Evaluation of the accuracy of deformable image registration on MRI with a physical phantom

Multiscale feature fusion network for 3D head MRI image registration

BACKGROUND

Image registration technology has become an important medical image preprocessing step with the wide application of computer-aided diagnosis technology in various medical image analysis tasks.

PURPOSE

We propose a multiscale feature fusion registration based on deep learning to achieve the accurate registration and fusion of head magnetic resonance imaging (MRI) and solve the problem that general registration methods cannot handle the complex spatial information and position information of head MRI.

METHODS

Our proposed multiscale feature fusion registration network consists of three sequentially trained modules. The first is an affine registration module that implements affine transformation; the second is to realize non-rigid transformation, a deformable registration module composed of top-down and bottom-up feature fusion subnetworks in parallel; and the third is a deformable registration module that also realizes non-rigid transformation and is composed of two feature fusion subnetworks in series. The network decomposes the deformation field of large displacement into multiple deformation fields of small displacement by multiscale registration and registration, which reduces the difficulty of registration. Moreover, multiscale information in head MRI is learned in a targeted manner, which improves the registration accuracy, by connecting the two feature fusion subnetworks.

RESULTS

We used 29 3D head MRIs for training and seven volumes for testing and calculated the values of the registration evaluation metrics for the new algorithm to register anterior and posterior lateral pterygoid muscles. The Dice similarity coefficient was 0.745 ± 0.021, the Hausdorff distance was 3.441 ± 0.935 mm, the Average surface distance was 0.738 ± 0.098 mm, and the Standard deviation of the Jacobian matrix was 0.425 ± 0.043. Our new algorithm achieved a higher registration accuracy compared with state-of-the-art registration methods.

CONCLUSIONS

Our proposed multiscale feature fusion registration network can realize end-to-end deformable registration of 3D head MRI, which can effectively cope with the characteristics of large deformation displacement and the rich details of head images and provide reliable technical support for the diagnosis and analysis of head diseases.

Detecting and quantifying spatial misalignment between longitudinal kilovoltage computed tomography (kVCT) scans of the head and neck by using convolutional neural networks (CNNs)

BACKGROUND

Adaptive radiotherapy (ART) aims to address anatomical modifications appearing during the treatment of patients by modifying the planning treatment according to the daily positioning image. Clinical implementation of ART relies on the quality of the deformable image registration (DIR) algorithms included in the ART workflow. To translate ART into clinical practice, automatic DIR assessment is needed.

OBJECTIVE

This article aims to estimate spatial misalignment between two head and neck kilovoltage computed tomography (kVCT) images by using two convolutional neural networks (CNNs).

METHODS

The first CNN quantifies misalignments between 0 mm and 15 mm and the second CNN detects and classifies misalignments into two classes (poor alignment and good alignment). Both networks take pairs of patches of 33x33x33 mm3 as inputs and use only the image intensity information. The training dataset was built by deforming kVCT images with basis splines (B-splines) to simulate DIR error maps. The test dataset was built using 2500 landmarks, consisting of hard and soft landmark tissues annotated by 6 clinicians at 10 locations.

RESULTS

The quantification CNN reaches a mean error of 1.26 mm (± 1.75 mm) on the landmark set which, depending on the location, has annotation errors between 1 mm and 2 mm. The errors obtained for the quantification network fit the computed interoperator error. The classification network achieves an overall accuracy of 79.32%, and although the classification network overdetects poor alignments, it performs well (i.e., it achieves a rate of 90.4%) in detecting poor alignments when given one.

CONCLUSION

The performances of the networks indicate the feasibility of using CNNs for an agnostic and generic approach to misalignment quantification and detection.

Prediction of adaptive strategies based on deformation vector field features for MR-guided adaptive radiotherapy of prostate cancer

BACKGROUND

The selection of adaptive strategies in MR-guided adaptive radiotherapy (MRgART) usually relies on subjective review of anatomical changes. However, this kind of review may lead to improper selection of adaptive strategy for some fractions.

