1
|
Cardiac MRI segmentation using shifted-window multilayer perceptron mixer networks. Phys Med Biol 2024. [PMID: 38744300 DOI: 10.1088/1361-6560/ad4b91] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 05/16/2024]
Abstract
PURPOSE In this work, we proposed a deep-learning segmentation algorithm for cardiac magnetic resonance imaging (MRI) to aid in contouring of the left ventricle (LV), right ventricle (RV), and Myocardium (Myo).
Methods: We proposed a shifted window multilayer perceptron (Swin-MLP) mixer network which is built upon a 3D U-shaped symmetric encoder-decoder structure. We evaluated our proposed network using public data from 100 individuals. The network performance was quantitatively evaluated using 3D volume similarity between the ground truth contours and the predictions using Dice score coefficient, sensitivity, and precision as well as 2D surface similarity using Hausdorff distance (HD), mean surface distance (MSD) and residual mean square distance (RMSD). We benchmarked the performance against two other current leading edge networks known as Dynamic UNet and Swin-UNetr on the same public dataset.
Results: The proposed network achieved the following volume similarity metrics when averaged over three cardiac segments: Dice = 0.952±0.017, precision = 0.948±0.016, sensitivity = 0.956±0.022. The average surface similarities were HD = 1.521±0.121 mm, MSD = 0.266±0.075 mm, and RMSD = 0.668±0.288 mm. The network shows statistically significant improvement in comparison to the Dynamic UNet and Swin-UNetr algorithms for most volumetric and surface metrics with p-value less than 0.05. Overall, the proposed Swin-MLP mixer network demonstrates better or comparable performance than competing methods.
Conclusions: The proposed Swin-MLP mixer network demonstrates more accurate segmentation performance compared to current leading edge methods. This robust method demonstrates the potential to streamline clinical workflows for multiple applications.
Collapse
|
2
|
Unsupervised MRI motion artifact disentanglement: introducing MAUDGAN. Phys Med Biol 2024. [PMID: 38714192 DOI: 10.1088/1361-6560/ad4845] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 05/09/2024]
Abstract
OBJECTIVE This study developed an unsupervised motion artifact reduction method for MRI images of patients with brain tumors. The proposed novel design uses multi-parametric multicenter contrast-enhanced T1W (ceT1W) and T2-FLAIR MRI images.
Approach: The proposed framework included two generators, two discriminators, and two feature extractor networks. A 3-fold cross-validation was used to train and fine-tune the hyperparameters of the proposed model using 230 brain MRI images with tumors, which were then tested on 148 patients' in-vivo datasets. An ablation was performed to evaluate the model's compartments. Our model was compared with Pix2pix and CycleGAN. Six evaluation metrics were reported, including normalized mean squared error (NMSE), structural similarity index (SSIM), multi-scale-SSIM (MS-SSIM), peak signal-to-noise ratio (PSNR), visual information fidelity (VIF), and multi-scale gradient magnitude similarity deviation (MS-GMSD). Artifact reduction and consistency of tumor regions, image contrast, and sharpness were evaluated by three evaluators using Likert scales and compared with ANOVA and Tukey's HSD tests.
Main results: On average, our method outperforms comparative models to remove heavy motion artifacts with the lowest NMSE (18.34±5.07%) and MS-GMSD (0.07±0.03) for heavy motion artifact level. Additionally, our method creates motion-free images with the highest SSIM (0.93±0.04), PSNR (30.63±4.96), and VIF (0.45±0.05) values, along with comparable MS-SSIM (0.96±0.31). Similarly, our method outperformed comparative models in removing in-vivo motion artifacts for different distortion levels except for MS- SSIM and VIF, which have comparable performance with CycleGAN. Moreover, our method had a consistent performance for different artifact levels. For the heavy level of motion artifacts, our method got the highest Likert scores of 2.82±0.52, 1.88±0.71, and 1.02±0.14 (p-values<<0.0001) for our method, CycleGAN, and Pix2pix respectively. Similar trends were also found for other motion artifact levels.
Significance: Our proposed unsupervised method was demonstrated to reduce motion artifacts from the ceT1W brain images under a multi-parametric framework.
Collapse
|
3
|
Advancing medical imaging with language models: featuring a spotlight on ChatGPT. Phys Med Biol 2024; 69:10TR01. [PMID: 38537293 DOI: 10.1088/1361-6560/ad387d] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/16/2023] [Accepted: 03/27/2024] [Indexed: 05/04/2024]
Abstract
This review paper aims to serve as a comprehensive guide and instructional resource for researchers seeking to effectively implement language models in medical imaging research. First, we presented the fundamental principles and evolution of language models, dedicating particular attention to large language models. We then reviewed the current literature on how language models are being used to improve medical imaging, emphasizing a range of applications such as image captioning, report generation, report classification, findings extraction, visual question response systems, interpretable diagnosis and so on. Notably, the capabilities of ChatGPT were spotlighted for researchers to explore its further applications. Furthermore, we covered the advantageous impacts of accurate and efficient language models in medical imaging analysis, such as the enhancement of clinical workflow efficiency, reduction of diagnostic errors, and assistance of clinicians in providing timely and accurate diagnoses. Overall, our goal is to have better integration of language models with medical imaging, thereby inspiring new ideas and innovations. It is our aspiration that this review can serve as a useful resource for researchers in this field, stimulating continued investigative and innovative pursuits of the application of language models in medical imaging.
