Medical Image Translation Research Articles

Reliable multi-modal medical image-to-image translation independent of pixel-wise aligned data.

The current mainstream multi-modal medical image-to-image translation methods face a contradiction. Supervised methods with outstanding performance rely on pixel-wise aligned training data to constrain the model optimization. However, obtaining pixel-wise aligned multi-modal medical image datasets is challenging. Unsupervised methods can be trained without paired data, but their reliability cannot be guaranteed. At present, there is no ideal multi-modal medical image-to-image translation method that can generate reliable translation results without the need for pixel-wise aligned data. This work aims to develop a novel medical image-to-image translation model that is independent of pixel-wise aligned data (MITIA), enabling reliable multi-modal medical image-to-image translation under the condition of misaligned trainingdata. The proposed MITIA model utilizes a prior extraction network composed of a multi-modal medical image registration module and a multi-modal misalignment error detection module to extract pixel-level prior information from training data with misalignment errors to the largest extent. The extracted prior information is then used to construct a regularization term to constrain the optimization of the unsupervised cycle-consistent Generative Adversarial Network model, restricting its solution space and thereby improving the performance and reliability of the generator. We trained the MITIA model using six datasets containing different misalignment errors and two well-aligned datasets. Subsequently, we conducted quantitative analysis using peak signal-to-noise ratio and structural similarity as metrics. Moreover, we compared the proposed method with six other state-of-the-art image-to-image translationmethods. The results of both quantitative analysis and qualitative visual inspection indicate that MITIA achieves superior performance compared to the competing state-of-the-art methods, both on misaligned data and aligned data. Furthermore, MITIA shows more stability in the presence of misalignment errors in the training data, regardless of their severity or type. The proposed method achieves outstanding performance in multi-modal medical image-to-image translation tasks without aligned training data. Due to the difficulty in obtaining pixel-wise aligned data for medical image translation tasks, MITIA is expected to generate significant application value in this scenario compared to existingmethods.

Cascaded Multi-path Shortcut Diffusion Model for Medical Image Translation

Image-to-image translation is a vital component in medical imaging processing, with many uses in a wide range of imaging modalities and clinical scenarios. Previous methods include Generative Adversarial Networks (GANs) and Diffusion Models (DMs), which offer realism but suffer from instability and lack uncertainty estimation. Even though both GAN and DM methods have individually exhibited their capability in medical image translation tasks, the potential of combining a GAN and DM to further improve translation performance and to enable uncertainty estimation remains largely unexplored. In this work, we address these challenges by proposing a Cascade Multi-path Shortcut Diffusion Model (CMDM) for high-quality medical image translation and uncertainty estimation. To reduce the required number of iterations and ensure robust performance, our method first obtains a conditional GAN-generated prior image that will be used for the efficient reverse translation with a DM in the subsequent step. Additionally, a multi-path shortcut diffusion strategy is employed to refine translation results and estimate uncertainty. A cascaded pipeline further enhances translation quality, incorporating residual averaging between cascades. We collected three different medical image datasets with two sub-tasks for each dataset to test the generalizability of our approach. Our experimental results found that CMDM can produce high-quality translations comparable to state-of-the-art methods while providing reasonable uncertainty estimations that correlate well with the translation error.

Target-Guided Diffusion Models for Unpaired Cross-Modality Medical Image Translation.

In a clinical setting, the acquisition of certain medical image modality is often unavailable due to various considerations such as cost, radiation, etc. Therefore, unpaired cross-modality translation techniques, which involve training on the unpaired data and synthesizing the target modality with the guidance of the acquired source modality, are of great interest. Previous methods for synthesizing target medical images are to establish one-shot mapping through generative adversarial networks (GANs). As promising alternatives to GANs, diffusion models have recently received wide interests in generative tasks. In this paper, we propose a target-guided diffusion model (TGDM) for unpaired cross-modality medical image translation. For training, to encourage our diffusion model to learn more visual concepts, we adopted a perception prioritized weight scheme (P2W) to the training objectives. For sampling, a pre-trained classifier is adopted in the reverse process to relieve modality-specific remnants from source data. Experiments on both brain MRI-CT and prostate MRI-US datasets demonstrate that the proposed method achieves a visually realistic result that mimics a vivid anatomical section of the target organ. In addition, we have also conducted a subjective assessment based on the synthesized samples to further validate the clinical value of TGDM.