PURPOSE

The purpose of this study was to develop prediction models based on deformation vector field (DVF) features for automatic and accurate strategy selection, using prostate cancer as an example.

METHODS

100 fractions of 20 prostate cancer patients were retrospectively selected in this study. Treatment plans using both adapt to position (ATP) strategy and adapt to shape (ATS) strategy were generated. Optimal adaptive strategy was determined according to dosimetric evaluation. DVFs of the deformable image registration (DIR) of daily MRI and CT simulation scans were extracted. The shape, first order statistics, and spatial features were extracted from the DVFs, subjected to further selection using the minimum redundancy maximum relevance (mRMR) method. The number of features (F_n ) was hyper-tuned using bootstrapping method, and then F_n indicating a peak area under the curve (AUC) value was used to construct three prediction models.

RESULTS

According to subjective review, the ATS strategy was adopted for all 100 fractions. However, the evaluation results showed that the ATP strategy could have met the clinical requirements for 23 (23%) fractions. The three prediction models showed high prediction performance, with the best performing model achieving an AUC value of 0.896, corresponding accuracy (ACC), sensitivity (SEN) and specificity (SPC) of 0.9, 0.958, and 0.667, respectively. The features used to construct prediction models included four features extracted from y direction of DVF (DVF_y ) and mask, one feature from z direction of DVF (DVF_z ). It indicated that the deformation along the anterior-posterior direction had a greater impact on determining the adaptive strategy than other directions.

CONCLUSIONS

DVF-feature-based models could accurately predict the adaptive strategy and avoid unnecessary selection of time-consuming ATS strategy, which consequently improves the efficiency of the MRgART process.

Evaluation of Multisource Adaptive MRI Fusion for Gross Tumor Volume Delineation of Hepatocellular Carcinoma

Purpose

Tumor delineation plays a critical role in radiotherapy for hepatocellular carcinoma (HCC) patients. The incorporation of MRI might improve the ability to correctly identify tumor boundaries and delineation consistency. In this study, we evaluated a novel Multisource Adaptive MRI Fusion (MAMF) method in HCC patients for tumor delineation.

Methods

Ten patients with HCC were included in this study retrospectively. Contrast-enhanced T1-weighted MRI at portal-venous phase (T1W_PP), contrast-enhanced T1-weighted MRI at 19-min delayed phase (T1W_DP), T2-weighted (T2W), and diffusion-weighted MRI (DWI) were acquired on a 3T MRI scanner and imported to in-house-developed MAMF software to generate synthetic MR fusion images. The original multi-contrast MR image sets were registered to planning CT by deformable image registration (DIR) using MIM. Four observers independently delineated gross tumor volumes (GTVs) on the planning CT, four original MR image sets, and the fused MRI for all patients. Tumor contrast-to-noise ratio (CNR) and Dice similarity coefficient (DSC) of the GTVs between each observer and a reference observer were measured on the six image sets. Inter-observer and inter-patient mean, SD, and coefficient of variation (CV) of the DSC were evaluated.

Results

Fused MRI showed the highest tumor CNR compared to planning CT and original MR sets in the ten patients. The mean ± SD tumor CNR was 0.72 ± 0.73, 3.66 ± 2.96, 4.13 ± 3.98, 4.10 ± 3.17, 5.25 ± 2.44, and 9.82 ± 4.19 for CT, T1W_PP, T2W, DWI, T1W_DP, and fused MRI, respectively. Fused MRI has the minimum inter-observer and inter-patient variations as compared to original MR sets and planning CT sets. GTV delineation inter-observer mean DSC across the ten patients was 0.81 ± 0.09, 0.85 ± 0.08, 0.88 ± 0.04, 0.89 ± 0.08, 0.90 ± 0.04, and 0.95 ± 0.02 for planning CT, T1W_PP, T2W, DWI, T1W_DP, and fused MRI, respectively. The patient mean inter-observer CV of DSC was 3.3%, 3.2%, 1.7%, 2.6%, 1.5%, and 0.9% for planning CT, T1W_PP, T2W, DWI, T1W_DP, and fused MRI, respectively.