Collapse
|
4
|
Synthetic CT generation from MRI using 3D transformer-based denoising diffusion model. Med Phys 2024; 51:2538-2548. [PMID: 38011588 PMCID: PMC10994752 DOI: 10.1002/mp.16847] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/28/2023] [Revised: 11/02/2023] [Accepted: 11/03/2023] [Indexed: 11/29/2023] Open
Abstract
BACKGROUND AND PURPOSE Magnetic resonance imaging (MRI)-based synthetic computed tomography (sCT) simplifies radiation therapy treatment planning by eliminating the need for CT simulation and error-prone image registration, ultimately reducing patient radiation dose and setup uncertainty. In this work, we propose a MRI-to-CT transformer-based improved denoising diffusion probabilistic model (MC-IDDPM) to translate MRI into high-quality sCT to facilitate radiation treatment planning. METHODS MC-IDDPM implements diffusion processes with a shifted-window transformer network to generate sCT from MRI. The proposed model consists of two processes: a forward process, which involves adding Gaussian noise to real CT scans to create noisy images, and a reverse process, in which a shifted-window transformer V-net (Swin-Vnet) denoises the noisy CT scans conditioned on the MRI from the same patient to produce noise-free CT scans. With an optimally trained Swin-Vnet, the reverse diffusion process was used to generate noise-free sCT scans matching MRI anatomy. We evaluated the proposed method by generating sCT from MRI on an institutional brain dataset and an institutional prostate dataset. Quantitative evaluations were conducted using several metrics, including Mean Absolute Error (MAE), Peak Signal-to-Noise Ratio (PSNR), Multi-scale Structure Similarity Index (SSIM), and Normalized Cross Correlation (NCC). Dosimetry analyses were also performed, including comparisons of mean dose and target dose coverages for 95% and 99%. RESULTS MC-IDDPM generated brain sCTs with state-of-the-art quantitative results with MAE 48.825 ± 21.491 HU, PSNR 26.491 ± 2.814 dB, SSIM 0.947 ± 0.032, and NCC 0.976 ± 0.019. For the prostate dataset: MAE 55.124 ± 9.414 HU, PSNR 28.708 ± 2.112 dB, SSIM 0.878 ± 0.040, and NCC 0.940 ± 0.039. MC-IDDPM demonstrates a statistically significant improvement (with p < 0.05) in most metrics when compared to competing networks, for both brain and prostate synthetic CT. Dosimetry analyses indicated that the target dose coverage differences by using CT and sCT were within ± 0.34%. CONCLUSIONS We have developed and validated a novel approach for generating CT images from routine MRIs using a transformer-based improved DDPM. This model effectively captures the complex relationship between CT and MRI images, allowing for robust and high-quality synthetic CT images to be generated in a matter of minutes. This approach has the potential to greatly simplify the treatment planning process for radiation therapy by eliminating the need for additional CT scans, reducing the amount of time patients spend in treatment planning, and enhancing the accuracy of treatment delivery.
Collapse
|
5
|
CBCT-Based synthetic CT image generation using conditional denoising diffusion probabilistic model. Med Phys 2024; 51:1847-1859. [PMID: 37646491 DOI: 10.1002/mp.16704] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/08/2023] [Revised: 07/17/2023] [Accepted: 08/08/2023] [Indexed: 09/01/2023] Open
Abstract
BACKGROUND Daily or weekly cone-beam computed tomography (CBCT) scans are commonly used for accurate patient positioning during the image-guided radiotherapy (IGRT) process, making it an ideal option for adaptive radiotherapy (ART) replanning. However, the presence of severe artifacts and inaccurate Hounsfield unit (HU) values prevent its use for quantitative applications such as organ segmentation and dose calculation. To enable the clinical practice of online ART, it is crucial to obtain CBCT scans with a quality comparable to that of a CT scan. PURPOSE This work aims to develop a conditional diffusion model to perform image translation from the CBCT to the CT distribution for the image quality improvement of CBCT. METHODS The proposed method is a conditional denoising diffusion probabilistic model (DDPM) that utilizes a time-embedded U-net architecture with residual and attention blocks to gradually transform the white Gaussian noise sample to the target CT distribution conditioned on the CBCT. The model was trained on deformed planning CT (dpCT) and CBCT image pairs, and its feasibility was verified in brain patient study and head-and-neck (H&N) patient study. The performance of the proposed algorithm was evaluated using mean absolute error (MAE), peak signal-to-noise ratio (PSNR) and normalized cross-correlation (NCC) metrics on generated synthetic CT (sCT) samples. The proposed method was also compared to four other diffusion model-based sCT generation methods. RESULTS In the brain patient study, the MAE, PSNR, and NCC of the generated sCT were 25.99 HU, 30.49 dB, and 0.99, respectively, compared to 40.63 HU, 27.87 dB, and 0.98 of the CBCT images. In the H&N patient study, the metrics were 32.56 HU, 27.65 dB, 0.98 and 38.99 HU, 27.00, 0.98 for sCT and CBCT, respectively. Compared to the other four diffusion models and one Cycle generative adversarial network (Cycle GAN), the proposed method showed superior results in both visual quality and quantitative analysis. CONCLUSIONS The proposed conditional DDPM method can generate sCT from CBCT with accurate HU numbers and reduced artifacts, enabling accurate CBCT-based organ segmentation and dose calculation for online ART.