SeRL: Style-embedding representation learning for unsupervised CT images synthesis from unpaired MR images

Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related deaths worldwide, and the fastest-growing cause of cancer deaths in the United States. Computed tomography (CT) and magnetic resonance (MR) imaging are the key image-based examination methods for HCC patients. Patients may alternatively undergo CT and MR exams given a radiation dose of the former and the cost of the latter. Due to organ representation differences between CT and MR, it’s a great challenge to detect the tumors from a time series of data consisting of both CT and MR by a single model. The annotations for these time series data will consume much time and labor for the physicians. Thus, we propose our style-embedding representation learning (SeRL) for unsupervised and unpaired abdomen CT and MR translation. Different from current medical image translation models, the style-representation information from real CT and real MR images has been embedded in the translation process to bypass some local minima during the convergence process and improve the synthesis results. Our patch-corrosion augmentation method enhances the style-embedding representation learning by bringing more diversity to the training data. Combined with the self-attention module, SeRL eliminates the noise caused by low grayscale pixel values during translation. Results on the unpaired HCC patient's CT and MR images show that our proposed SeRL is able to generate high quality CT images from MR ones. Evaluations such as Frechet Inception Distance (FID), Sliced Wasserstein Distance (SWD), and liver segmentation dice score are utilized to demonstrate our advantages over other state-of-the-art unsupervised methods.

Zero-Shot Medical Image Translation via Frequency-Guided Diffusion Models.

Recently, the diffusion model has emerged as a superior generative model that can produce high quality and realistic images. However, for medical image translation, the existing diffusion models are deficient in accurately retaining structural information since the structure details of source domain images are lost during the forward diffusion process and cannot be fully recovered through learned reverse diffusion, while the integrity of anatomical structures is extremely important in medical images. For instance, errors in image translation may distort, shift, or even remove structures and tumors, leading to incorrect diagnosis and inadequate treatments. Training and conditioning diffusion models using paired source and target images with matching anatomy can help. However, such paired data are very difficult and costly to obtain, and may also reduce the robustness of the developed model to out-of-distribution testing data. We propose a frequency-guided diffusion model (FGDM) that employs frequency-domain filters to guide the diffusion model for structure-preserving image translation. Based on its design, FGDM allows zero-shot learning, as it can be trained solely on the data from the target domain, and used directly for source-to-target domain translation without any exposure to the source-domain data during training. We evaluated it on three cone-beam CT (CBCT)-to-CT translation tasks for different anatomical sites, and a cross-institutional MR imaging translation task. FGDM outperformed the state-of-the-art methods (GAN-based, VAE-based, and diffusion-based) in metrics of Fréchet Inception Distance (FID), Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity Index Measure (SSIM), showing its significant advantages in zero-shot medical image translation.

Mutual Information Guided Diffusion for Zero-Shot Cross-Modality Medical Image Translation.

Cross-modality data translation has attracted great interest in medical image computing. Deep generative models show performance improvement in addressing related challenges. Nevertheless, as a fundamental challenge in image translation, the problem of zero-shot learning cross-modality image translation with fidelity remains unanswered. To bridge this gap, we propose a novel unsupervised zero-shot learning method called Mutual Information guided Diffusion Model, which learns to translate an unseen source image to the target modality by leveraging the inherent statistical consistency of Mutual Information between different modalities. To overcome the prohibitive high dimensional Mutual Information calculation, we propose a differentiable local-wise mutual information layer for conditioning the iterative denoising process. The Local-wise-Mutual-Information-Layer captures identical cross-modality features in the statistical domain, offering diffusion guidance without relying on direct mappings between the source and target domains. This advantage allows our method to adapt to changing source domains without the need for retraining, making it highly practical when sufficient labeled source domain data is not available. We demonstrate the superior performance of MIDiffusion in zero-shot cross-modality translation tasks through empirical comparisons with other generative models, including adversarial-based and diffusion-based models. Finally, we showcase the real-world application of MIDiffusion in 3D zero-shot learning-based cross-modality image segmentation tasks.

Unsupervised Medical Image Translation With Adversarial Diffusion Models.