Conclusion

The results demonstrated that the fused MRI generated using the MAMF method can enhance tumor CNR and improve inter-observer consistency of GTV delineation in HCC as compared to planning CT and four commonly used MR image sets (T1W_PP, T1W_DP, T2W, and DWI). The MAMF method holds great promise in MRI applications in HCC radiotherapy treatment planning.

Clinical Assessment of a Novel Ring Gantry Linear Accelerator-Mounted Helical Fan-Beam kVCT System

Purpose

To assess clinically relevant image quality metrics (IQMs) of helical fan beam kilovoltage (kV) fan beam computed tomography (CT).

Methods and Materials

kVCT IQMs were evaluated on an Accuray Radixact unit equipped with helical fan beam kVCT to assess the capabilities of this newly available modality. kVCT IQMs were evaluated and compared to a kVCT simulator and linear accelerator-based cone beam CTs (CBCT) using a commercial CBCT image quality phantom. kVCTs were acquired on the Accuray Radixact for all combinations of kVp and mAs in fine mode using a 440-mm field of view (FOV). Evaluated IQMs were spatial resolution, overall uniformity, subject contrast, contrast-to-noise ratio (CNR), and effective slice thickness. Imaging dose was assessed for planar kV imaging.

Results

On this kVCT system spatial resolution and contrast were consistent across all settings with 0.28 ± 0.03 lp/mm and 9.8% ± 0.7% (both 95% confidence interval). CNR strongly depended on selected mode (views per rotation) and body size (mA per view) and ranged between 7.9 and 34.9. Overall uniformity was greater than 97% for all settings. Large FOV was not found to substantially affect the IQMs whereas small FOV affected IQMs due to its effect on pitch. Technique-matched CT simulator scans were comparable for uniformity and contrast, while spatial resolution was higher (0.43 ± 0.06 lp/mm), and CNR was between 4% (140 kVp) and 51% (100 kVp) lower. For kV-CBCT, spatial resolutions ranging from 0.37 to 0.44 lp/mm were achieved with comparable contrast, CNR, and uniformity to kVCT. All kVCT scans exhibit imaging artifacts due to helical acquisition. Clinical acquisitions of megavoltage (MV) CT, kV-CBCT, and kVCT on the same patient showed improved and comparable image quality of kVCT compared to MVCT and kV-CBCT, respectively.

Conclusions

Helical fan beam kVCT allows for daily image guidance for localization and setup verification with comparable performance to existing kV-CBCT systems. Scan parameters must be selected carefully to maximize image quality for the desired tasks. Due to the large effective slice thicknesses for all parameter combinations, kVCT scans should not be used for simulation or planning of stereotactic procedures. Finally, improved image quality over MVCT has the potential to greatly improve manual and automated adaptive monitoring and planning.

Quantitative assessment of intra- and inter-modality deformable image registration of the heart, left ventricle, and thoracic aorta on longitudinal 4D-CT and MR images

Purpose

Magnetic resonance imaging (MRI)‐based investigations into radiotherapy (RT)‐induced cardiotoxicity require reliable registrations of magnetic resonance (MR) imaging to planning computed tomography (CT) for correlation to regional dose. In this study, the accuracy of intra‐ and inter‐modality deformable image registration (DIR) of longitudinal four‐dimensional CT (4D‐CT) and MR images were evaluated for heart, left ventricle (LV), and thoracic aorta (TA).

Methods and materials

Non‐cardiac‐gated 4D‐CT and T1 volumetric interpolated breath‐hold examination (T1‐VIBE) MRI datasets from five lung cancer patients were obtained at two breathing phases (inspiration/expiration) and two time points (before treatment and 5 weeks after initiating RT). Heart, LV, and TA were manually contoured. Each organ underwent three intramodal DIRs ((A) CT modality over time, (B) MR modality over time, and (C) MR contrast effect at the same time) and two intermodal DIRs ((D) CT/MR multimodality at same time and (E) CT/MR multimodality over time). Hausdorff distance (HD), mean distance to agreement (MDA), and Dice were evaluated and assessed for compliance with American Association of Physicists in Medicine (AAPM) Task Group (TG)‐132 recommendations.