Collapse
|
6
|
High-resolution MRI synthesis using a data-driven framework with denoising diffusion probabilistic modeling. Phys Med Biol 2024; 69:045001. [PMID: 38241726 PMCID: PMC10839468 DOI: 10.1088/1361-6560/ad209c] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/20/2023] [Revised: 01/08/2024] [Accepted: 01/19/2024] [Indexed: 01/21/2024]
Abstract
Objective. High-resolution magnetic resonance imaging (MRI) can enhance lesion diagnosis, prognosis, and delineation. However, gradient power and hardware limitations prohibit recording thin slices or sub-1 mm resolution. Furthermore, long scan time is not clinically acceptable. Conventional high-resolution images generated using statistical or analytical methods include the limitation of capturing complex, high-dimensional image data with intricate patterns and structures. This study aims to harness cutting-edge diffusion probabilistic deep learning techniques to create a framework for generating high-resolution MRI from low-resolution counterparts, improving the uncertainty of denoising diffusion probabilistic models (DDPM).Approach. DDPM includes two processes. The forward process employs a Markov chain to systematically introduce Gaussian noise to low-resolution MRI images. In the reverse process, a U-Net model is trained to denoise the forward process images and produce high-resolution images conditioned on the features of their low-resolution counterparts. The proposed framework was demonstrated using T2-weighted MRI images from institutional prostate patients and brain patients collected in the Brain Tumor Segmentation Challenge 2020 (BraTS2020).Main results. For the prostate dataset, the bicubic interpolation model (Bicubic), conditional generative-adversarial network (CGAN), and our proposed DDPM framework improved the noise quality measure from low-resolution images by 4.4%, 5.7%, and 12.8%, respectively. Our method enhanced the signal-to-noise ratios by 11.7%, surpassing Bicubic (9.8%) and CGAN (8.1%). In the BraTS2020 dataset, the proposed framework and Bicubic enhanced peak signal-to-noise ratio from resolution-degraded images by 9.1% and 5.8%. The multi-scale structural similarity indexes were 0.970 ± 0.019, 0.968 ± 0.022, and 0.967 ± 0.023 for the proposed method, CGAN, and Bicubic, respectively.Significance. This study explores a deep learning-based diffusion probabilistic framework for improving MR image resolution. Such a framework can be used to improve clinical workflow by obtaining high-resolution images without penalty of the long scan time. Future investigation will likely focus on prospectively testing the efficacy of this framework with different clinical indications.
Collapse
|
7
|
Deep learning in MRI-guided radiation therapy: A systematic review. J Appl Clin Med Phys 2024; 25:e14155. [PMID: 37712893 PMCID: PMC10860468 DOI: 10.1002/acm2.14155] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/21/2023] [Revised: 05/10/2023] [Accepted: 08/21/2023] [Indexed: 09/16/2023] Open
Abstract
Recent advances in MRI-guided radiation therapy (MRgRT) and deep learning techniques encourage fully adaptive radiation therapy (ART), real-time MRI monitoring, and the MRI-only treatment planning workflow. Given the rapid growth and emergence of new state-of-the-art methods in these fields, we systematically review 197 studies written on or before December 31, 2022, and categorize the studies into the areas of image segmentation, image synthesis, radiomics, and real time MRI. Building from the underlying deep learning methods, we discuss their clinical importance and current challenges in facilitating small tumor segmentation, accurate x-ray attenuation information from MRI, tumor characterization and prognosis, and tumor motion tracking. In particular, we highlight the recent trends in deep learning such as the emergence of multi-modal, visual transformer, and diffusion models.
Collapse
|
8
|
IODINE-125 PLAQUE BRACHYTHERAPY FOR DIFFUSE CHOROIDAL HEMANGIOMA. Retin Cases Brief Rep 2024; 18:51-58. [PMID: 36007192 DOI: 10.1097/icb.0000000000001334] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 06/15/2023]
Abstract
PURPOSE To report 6 cases of diffuse choroidal hemangioma in children treated with iodine-125 plaque brachytherapy at a single tertiary care center. METHODS Retrospective case series. RESULTS Six pediatric patients diagnosed with diffuse choroidal hemangioma were included in the study. Preplaque visual acuity ranged from 20/150 to no light perception. All patients had extensive serous retinal detachment at presentation. An iodine-125 radioactive plaque was placed on the affected eye to administer a dose of 34.2-42.1 Gy to the tumor apex over a median of 4 days. Tumor regression and subretinal fluid resolution were observed in all eyes within 17 months of treatment. Visual acuity improved in two patients. Radiation-induced cataract and subretinal fibrosis were documented in one case, and one patient developed radiation retinopathy. No patients developed neovascular glaucoma within the follow-up time of 12-65 months. CONCLUSION Iodine-125 plaque radiotherapy is an effective option for diffuse choroidal hemangioma, although there is a risk for radiation-induced complications.