Imputation of missing images via source-to-target modality translation can improve diversity in medical imaging protocols. A pervasive approach for synthesizing target images involves one-shot mapping through generative adversarial networks (GAN). Yet, GAN models that implicitly characterize the image distribution can suffer from limited sample fidelity. Here, we propose a novel method based on adversarial diffusion modeling, SynDiff, for improved performance in medical image translation. To capture a direct correlate of the image distribution, SynDiff leverages a conditional diffusion process that progressively maps noise and source images onto the target image. For fast and accurate image sampling during inference, large diffusion steps are taken with adversarial projections in the reverse diffusion direction. To enable training on unpaired datasets, a cycle-consistent architecture is devised with coupled diffusive and non-diffusive modules that bilaterally translate between two modalities. Extensive assessments are reported on the utility of SynDiff against competing GAN and diffusion models in multi-contrast MRI and MRI-CT translation. Our demonstrations indicate that SynDiff offers quantitatively and qualitatively superior performance against competing baselines.

Multi-category domain-dependent feature-based medical image translation

Multi-category domain-dependent feature-based medical image translation

Machine Learning for Medical Image Translation: A Systematic Review.

CT scans are often the first and only form of brain imaging that is performed to inform treatment plans for neurological patients due to its time- and cost-effective nature. However, MR images give a more detailed picture of tissue structure and characteristics and are more likely to pick up abnormalities and lesions. The purpose of this paper is to review studies which use deep learning methods to generate synthetic medical images of modalities such as MRI and CT. A literature search was performed in March 2023, and relevant articles were selected and analyzed. The year of publication, dataset size, input modality, synthesized modality, deep learning architecture, motivations, and evaluation methods were analyzed. A total of 103 studies were included in this review, all of which were published since 2017. Of these, 74% of studies investigated MRI to CT synthesis, and the remaining studies investigated CT to MRI, Cross MRI, PET to CT, and MRI to PET. Additionally, 58% of studies were motivated by synthesizing CT scans from MRI to perform MRI-only radiation therapy. Other motivations included synthesizing scans to aid diagnosis and completing datasets by synthesizing missing scans. Considerably more research has been carried out on MRI to CT synthesis, despite CT to MRI synthesis yielding specific benefits. A limitation on medical image synthesis is that medical datasets, especially paired datasets of different modalities, are lacking in size and availability; it is therefore recommended that a global consortium be developed to obtain and make available more datasets for use. Finally, it is recommended that work be carried out to establish all uses of the synthesis of medical scans in clinical practice and discover which evaluation methods are suitable for assessing the synthesized images for these needs.

Attentive continuous generative self-training for unsupervised domain adaptive medical image translation.

Self-training is an important class of unsupervised domain adaptation (UDA) approaches that are used to mitigate the problem of domain shift, when applying knowledge learned from a labeled source domain to unlabeled and heterogeneous target domains. While self-training-based UDA has shown considerable promise on discriminative tasks, including classification and segmentation, through reliable pseudo-label filtering based on the maximum softmax probability, there is a paucity of prior work on self-training-based UDA for generative tasks, including image modality translation. To fill this gap, in this work, we seek to develop a generative self-training (GST) framework for domain adaptive image translation with continuous value prediction and regression objectives. Specifically, we quantify both aleatoric and epistemic uncertainties within our GST using variational Bayes learning to measure the reliability of synthesized data. We also introduce a self-attention scheme that de-emphasizes the background region to prevent it from dominating the training process. The adaptation is then carried out by an alternating optimization scheme with target domain supervision that focuses attention on the regions with reliable pseudo-labels. We evaluated our framework on two cross-scanner/center, inter-subject translation tasks, including tagged-to-cine magnetic resonance (MR) image translation and T1-weighted MR-to-fractional anisotropy translation. Extensive validations with unpaired target domain data showed that our GST yielded superior synthesis performance in comparison to adversarial training UDA methods.