Results

Mean values of HD, MDA, and Dice under all registration scenarios for each region of interest ranged between 8.7 and 16.8 mm, 1.0 and 2.6 mm, and 0.85 and 0.95, respectively, and were within the TG‐132 recommended range (MDA < 3 mm, Dice > 0.8). Intramodal DIR showed slightly better results compared to intermodal DIR. Heart and TA demonstrated higher registration accuracy compared to LV for all scenarios except for HD and Dice values in Group A. Significant differences for each metric and tissue of interest were noted between Groups B and D and between Groups B and E. MDA and Dice significantly differed between LV and heart in all registrations except for MDA in Group E.

Conclusions

DIR of the heart, LV, and TA between non‐cardiac‐gated longitudinal 4D‐CT and MRI across two modalities, breathing phases, and pre/post‐contrast is acceptably accurate per AAPM TG‐132 guidelines. This study paves the way for future evaluation of RT‐induced cardiotoxicity and its related factors using multimodality DIR.

Magnetic Resonance Imaging Images under Deep Learning in the Identification of Tuberculosis and Pneumonia

This work aimed to explore the application value of deep learning-based magnetic resonance imaging (MRI) images in the identification of tuberculosis and pneumonia, in order to provide a certain reference basis for clinical identification. In this study, 30 pulmonary tuberculosis patients and 27 pneumonia patients who were hospitalized were selected as the research objects, and they were divided into a pulmonary tuberculosis group and a pneumonia group. MRI examination based on noise reduction algorithms was used to observe and compare the signal-to-noise ratio (SNR) and carrier-to-noise ratio (CNR) of the images. In addition, the apparent diffusion coefficient (ADC) value for the diagnosis efficiency of lung parenchymal lesions was analyzed, and the best b value was selected. The results showed that the MRI image after denoising by the deep convolutional neural network (DCNN) algorithm was clearer, the edges of the lung tissue were regular, the inflammation signal was higher, and the SNR and CNR were better than before, which were 119.79 versus 83.43 and 12.59 versus 7.21, respectively. The accuracy of MRI based on a deep learning algorithm in the diagnosis of pulmonary tuberculosis and pneumonia was significantly improved (96.67% vs. 70%, 100% vs. 62.96%) (P < 0.05). With the increase in b value, the CNR and SNR of MRI images all showed a downward trend (P < 0.05). Therefore, it was found that the shadow of tuberculosis lesions under a specific sequence was higher than that of pneumonia in the process of identifying tuberculosis and pneumonia, which reflected the importance of deep learning MRI images in the differential diagnosis of tuberculosis and pneumonia, thereby providing reference basis for clinical follow-up diagnosis and treatment.

Comparison of rigid and deformable coregistration between mpMRI and CT images in radiotherapy of prostate bed cancer recurrence

PURPOSE

To evaluate the accuracy of rigid coregistration between multiparametric magnetic resonance (mpMR) and computed tomography (CT) images for radiotherapy of prostate bed cancer recurrence.

MATERIALS AND METHOD

Fifty-three patients (59 nodules) accrued in a prospective study on salvage radiotherapy for prostatic bed recurrence were suitable for the analysis. Patients underwent a pre radiotherapy mpMR exam and a planning CT in the same treatment position and with control of organ filling. The site of recurrence was delineated on mpMR images and contours transferred on planning CT images using both rigid and deformable registrations. Coregistrations were evaluated by mathematical operators that quantify deformation (Jacobian determinant and vector curl) and similarity indices (Dice and Jaccard coefficients). Dose coverage was evaluated.