Collapse
|
9
|
MGMT promoter methylation prediction based on multiparametric MRI via vision graph neural network. J Med Imaging (Bellingham) 2024; 11:014503. [PMID: 38370421 PMCID: PMC10869845 DOI: 10.1117/1.jmi.11.1.014503] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/30/2023] [Revised: 12/24/2023] [Accepted: 01/29/2024] [Indexed: 02/20/2024] Open
Abstract
Purpose Glioblastoma (GBM) is aggressive and malignant. The methylation status of the O 6 -methylguanine-DNA methyltransferase (MGMT) promoter in GBM tissue is considered an important biomarker for developing the most effective treatment plan. Although the standard method for assessing the MGMT promoter methylation status is via bisulfite modification and deoxyribonucleic acid (DNA) sequencing of biopsy or surgical specimens, a secondary automated method based on medical imaging may improve the efficiency and accuracy of those tests. Approach We propose a deep vision graph neural network (ViG) using multiparametric magnetic resonance imaging (MRI) to predict the MGMT promoter methylation status noninvasively. Our model was compared to the RSNA radiogenomic classification winners. The dataset includes 583 usable patient cases. Combinations of MRI sequences were compared. Our multi-sequence fusion strategy was compared with those using single MR sequences. Results Our best model [Fluid Attenuated Inversion Recovery (FLAIR), T1-weighted pre-contrast (T1w), T2-weighted (T2)] outperformed the winning models with a test area under the curve (AUC) of 0.628, an accuracy of 0.632, a precision of 0.646, a recall of 0.677, a specificity of 0.581, and an F1 score of 0.661. Compared to the winning models with single MR sequences, our ViG utilizing fused-MRI showed a significant improvement statistically in AUC scores, which are FLAIR (p = 0.042 ), T1w (p = 0.017 ), T1wCE (p = 0.001 ), and T2 (p = 0.018 ). Conclusions Our model is superior to challenge champions. A graph representation of the medical images enabled good handling of complexity and irregularity. Our work provides an automatic secondary check pipeline to ensure the correctness of MGMT methylation status prediction.
Collapse
|
10
|
Hippocampus substructure segmentation using morphological vision transformer learning. Phys Med Biol 2023; 68:235013. [PMID: 37972414 PMCID: PMC10690959 DOI: 10.1088/1361-6560/ad0d45] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/12/2023] [Revised: 11/01/2023] [Accepted: 11/15/2023] [Indexed: 11/19/2023]
Abstract
The hippocampus plays a crucial role in memory and cognition. Because of the associated toxicity from whole brain radiotherapy, more advanced treatment planning techniques prioritize hippocampal avoidance, which depends on an accurate segmentation of the small and complexly shaped hippocampus. To achieve accurate segmentation of the anterior and posterior regions of the hippocampus from T1 weighted (T1w) MR images, we developed a novel model, Hippo-Net, which uses a cascaded model strategy. The proposed model consists of two major parts: (1) a localization model is used to detect the volume-of-interest (VOI) of hippocampus. (2) An end-to-end morphological vision transformer network (Franchietal2020Pattern Recognit.102107246, Ranemetal2022 IEEE/CVF Conf. on Computer Vision and Pattern Recognition Workshops (CVPRW) pp 3710-3719) is used to perform substructures segmentation within the hippocampus VOI. The substructures include the anterior and posterior regions of the hippocampus, which are defined as the hippocampus proper and parts of the subiculum. The vision transformer incorporates the dominant features extracted from MR images, which are further improved by learning-based morphological operators. The integration of these morphological operators into the vision transformer increases the accuracy and ability to separate hippocampus structure into its two distinct substructures. A total of 260 T1w MRI datasets from medical segmentation decathlon dataset were used in this study. We conducted a five-fold cross-validation on the first 200 T1w MR images and then performed a hold-out test on the remaining 60 T1w MR images with the model trained on the first 200 images. In five-fold cross-validation, the Dice similarity coefficients were 0.900 ± 0.029 and 0.886 ± 0.031 for the hippocampus proper and parts of the subiculum, respectively. The mean surface distances (MSDs) were 0.426 ± 0.115 mm and 0.401 ± 0.100 mm for the hippocampus proper and parts of the subiculum, respectively. The proposed method showed great promise in automatically delineating hippocampus substructures on T1w MR images. It may facilitate the current clinical workflow and reduce the physicians' effort.
Collapse
|
11
|
Image-Domain Material Decomposition for Dual-energy CT using Unsupervised Learning with Data-fidelity Loss. ARXIV 2023:arXiv:2311.10641v1. [PMID: 38013889 PMCID: PMC10680906] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Subscribe] [Scholar Register] [Indexed: 11/29/2023]
Abstract
BACKGROUND Dual-energy CT (DECT) and material decomposition play vital roles in quantitative medical imaging. However, the decomposition process may suffer from significant noise amplification, leading to severely degraded image signal-to-noise ratios (SNRs). While existing iterative algorithms perform noise suppression using different image priors, these heuristic image priors cannot accurately represent the features of the target image manifold. Although deep learning-based decomposition methods have been reported, these methods are in the supervised-learning framework requiring paired data for training, which is not readily available in clinical settings. PURPOSE This work aims to develop an unsupervised-learning framework with data-measurement consistency for image-domain material decomposition in DECT.