Constrained CycleGAN for effective generation of ultrasound sector images of improved spatial resolution

Objective. A phased or a curvilinear array produces ultrasound (US) images with a sector field of view, which inherently exhibits spatially-varying image resolution with inferior quality in the far zone and towards the two sides azimuthally. Sector US images with improved spatial resolutions are favorable for accurate quantitative analysis of large and dynamic organs, such as the heart. Therefore, this study aims to translate US images with spatially-varying resolution to ones with less spatially-varying resolution. CycleGAN has been a prominent choice for unpaired medical image translation; however, it neither guarantees structural consistency nor preserves backscattering patterns between input and generated images for unpaired US images. Approach. To circumvent this limitation, we propose a constrained CycleGAN (CCycleGAN), which directly performs US image generation with unpaired images acquired by different ultrasound array probes. In addition to conventional adversarial and cycle-consistency losses of CycleGAN, CCycleGAN introduces an identical loss and a correlation coefficient loss based on intrinsic US backscattered signal properties to constrain structural consistency and backscattering patterns, respectively. Instead of post-processed B-mode images, CCycleGAN uses envelope data directly obtained from beamformed radio-frequency signals without any other nonlinear postprocessing. Main results. In vitro phantom results demonstrate that CCycleGAN successfully generates images with improved spatial resolution as well as higher peak signal-to-noise ratio and structural similarity compared with benchmarks. Significance. CCycleGAN-generated US images of the in vivo human beating heart further facilitate higher quality heart wall motion estimation than benchmarks-generated ones, particularly in deep regions. The codes are available at https://github.com/xfsun99/CCycleGAN-TF2

Basic GAN Models and the Application In Medical Image Field

Generative adversarial networks (GANs) have shown its prominent performance to the world and be exploited in many areas. For example, generating human face image, translating image between different domain, generating supplementary data for tasks which lacks training and labeled data. Specially, GANs also shows its potential in medical image analysis. It can tackle the problems in medical image analysis such as medical image translation, segmentation, reconstruction, detection and image classification. Being capable of generating images at amazing level of realisim, GANs gives the opportunity to resolve the problem that labeled data for those rare diseases cannot meet the need for medical field. In this paper, we give an overview of GANs and its application in medical image analysis. Defects and advantages of those GANs methods are also thoroughly reviewed. We also discuss its potential improvement in future. We review those most frequently methods published until now. And essential details about papers we discussed in this paper and access to that paper is attached at the endpage.

Multi-domain medical image translation generation for lung image classification based on generative adversarial networks

ObjectiveLung image classification-assisted diagnosis has a large application market. Aiming at the problems of poor attention to existing translation models, the insufficient ability of key transfer and generation, insufficient quality of generated images, and lack of detailed features, this paper conducts research on lung medical image translation and lung image classification based on generative adversarial networks. MethodsThis paper proposes a medical image multi-domain translation algorithm MI-GAN based on the key migration branch. After the actual analysis of the imbalanced medical image data, the key target domain images are selected, the key migration branch is established, and a single generator is used to complete the medical image multi-domain translation. The conversion between domains ensures the attention performance of the medical image multi-domain translation model and the quality of the synthesized images. At the same time, a lung image classification model based on synthetic image data augmentation is proposed. The synthetic lung CT medical images and the original real medical images are used as the training set together to study the performance of the auxiliary diagnosis model in the classification of normal healthy subjects, and also of the mild and severe COVID-19 patients. ResultsBased on the chest CT image dataset, MI-GAN has completed the mutual conversion and generation of normal lung images without disease, viral pneumonia and Mild COVID-19 images. The synthetic images GAN-test and GAN-train indicators reached, respectively 92.188% and 85.069%, compared with other generative models in terms of authenticity and diversity, there is a considerable improvement. The accuracy rate of pneumonia diagnosis of the lung image classification model is 93.85%, which is 3.1% higher than that of the diagnosis model trained only with real images; the sensitivity of disease diagnosis is 96.69%, a relative improvement of 7.1%. 1%, the specificity was 89.70%; the area under the ROC curve (AUC) increased from 94.00% to 96.17%. ConclusionIn this paper, a multi-domain translation model of medical images based on the key transfer branch is proposed, which enables the translation network to have key transfer and attention performance. It is verified on lung CT images and achieved good results. The required medical images are synthesized by the above medical image translation model, and the effectiveness of the synthesized images on the lung image classification network is verified experimentally.