RESULTS

Deformable registration did not change volumes, (p = 0.92 MW test). The Jacobian coefficient and the vector curl revealed no important image deformations. Dice and Jaccard coefficients indicated dislocation of the nodule volumes. Dislocation magnitude was d = (5.6 ± 3.1) mm. Organ filling was not correlated with deformation or dislocation. Volumes were covered by the 95% isodose in 96% of cases when rigid registration was performed versus 75% of cases when deformed.

CONCLUSIONS

Rigid image coregistration is sufficiently accurate in this setting. The results indicate that the deformable registration tends to shrink the voxels and to dislocate the ROI, the adopted expansion for the recurrence volume adequately accounts for the observed deformation and dislocation, provided that organ filling is controlled.

Evaluation of performance of pelvic CT-MR deformable image registration using two software programs

We assessed the accuracy of deformable image registration (DIR) accuracy between CT and MR images using an open-source software (Elastix, from Utrecht Medical Center) and a commercial software (Velocity AI Ver. 3.2.0 from Varian Medical Systems, Palo Alto, CA, USA) software. Five male patients' pelvic regions were studied using publicly available CT, T1-weighted (T1w) MR, and T2-weighted (T2w) MR images. In the cost function of the Elastix, we used six DIR parameter settings with different regularization weights (Elastix0, Elastix0.01, Elastix0.1, Elastix1, Elastix10, and Elastix100). We used MR Corrected Deformable algorithm for Velocity AI. The Dice similarity coefficient (DSC) and mean distance to agreement (MDA) for the prostate, bladder, rectum and left and right femoral heads were used to evaluate DIR accuracy. Except for the bladder, most algorithms produced good DSC and MDA results for all organs. In our study, the mean DSCs for the bladder ranged from 0.75 to 0.88 (CT-T1w) and from 0.72 to 0.76 (CT-T2w). Similarly, the mean MDA ranges were 2.4 to 4.9 mm (CT-T1w), 4.6 to 5.3 mm (CT-T2w). For the Elastix, CT-T1w was outperformed CT-T2w for both DSCs and MDAs at Elastix0, Elastix0.01, and Elastix0.1. In the case of Velocity AI, no significant differences in DSC and MDA of all organs were observed. This implied that the DIR accuracy of CT and MR images might differ depending on the sequence used.

Development of a physical geometric phantom for deformable image registration credentialing of radiotherapy centers for a clinical trial

PURPOSE

This study aimed to develop a physical geometric phantom for the deformable image registration (DIR) credentialing of radiotherapy centers for a clinical trial and tested the feasibility of the proposed phantom at multiple domestic and international institutions.

METHODS AND MATERIALS

The phantom reproduced tumor shrinkage, rectum shape change, and body shrinkage using several physical phantoms with custom inserts. We tested the feasibility of the proposed phantom using 5 DIR patterns at 17 domestic and 2 international institutions (21 datasets). Eight institutions used the MIM software (MIM Software Inc, Cleveland, OH); seven used Velocity (Varian Medical Systems, Palo Alto, CA), and six used RayStation (RaySearch Laboratories, Stockholm, Sweden). The DIR accuracy was evaluated using the Dice similarity coefficient (DSC) and Hausdorff distance (HD).

RESULTS

The mean and one standard deviation (SD) values (range) of DSC were 0.909 ± 0.088 (0.434-0.984) and 0.909 ± 0.048 (0.726-0.972) for tumor and rectum proxies, respectively. The mean and one SD values (range) of the HD value were 5.02 ± 3.32 (1.53-20.35) and 5.79 ± 3.47 (1.22-21.48) (mm) for the tumor and rectum proxies, respectively. In three patterns evaluating the DIR accuracy within the entire phantom, 61.9% of the data had more than a DSC of 0.8 in both tumor and rectum proxies. In two patterns evaluating the DIR accuracy by focusing on tumor and rectum proxies, all data had more than a DSC of 0.8 in both tumor and rectum proxies.

CONCLUSIONS

The wide range of DIR performance highlights the importance of optimizing the DIR process. Thus, the proposed method has considerable potential as an evaluation tool for DIR credentialing and quality assurance.