Collapse
|
12
|
A Novel Brachytherapy Needle Design to Reduce Urethra Dose in HDR Prostate Brachytherapy. Int J Radiat Oncol Biol Phys 2023; 117:e415-e416. [PMID: 37785372 DOI: 10.1016/j.ijrobp.2023.06.1565] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/04/2023]
Abstract
PURPOSE/OBJECTIVE(S) Ultrasound-guided high-dose-rate (HDR) prostate brachytherapy is a safe and effective treatment option for prostate cancer patients; however, some patients still experience acute and late genitourinary (GU) toxicity. It has been reported that urethra dose is associated with the incidence and severity of GU toxicity. Therefore, a technique that can further spare the urethra while ensuring adequate target coverage is highly desirable. Intensity modulated brachytherapy (IMBT) designs, such as rotating shield brachytherapy (RSBT), offer ideal dosimetry on-paper but are challenging to implement clinically due to the need for high precision in moving mechanisms and the use of non-conventional HDR sources, such as 153Gd, electronic, and 169Yb. In this study, we propose a novel clinically implementable solution based on the direction modulated brachytherapy (DMBT) design concept, which has no moving parts and works effectively with the commercially available 192Ir sources. MATERIALS/METHODS The widely used GammaMedPlus (GMP) 192Ir sources, with outer diameters of 0.9 mm, was simulated using the GEANT4 Monte Carlo (MC) simulation code. The novel DMBT needle concept consists of a 14-gauge nitinol needle with an outer diameter of 2.1 mm and a crust thickness of 0.1 mm, which housed a platinum shield inside. A single groove, consistent with the outer diameter of each source, was incorporated inside the platinum shield to accommodate the HDR source. The maximum thickness of the shield was 0.8 mm. To evaluate the effectiveness of the DMBT needle concept in reducing urethra dose, two proof-of-concept in silico DMBT plans were created from the standard clinical plan by replacing two needles close to the urethra with the DMBT needles. The dosimetrical comparisons between the two DMBT plans, and the clinical BT plan were done by assessing the target coverage, and urethral, bladder and rectum dose. RESULTS The MC results showed that the use of the novel DMBT needle could reduce the radiation dose by approximately 35% at 1 cm distance from the needle behind the platinum shield, as compared to the unshielded side. Additionally, when using the same dose volume histogram (DVH) planning criteria as the original plan, the DMBT plan reduced the maximum urethral dose at the pre-apical region by 16% and 28% for 0 mm and 2 mm planning organ at risk (PRV) margins, respectively, while maintaining equivalent V100 and D90 target coverage. CONCLUSION The novel DMBT needle design offers a promising solution for reducing the dose to the pre-apical region of the urethra without compromising target coverage, increasing the dose to other organs-at-risk, or increasing the treatment time.
Collapse
|
13
|
2D medical image synthesis using transformer-based denoising diffusion probabilistic model. Phys Med Biol 2023; 68. [PMID: 37015231 PMCID: PMC10160739 DOI: 10.1088/1361-6560/acca5c] [Citation(s) in RCA: 8] [Impact Index Per Article: 8.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/11/2023] [Accepted: 04/04/2023] [Indexed: 04/06/2023]
Abstract
OBJECTIVE Artificial intelligence (AI) methods have gained popularity in medical imaging research. The size and scope of the training image datasets needed for successful AI model deployment does not always have the desired scale. In this paper, we introduce a medical image synthesis framework aimed at addressing the challenge of limited training datasets for AI models.

Approach: The proposed 2D image synthesis framework is based on a diffusion model using a Swin-transformer-based network. This model consists of a forward Gaussian noise process and a reverse process using the Transformer-based Diffusion model for denoising. Training data includes four image datasets: chest X-rays, heart MRI, pelvic CT, and abdomen CT. We evaluated the authenticity, quality, and diversity of the synthetic images using visual Turing assessments conducted by three medical physicists, and four quantitative evaluations: the Inception score (IS), Fréchet Inception Distance score (FID), feature similarity and diversity score (DS, indicating diversity similarity) between the synthetic and true images. To leverage the framework value for training AI models, we conducted COVID-19 classification tasks using real images, synthetic images, and mixtures of both images. 

Main results: Visual Turing assessments showed an average accuracy of 0.64 (accuracy converging to 50% indicates a better realistic visual appearance of the synthetic images), sensitivity of 0.79, and specificity of 0.50. Average quantitative accuracy obtained from all datasets were IS=2.28, FID=37.27, FDS=0.20, and DS=0.86. For the COVID-19 classification task, the baseline network obtained an accuracy of 0.88 using a pure real dataset, 0.89 using a pure synthetic dataset, and 0.93 using a dataset mixed of real and synthetic data.

Significance: A image synthesis framework was demonstrated for medical image synthesis, which can generate high-quality medical images of different imaging modalities with the purpose of supplementing existing training sets for AI model deployment. This method has potential applications in many data-driven medical imaging research.
.
Collapse
|
14
|
Deep learning-based thoracic CBCT correction with histogram matching. Biomed Phys Eng Express 2021; 7. [PMID: 34654011 DOI: 10.1088/2057-1976/ac3055] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/03/2021] [Accepted: 10/15/2021] [Indexed: 12/25/2022]
Abstract
Kilovoltage cone-beam computed tomography (CBCT)-based image-guided radiation therapy (IGRT) is used for daily delivery of radiation therapy, especially for stereotactic body radiation therapy (SBRT), which imposes particularly high demands for setup accuracy. The clinical applications of CBCTs are constrained, however, by poor soft tissue contrast, image artifacts, and instability of Hounsfield unit (HU) values. Here, we propose a new deep learning-based method to generate synthetic CTs (sCT) from thoracic CBCTs. A deep-learning model which integrates histogram matching (HM) into a cycle-consistent adversarial network (Cycle-GAN) framework, called HM-Cycle-GAN, was trained to learn mapping between thoracic CBCTs and paired planning CTs. Perceptual supervision was adopted to minimize blurring of tissue interfaces. An informative maximizing loss was calculated by feeding CBCT into the HM-Cycle-GAN to evaluate the image histogram matching between the planning CTs and the sCTs. The proposed algorithm was evaluated using data from 20 SBRT patients who each received 5 fractions and therefore 5 thoracic CBCTs. To reduce the effect of anatomy mismatch, original CBCT images were pre-processed via deformable image registrations with the planning CT before being used in model training and result assessment. We used planning CTs as ground truth for the derived sCTs from the correspondent co-registered CBCTs. The mean absolute error (MAE), peak signal-to-noise ratio (PSNR), and normalized cross-correlation (NCC) indices were adapted as evaluation metrics of the proposed algorithm. Assessments were done using Cycle-GAN as the benchmark. The average MAE, PSNR, and NCC of the sCTs generated by our method were 66.2 HU, 30.3 dB, and 0.95, respectively, over all CBCT fractions. Superior image quality and reduced noise and artifact severity were seen using the proposed method compared to the results from the standard Cycle-GAN method. Our method could therefore improve the accuracy of IGRT and corrected CBCTs could help improve online adaptive RT by offering better contouring accuracy and dose calculation.
