in 11 of 19 patients (57.9%), and in 67 of 266 biopsy cores (25.2%). There was no significant difference in prostate-volume measurement between the automatic and manual methods (p = 0.89). In conclusion, nonrigid registration reduces targeting errors.

show abstract

Section: Discussionmentioning

confidence: 73%

“…TRE was re-measured for nonrigid registered images. A more detailed description of the technique was published in a previous study [10].…”

Section: Image Fusion Registration and Error Measurementsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Comparison of Accuracies between Real-Time Nonrigid and Rigid Registration in the MRI–US Fusion Biopsy of the Prostate

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…A majority of the methods predict slices localization with respect to a fixed reference volume by only inputting the moving slices. Few studies combined both 2D and 3D inputs in the same model 25,26 . Even if most of these studies aim to solve 2D/3D registration, comparing them remains difficult, as data involved may be very different in terms of modality (mostly MRI, vs.…”

Section: D/3d Deep Registration For Real-time Navigation Assistancementioning

confidence: 99%

“…We find several main structures among all these studies: (i) some use a hierarchical structure 6,7,[23][24][25][26][27] . It allows a coarse-to-fine registration by using results from a previous step to initialize the next one.…”

Section: D/3d Deep Registration For Real-time Navigation Assistancementioning

confidence: 99%

2D/3D deep registration for real-time prostate biopsy navigation

Dupuy

Beitone

Troccaz

et al. 2021

Medical Imaging 2021: Image-Guided Procedures, Robotic Interventions, and Modeling

View full text Add to dashboard Cite

The accuracy of biopsy sampling and the related tumor localization are major issues for prostate cancer diagnosis and therapy. However, the ability to navigate accurately to biopsy targets faces several difficulties coming from both transrectal ultrasound (TRUS) image guidance properties and prostate motion or deformation. To reduce inaccuracy and exam duration, the main objective of this study is to develop a real-time navigation assistance. The aim is to provide the current probe position and orientation with respect to the deformable organ and the next biopsy targets. We propose a deep learning real-time 2D/3D registration method based on Convolutional Neural Networks (CNN) to localize the current 2D US image relative to the available 3D TRUS reference volume. We experiment several scenarii combining different input data including: pair of successive 2D US images, the optical flow between them and current probe tracking information. The main novelty of our study is to consider prior navigation trajectory information by introducing previous registration result. This model is evaluated on clinical data through simulated biopsy trajectories. The results highlight significant improvement by exploiting trajectory information especially through prior registration results and probe tracking parameters. With such trajectory information, we achieve an average registration error of 2.21 mm ± 2.89. The network demonstrates efficient generalization capabilities on new patients and new trajectories, which is promising for successful continuous tracking during biopsy procedure.

show abstract

MR to ultrasound image registration with segmentation‐based learning for HDR prostate brachytherapy

Chen

et al. 2021

Medical Physics

View full text Add to dashboard Cite

Purpose Propagation of contours from high‐quality magnetic resonance (MR) images to treatment planning ultrasound (US) images with severe needle artifacts is a challenging task, which can greatly aid the organ contouring in high dose rate (HDR) prostate brachytherapy. In this study, a deep learning approach was developed to automatize this registration procedure for HDR brachytherapy practice. Methods Because of the lack of training labels and difficulty of accurate registration from inferior image quality, a new segmentation‐based registration framework was proposed for this multi‐modality image registration problem. The framework consisted of two segmentation networks and a deformable registration network, based on the weakly ‐supervised registration strategy. Specifically, two 3D V‐Nets were trained for the prostate segmentation on the MR and US images separately, to generate the weak supervision labels for the registration network training. Besides the image pair, the corresponding prostate probability maps from the segmentation were further fed to the registration network to predict the deformation matrix, and an augmentation method was designed to randomly scale the input and label probability maps during the registration network training. The overlap between the deformed and fixed prostate contours was analyzed to evaluate the registration accuracy. Three datasets were collected from our institution for the MR and US image segmentation networks, and the registration network learning, which contained 121, 104, and 63 patient cases, respectively. Results The mean Dice similarity coefficient (DSC) results of the two prostate segmentation networks are 0.86 ± 0.05 and 0.90 ± 0.03, for MR images and the US images after the needle insertion, respectively. The mean DSC, center‐of‐mass (COM) distance, Hausdorff distance (HD), and averaged symmetric surface distance (ASSD) results for the registration of manual prostate contours were 0.87 ± 0.05, 1.70 ± 0.89 mm, 7.21 ± 2.07 mm, 1.61 ± 0.64 mm, respectively. By providing the prostate probability map from the segmentation to the registration network, as well as applying the random map augmentation method, the evaluation results of the four metrics were all improved, such as an increase in DSC from 0.83 ± 0.08 to 0.86 ± 0.06 and from 0.86 ± 0.06 to 0.87 ± 0.05, respectively. Conclusions A novel segmentation‐based registration framework was proposed to automatically register prostate MR images to the treatment planning US images with metal artifacts, which not only largely saved the labor work on the data preparation, but also improved the registration accuracy. The evaluation results showed the potential of this approach in HDR prostate brachytherapy practice.

show abstract

Rigid and deformable corrections in real-time using deep learning for prostate fusion biopsy

Cited by 5 publications

References 0 publications

Comparison of Accuracies between Real-Time Nonrigid and Rigid Registration in the MRI–US Fusion Biopsy of the Prostate

Comparison of Accuracies between Real-Time Nonrigid and Rigid Registration in the MRI–US Fusion Biopsy of the Prostate

2D/3D deep registration for real-time prostate biopsy navigation

MR to ultrasound image registration with segmentation‐based learning for HDR prostate brachytherapy

Contact Info

Product

Resources

About