H&E-like staining of OCT images of human skin via generative adversarial network

Noninvasive and high-speed optical coherence tomography (OCT) systems have been widely deployed for daily clinical uses. High-resolution OCTs are advancing rapidly; however, grey-level OCT images are not easy to read for pathologists due to the lack of diagnosis specificity compared with hematoxylin and eosin (H&E) stained images. This work presents an OCT to H&E image translation model to convert the OCT images to H&E-like stained images using unpaired OCT and H&E datasets. “H&E like” means the stratum corneum (SC) boundary and the dermal-epidermal junction (DEJ) of the OCT and the translated images are consistent. Pre-trained segmentation models for the DEJ and the SC are exploited to enhance the performance of anatomical image translation and reduce the DEJ and SC lower boundary errors to ±2.3 and ±1.7 μm, respectively. A pre-trained VGG16 network extracts the features of the nuclei. Pearson's correlation coefficient of the nuclei location and size consistency is 84% ± 1%. As a result, in vivo medical image translation accuracy with cellular resolution was achieved.

Swin transformer-based GAN for multi-modal medical image translation.

Medical image-to-image translation is considered a new direction with many potential applications in the medical field. The medical image-to-image translation is dominated by two models, including supervised Pix2Pix and unsupervised cyclic-consistency generative adversarial network (GAN). However, existing methods still have two shortcomings: 1) the Pix2Pix requires paired and pixel-aligned images, which are difficult to acquire. Nevertheless, the optimum output of the cycle-consistency model may not be unique. 2) They are still deficient in capturing the global features and modeling long-distance interactions, which are critical for regions with complex anatomical structures. We propose a Swin Transformer-based GAN for Multi-Modal Medical Image Translation, named MMTrans. Specifically, MMTrans consists of a generator, a registration network, and a discriminator. The Swin Transformer-based generator enables to generate images with the same content as source modality images and similar style information of target modality images. The encoder part of the registration network, based on Swin Transformer, is utilized to predict deformable vector fields. The convolution-based discriminator determines whether the target modality images are similar to the generator or from the real images. Extensive experiments conducted using the public dataset and clinical datasets showed that our network outperformed other advanced medical image translation methods in both aligned and unpaired datasets and has great potential to be applied in clinical applications.

Multimodal image translation via deep learning inference model trained in video domain

BackgroundCurrent medical image translation is implemented in the image domain. Considering the medical image acquisition is essentially a temporally continuous process, we attempt to develop a novel image translation framework via deep learning trained in video domain for generating synthesized computed tomography (CT) images from cone-beam computed tomography (CBCT) images.MethodsFor a proof-of-concept demonstration, CBCT and CT images from 100 patients were collected to demonstrate the feasibility and reliability of the proposed framework. The CBCT and CT images were further registered as paired samples and used as the input data for the supervised model training. A vid2vid framework based on the conditional GAN network, with carefully-designed generators, discriminators and a new spatio-temporal learning objective, was applied to realize the CBCT–CT image translation in the video domain. Four evaluation metrics, including mean absolute error (MAE), peak signal-to-noise ratio (PSNR), normalized cross-correlation (NCC), and structural similarity (SSIM), were calculated on all the real and synthetic CT images from 10 new testing patients to illustrate the model performance.ResultsThe average values for four evaluation metrics, including MAE, PSNR, NCC, and SSIM, are 23.27 ± 5.53, 32.67 ± 1.98, 0.99 ± 0.0059, and 0.97 ± 0.028, respectively. Most of the pixel-wise hounsfield units value differences between real and synthetic CT images are within 50. The synthetic CT images have great agreement with the real CT images and the image quality is improved with lower noise and artifacts compared with CBCT images.ConclusionsWe developed a deep-learning-based approach to perform the medical image translation problem in the video domain. Although the feasibility and reliability of the proposed framework were demonstrated by CBCT–CT image translation, it can be easily extended to other types of medical images. The current results illustrate that it is a very promising method that may pave a new path for medical image translation research.