Collapse
|
15
|
Automatic multi-catheter detection using deeply supervised convolutional neural network in MRI-guided HDR prostate brachytherapy. Med Phys 2020; 47:4115-4124. [PMID: 32484573 DOI: 10.1002/mp.14307] [Citation(s) in RCA: 17] [Impact Index Per Article: 4.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/19/2020] [Revised: 05/19/2020] [Accepted: 05/24/2020] [Indexed: 12/19/2022] Open
Abstract
PURPOSE High-dose-rate (HDR) brachytherapy is an established technique to be used as monotherapy option or focal boost in conjunction with external beam radiation therapy (EBRT) for treating prostate cancer. Radiation source path reconstruction is a critical procedure in HDR treatment planning. Manually identifying the source path is labor intensive and time inefficient. In recent years, magnetic resonance imaging (MRI) has become a valuable imaging modality for image-guided HDR prostate brachytherapy due to its superb soft-tissue contrast for target delineation and normal tissue contouring. The purpose of this study is to investigate a deep-learning-based method to automatically reconstruct multiple catheters in MRI for prostate cancer HDR brachytherapy treatment planning. METHODS Attention gated U-Net incorporated with total variation (TV) regularization model was developed for multi-catheter segmentation in MRI. The attention gates were used to improve the accuracy of identifying small catheter points, while TV regularization was adopted to encode the natural spatial continuity of catheters into the model. The model was trained using the binary catheter annotation images offered by experienced physicists as ground truth paired with original MRI images. After the network was trained, MR images of a new prostate cancer patient receiving HDR brachytherapy were fed into the model to predict the locations and shapes of all the catheters. Quantitative assessments of our proposed method were based on catheter shaft and tip errors compared to the ground truth. RESULTS Our method detected 299 catheters from 20 patients receiving HDR prostate brachytherapy with a catheter tip error of 0.37 ± 1.68 mm and a catheter shaft error of 0.93 ± 0.50 mm. For detection of catheter tips, our method resulted in 87% of the catheter tips within an error of less than ± 2.0 mm, and more than 71% of the tips can be localized within an absolute error of no >1.0 mm. For catheter shaft localization, 97% of catheters were detected with an error of <2.0 mm, while 63% were within 1.0 mm. CONCLUSIONS In this study, we proposed a novel multi-catheter detection method to precisely localize the tips and shafts of catheters in three-dimensional MRI images of HDR prostate brachytherapy. It paves the way for elevating the quality and outcome of MRI-guided HDR prostate brachytherapy.
Collapse
|
16
|
Development of topological insulator and topological crystalline insulator nanostructures. NANOTECHNOLOGY 2020; 31:192001. [PMID: 31962300 DOI: 10.1088/1361-6528/ab6dfc] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 06/10/2023]
Abstract
Topological insulators (TIs), a class of quantum materials with time reversal symmetry protected gapless Dirac-surface states, have attracted intensive research interests due to their exotic electronic properties. Topological crystalline insulators (TCIs), whose gapless surface states are protected by the crystal symmetry, have recently been proposed and experimentally verified as a new class of TIs. With high surface-to-volume ratio, nanoscale TI and TCI materials such as nanowires and nanoribbons can have significantly enhanced contribution from surface states in carrier transport and are thus ideally suited for the fundamental studies of topologically protected surface state transport and nanodevice fabrication. This article will review the synthesis and transport device measurements of TIs and TCIs nanostructures.
Collapse
|
17
|
Technical Note: A step-by-step guide to Temporally Feathered Radiation Therapy planning for head and neck cancer. J Appl Clin Med Phys 2020; 21:209-215. [PMID: 32383296 PMCID: PMC7386183 DOI: 10.1002/acm2.12893] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/31/2019] [Revised: 03/20/2020] [Accepted: 04/06/2020] [Indexed: 12/26/2022] Open
Abstract
Purpose Prior in silico simulations propose that Temporally Feathered Radiation Therapy (TFRT) may reduce toxicity related to head and neck radiation therapy. In this study we demonstrate a step‐by‐step guide to TFRT planning with modern treatment planning systems. Methods One patient with oropharyngeal cancer planned for definitive radiation therapy using intensity‐modulated radiation therapy (IMRT) techniques was replanned using the TFRT technique. Five organs at risk (OAR) were identified to be feathered. A “base plan” was first created based on desired planning target volumes (PTV) coverage, plan conformality, and OAR constraints. The base plan was then re‐optimized by modifying planning objectives, to generate five subplans. All beams from each subplan were imported onto one trial to create the composite TFRT plan. The composite TFRT plan was directly compared with the non‐TFRT IMRT plan. During plan assessment, the composite TFRT was first evaluated followed by each subplan to meet preset compliance criteria. Results The following organs were feathered: oral cavity, right submandibular gland, left submandibular gland, supraglottis, and OAR Pharynx. Prescription dose PTV coverage (>95%) was met in each subplan and the composite TFRT plan. Expected small variations in dose were observed among the plans. The percent variation between the high fractional dose and average low fractional dose was 29%, 28%, 24%, 19%, and 10% for the oral cavity, right submandibular, left submandibular, supraglottis, and OAR pharynx nonoverlapping with the PTV. Conclusions Temporally Feathered Radiation Therapy planning is possible with modern treatment planning systems. Modest dosimetric changes are observed with TFRT planning compared with non‐TFRT IMRT planning. We await the results of the current prospective trial to seeking to demonstrate the feasibility of TFRT in the modern clinical workflow (NCT03768856). Further studies will be required to demonstrate the potential benefit of TFRT over non‐TFRT IMRT Planning.
Collapse
|
18
|
Multi-needle Localization with Attention U-Net in US-guided HDR Prostate Brachytherapy. Med Phys 2020; 47:2735-2745. [PMID: 32155666 DOI: 10.1002/mp.14128] [Citation(s) in RCA: 23] [Impact Index Per Article: 5.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/13/2019] [Revised: 02/17/2020] [Accepted: 03/04/2020] [Indexed: 12/11/2022] Open
Abstract
PURPOSE Ultrasound (US)-guided high dose rate (HDR) prostate brachytherapy requests the clinicians to place HDR needles (catheters) into the prostate gland under transrectal US (TRUS) guidance in the operating room. The quality of the subsequent radiation treatment plan is largely dictated by the needle placements, which varies upon the experience level of the clinicians and the procedure protocols. Real-time plan dose distribution, if available, could be a vital tool to provide more subjective assessment of the needle placements, hence potentially improving the radiation plan quality and the treatment outcome. However, due to low signal-to-noise ratio (SNR) in US imaging, real-time multi-needle segmentation in 3D TRUS, which is the major obstacle for real-time dose mapping, has not been realized to date. In this study, we propose a deep learning-based method that enables accurate and real-time digitization of the multiple needles in the 3D TRUS images of HDR prostate brachytherapy. METHODS A deep learning model based on the U-Net architecture was developed to segment multiple needles in the 3D TRUS images. Attention gates were considered in our model to improve the prediction on the small needle points. Furthermore, the spatial continuity of needles was encoded into our model with total variation (TV) regularization. The combined network was trained on 3D TRUS patches with the deep supervision strategy, where the binary needle annotation images were provided as ground truth. The trained network was then used to localize and segment the HDR needles for a new patient's TRUS images. We evaluated our proposed method based on the needle shaft and tip errors against manually defined ground truth and compared our method with other state-of-art methods (U-Net and deeply supervised attention U-Net). RESULTS Our method detected 96% needles of 339 needles from 23 HDR prostate brachytherapy patients with 0.290 ± 0.236 mm at shaft error and 0.442 ± 0.831 mm at tip error. For shaft localization, our method resulted in 96% localizations with less than 0.8 mm error (needle diameter is 1.67 mm), while for tip localization, our method resulted in 75% needles with 0 mm error and 21% needles with 2 mm error (TRUS image slice thickness is 2 mm). No significant difference is observed (P = 0.83) on tip localization between our results with the ground truth. Compared with U-Net and deeply supervised attention U-Net, the proposed method delivers a significant improvement on both shaft error and tip error (P < 0.05). CONCLUSIONS We proposed a new segmentation method to precisely localize the tips and shafts of multiple needles in 3D TRUS images of HDR prostate brachytherapy. The 3D rendering of the needles could help clinicians to evaluate the needle placements. It paves the way for the development of real-time plan dose assessment tools that can further elevate the quality and outcome of HDR prostate brachytherapy.
Collapse
|
19
|
Abstract
A reply to the comment by A. Yu. Kuntsevich and V. M. Pudalov.
Collapse
|
20
|
Metal-insulator transition in variably doped (Bi(1-x)Sb(x))2Se3 nanosheets. NANOSCALE 2013; 5:4337-4343. [PMID: 23563061 DOI: 10.1039/c3nr01155k] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 06/02/2023]
Abstract
Topological insulators are novel quantum materials with metallic surface transport but insulating bulk behavior. Often, topological insulators are dominated by bulk contributions due to defect induced bulk carriers, making it difficult to isolate the more interesting surface transport characteristics. Here, we report the synthesis and characterization of nanosheets of a topological insulator Bi2Se3 with variable Sb-doping levels to control the electron carrier density and surface transport behavior. (Bi(1-x)Sb(x))2Se3 thin films of thickness less than 10 nm are prepared by epitaxial growth on mica substrates in a vapor transport setup. The introduction of Sb in Bi2Se3 effectively suppresses the room temperature electron density from ∼4 × 10(13) cm(-2) in pure Bi2Se3 (x = 0) to ∼2 × 10(12) cm(-2) in (Bi(1-x)Sb(x))2Se3 at x ∼ 0.15, while maintaining the metallic transport behavior. At x ≳ ∼0.20, a metal-insulator transition (MIT) is observed, indicating that the system has transformed into an insulator in which the metallic surface conduction is blocked. In agreement with the observed MIT, Raman spectroscopy reveals the emergence of vibrational modes arising from Sb-Sb and Sb-Se bonds at high Sb concentrations, confirming the appearance of the Sb2Se3 crystal structure in the sample. These results suggest that nanostructured chalcogenide films with controlled doping can be a tunable platform for fundamental studies and electronic applications of topological insulator systems.