Development of a GAN architecture based on integrating global and local information for paired and unpaired medical image translation

The purpose of this study is to develop and evaluate a novel end-to-end generative adversarial network for image-to-image translation of medical paired and unpaired data. The proposed deep framework is based on generative adversarial networks, in which blocks are designed and used to improve the estimated images and achieve more accurate results. A joint attention block is designed to capture and fuse the global and local information during the training process. A dual discriminator is employed to accurately determine whether the generated images are real or fake. A global guidance up-sampling block is proposed to reconstruct the local details of the estimated images with guidance of the low-level information of the shallow layers. Moreover, an edge detection network is utilized to make the tissue edges sharper. The loss function is composed of different criteria considering the geometric characteristics for obtaining more accurate results. Different translation tasks including PD/T2, CT/MR, and PET/CT are considered to evaluate the performance of the proposed deep framework. The evaluation results indicated averagely 28.28, 10.87, 31.30, 0.90, 0.92, 0.85, 24.80, and 0.042 for root-mean-squared error (RMSE), mean absolute error (MAE), peak signal to noise ratio (PSNR), dice similarity coefficient (DSC), normalized cross-correlation (NCC), structural similarity index measure (SSIM), learned perceptual image patch similarity (LPIPS) and frechet inception distance (FID) respectively. Evaluation of the results showed improvement of 106, 66, 12, 0.22, 0.17, 0.15, 58, and 0.39 for RMSE, MAE, PSNR, DSC, NCC, SSIM, LPIPS, and FID, respectively when the proposed blocks were used. Also, comparing the performance of the proposed network against other methods in the two modes of using paired and unpaired data showed the better performance of the proposed network.

Medical image translation using an edge-guided generative adversarial network with global-to-local feature fusion.

In this paper, we propose a framework based deep learning for medical image translation using paired and unpaired training data. Initially, a deep neural network with an encoder-decoder structure is proposed for image-to-image translation using paired training data. A multi-scale context aggregation approach is then used to extract various features from different levels of encoding, which are used during the corresponding network decoding stage. At this point, we further propose an edge-guided generative adversarial network for image-to-image translation based on unpaired training data. An edge constraint loss function is used to improve network performance in tissue boundaries. To analyze framework performance, we conducted five different medical image translation tasks. The assessment demonstrates that the proposed deep learning framework brings significant improvement beyond state-of-the-arts.

Cross-Modality Breast Image Translation with Improved Resolution Using Generative Adversarial Networks

Unpaired cross domain medical image translation is a challenging problem as the target image modality cannot be mapped directly from the input data distribution. The best approach used till date, was by Cycle generative adversarial network, which utilized the cycle consistency loss to perform the task. Although efficient, still the resultant image size was small and blurry. Recent trends show that due to change in lifestyle and increased exposure to carcinogens in different forms has increased occurrence of cancer. Statistics show that every one in eight women might develop breast cancer at some stage of her life. Hence, this paper focuses on a combination of two GANs, CycleGAN and Super resolution GAN is used in two stages to obtain translated breast images with improved resolution. The proposed model is tested on images of breast cancer patients to obtain CT Scan using PET scan and vice versa so that the patients are not exposed to an extremely potent dose of radiation. In order to ensure the presence of tumour in the estimated image, a simplified U-net feature extractor is also used. Quantitative studies are carried out for both the stages of simulation to establish the efficiency of the proposed model.

A novel deep learning conditional generative adversarial network for producing angiography images from retinal fundus photographs

Fluorescein angiography (FA) is a procedure used to image the vascular structure of the retina and requires the insertion of an exogenous dye with potential adverse side effects. Currently, there is only one alternative non-invasive system based on Optical coherence tomography (OCT) technology, called OCT angiography (OCTA), capable of visualizing retina vasculature. However, due to its cost and limited view, OCTA technology is not widely used. Retinal fundus photography is a safe imaging technique used for capturing the overall structure of the retina. In order to visualize retinal vasculature without the need for FA and in a cost-effective, non-invasive, and accurate manner, we propose a deep learning conditional generative adversarial network (GAN) capable of producing FA images from fundus photographs. The proposed GAN produces anatomically accurate angiograms, with similar fidelity to FA images, and significantly outperforms two other state-of-the-art generative algorithms (p<.001 and p<.0001). Furthermore, evaluations by experts shows that our proposed model produces such high quality FA images that are indistinguishable from real angiograms. Our model as the first application of artificial intelligence and deep learning to medical image translation, by employing a theoretical framework capable of establishing a shared feature-space between two domains (i.e. funduscopy and fluorescein angiography) provides an unrivaled way for the translation of images from one domain to the other.