Collapse
|
21
|
Ambipolar surface conduction in ternary topological insulator Bi₂(Te₁-xSex)₃ nanoribbons. ACS NANO 2013; 7:2126-2131. [PMID: 23441571 DOI: 10.1021/nn304684b] [Citation(s) in RCA: 5] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 06/01/2023]
Abstract
We report the composition- and gate voltage-induced tuning of transport properties in chemically synthesized Bi2(Te1-xSex)3 nanoribbons. It is found that increasing Se concentration effectively suppresses the bulk carrier transport and induces semiconducting behavior in the temperature-dependent resistance of Bi2(Te1-xSex)3 nanoribbons when x is greater than ∼10%. In Bi2(Te1-xSex)3 nanoribbons with x ≈ 20%, gate voltage enables ambipolar modulation of resistance (or conductance) in samples with thicknesses around or larger than 100 nm, indicating significantly enhanced contribution in transport from the gapless surface states.
Collapse
|
22
|
One-dimensional quantum confinement effect modulated thermoelectric properties in InAs nanowires. NANO LETTERS 2012; 12:6492-6497. [PMID: 23167670 DOI: 10.1021/nl304194c] [Citation(s) in RCA: 25] [Impact Index Per Article: 2.1] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 06/01/2023]
Abstract
We report electrical conductance and thermopower measurements on InAs nanowires synthesized by chemical vapor deposition. Gate modulation of the thermopower of individual InAs nanowires with a diameter around 20 nm is obtained over T = 40-300 K. At low temperatures (T < ∼100 K), oscillations in the thermopower and power factor concomitant with the stepwise conductance increases are observed as the gate voltage shifts the chemical potential of electrons in InAs nanowire through quasi-one-dimensional (1D) subbands. This work experimentally shows the possibility to modulate semiconductor nanowire's thermoelectric properties through 1D subband formation in the diffusive transport regime for electron, a long-sought goal in nanostructured thermoelectrics research. Moreover, we point out the scattering (or disorder) induced energy level broadening as the limiting factor in smearing out the 1D confinement enhanced thermoelectric power factor.
Collapse
|
23
|
Observation of infrared-active modes in Raman scattering from topological insulator nanoplates. NANOTECHNOLOGY 2012; 23:455703. [PMID: 23064215 DOI: 10.1088/0957-4484/23/45/455703] [Citation(s) in RCA: 5] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 06/01/2023]
Abstract
Two infrared (IR)-active vibrational modes, observed at 93 and 113 cm(-1) in Raman scattering, are evidence of an inversion symmetry breakdown in thin (~10 nm) nanoplates of topological insulator Bi(2)Te(3) as-grown on SiO(2). Both Raman and IR modes are preserved after typical device fabrication processes. In nanoplates transferred to another SiO(2) substrate via contact printing, however, the IR modes are absent, and the Raman spectra are similar to those from bulk samples. The differences between as-grown and transferred nanoplates may result from nanoplate-substrate interactions.
Collapse
|
24
|
Connecting the reentrant insulating phase and the zero-field metal-insulator transition in a 2D hole system. PHYSICAL REVIEW LETTERS 2012; 108:106404. [PMID: 22463433 DOI: 10.1103/physrevlett.108.106404] [Citation(s) in RCA: 5] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Received: 09/21/2011] [Indexed: 05/31/2023]
Abstract
We present the transport and capacitance measurements of 10 nm wide GaAs quantum wells with hole densities around the critical point of the 2D metal-insulator transition (critical density p(c) down to 0.8 × 10(10)/cm2, r(s) ∼ 36). For metallic hole density p(c) < p < p(c) + 0.15 × 10(10)/cm2, a reentrant insulating phase (RIP) is observed between the ν = 1 quantum Hall state and the zero-field metallic state and it is attributed to the formation of pinned Wigner crystal. Through studying the evolution of the RIP versus 2D hole density, we show that the RIP is incompressible and continuously connected to the zero-field insulator, suggesting a similar origin for these two phases.
Collapse
|
25
|
Two-dimensional transport-induced linear magneto-resistance in topological insulator Bi2Se3 nanoribbons. ACS NANO 2011; 5:7510-7516. [PMID: 21793543 DOI: 10.1021/nn2024607] [Citation(s) in RCA: 83] [Impact Index Per Article: 6.4] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/31/2023]
Abstract
We report the study of a novel linear magneto-resistance (MR) under perpendicular magnetic fields in Bi(2)Se(3) nanoribbons. Through angular dependence magneto-transport experiments, we show that this linear MR is purely due to two-dimensional (2D) transport, in agreement with the recently discovered linear MR from 2D topological surface state in bulk Bi(2)Te(3), and the linear MR of other gapless semiconductors and graphene. We further show that the linear MR of Bi(2)Se(3) nanoribbons persists to room temperature, underscoring the potential of exploiting topological insulator nanomaterials for room-temperature magneto-electronic applications.
Collapse
|