The scarcity and imbalance of medical image datasets hinder the development of robust computer-aided diagnosis (CAD) systems for breast cancer. This study explores the application of advanced generative models, based on generative artificial intelligence (GenAI), for the synthesis of digital breast ultrasound images. Using a hybrid Conditional Variational Autoencoder-Wasserstein Generative Adversarial Network (CVAE-WGAN) architecture, we developed a system to generate high-quality synthetic images conditioned on the class (malignant vs. normal/benign). These synthetic images, generated from the low-resolution BreastMNIST dataset and filtered for quality, were systematically integrated with real training data at different mixing ratios (W). The performance of a CNN classifier trained on these mixed datasets was evaluated against a baseline model trained only on real data balanced with SMOTE. The optimal integration (mixing weight W=0.25) produced a significant performance increase on the real test set: +8.17% in macro-average F1-score and +4.58% in accuracy compared to using real data alone. Analysis confirmed the originality of the generated samples. This approach offers a promising solution for overcoming data limitations in image-based breast cancer diagnostics, potentially improving the capabilities of CAD systems.

In modern healthcare, telemedicine, health records, and AI-driven diagnostics depend on medical image watermarking to secure chest X-rays for pneumonia diagnosis, ensuring data integrity, confidentiality, and authenticity. A 2024 study found over 70 % of healthcare institutions faced medical image data breaches. Yet, current methods falter in imperceptibility, robustness against attacks, and deployment efficiency. ViTU-Net integrates cutting-edge techniques to address these multifaceted challenges in medical image security and analysis. The model's core component, the Vision Transformer (ViT) encoder, efficiently captures global dependencies and spatial information, while the U-Net decoder enhances image reconstruction, with both components leveraging the Adaptive Hierarchical Spatial Attention (AHSA) module for improved spatial processing. Additionally, the patch-based LSB embedding mechanism ensures focused embedding of reversible fragile watermarks within each patch of the segmented non-diagnostic region (RONI), guided dynamically by adaptive masks derived from the attention mechanism, minimizing impact on diagnostic accuracy while maximizing precision and ensuring optimal utilization of spatial information. The hybrid meta-heuristic optimization algorithm, TuniBee Fusion, dynamically optimizes watermarking parameters, striking a balance between exploration and exploitation, thereby enhancing watermarking efficiency and robustness. The incorporation of advanced cryptographic techniques, including SHA-512 hashing and AES encryption, fortifies the model's security, ensuring the authenticity and confidentiality of watermarked medical images. A PSNR value of 60.7 dB, along with an NCC value of 0.9999 and an SSIM value of 1.00, underscores its effectiveness in preserving image quality, security, and diagnostic accuracy. Robustness analysis against a spectrum of attacks validates ViTU-Net's resilience in real-world scenarios.

Providing effective treatment and making informed clinical decisions are essential goals of modern medicine and clinical care. We are interested in simulating disease dynamics for clinical decision-making, leveraging recent advances in large generative models. To this end, we introduce the Medical World Model (MeWM), the first world model in medicine that visually predicts future disease states based on clinical decisions. MeWM comprises (i) vision-language models to serve as policy models, and (ii) tumor generative models as dynamics models. The policy model generates action plans, such as clinical treatments, while the dynamics model simulates tumor progression or regression under given treatment conditions. Building on this, we propose the inverse dynamics model that applies survival analysis to the simulated post-treatment tumor, enabling the evaluation of treatment efficacy and the selection of the optimal clinical action plan. As a result, the proposed MeWM simulates disease dynamics by synthesizing post-treatment tumors, with state-of-the-art specificity in Turing tests evaluated by radiologists. Simultaneously, its inverse dynamics model outperforms medical-specialized GPTs in optimizing individualized treatment protocols across all metrics. Notably, MeWM improves clinical decision-making for interventional physicians, boosting F1-score in selecting the optimal TACE protocol by 13%, paving the way for future integration of medical world models as the second readers.

Routine knee MRI protocols for 1.5 T and 3 T scanners, do not include T2*-w gradient echo (T2*W) images, which are useful in several clinical scenarios such as the assessment of cartilage, synovial blooming (deposition of hemosiderin), chondrocalcinosis and the evaluation of the physis in pediatric patients. Herein, we aimed to develop an open-source deep learning model that creates synthetic T2*W images of the knee using fat-suppressed intermediate-weighted images. A cycleGAN model was trained with 12,118 sagittal knee MR images and tested on an independent set of 2996 images. Diagnostic interchangeability of synthetic T2*W images was assessed against a series of findings. Voxel intensity of four tissues was evaluated with Bland-Altman plots. Image quality was assessed with the use of root mean squared error (NRMSE), structural similarity index (SSIM) and peak signal-to-noise ratio (PSNR). Code, model and a standalone executable file are provided on github. The model achieved a median NRMSE, PSNR and SSIM of 0.5, 17.4, and 0.5, respectively. Images were found interchangeable with an intraclass correlation coefficient >0.95 for all findings. Mean voxel intensity was equal between synthetic and conventional images. Four types of artifacts were identified: geometrical distortion (86/163 cases), object insertion/omission (11/163 cases), a wrap-around-like (26/163 cases) and an incomplete fat-suppression artifact (120/163 cases), which had a median 0 impact (no impact) on the diagnosis. In conclusion, the developed open-source GAN model creates synthetic T2*W images of the knee of high diagnostic value and quality. The identified artifacts had no or minor effect on the diagnostic value of the images.

The zero-watermarking methods provide a means of lossless, which was adopted to protect medical image copyright requiring high integrity. However, most existing studies have only focused on robustness and there has been little discussion about the analysis and experiment on discriminability. Therefore, this paper proposes a trusted robust zero-watermarking scheme for medical images based on Deep convolution neural network (DCNN) and the hyperchaotic encryption system. Firstly, the medical image is converted into several feature map matrices by the specific convolution layer of DCNN. Then, a stable Gram matrix is obtained by calculating the colinear correlation between different channels in feature map matrices. Finally, the Gram matrixes of the medical image and the feature map matrixes of the watermark image are fused by the trained DCNN to generate the zero-watermark. Meanwhile, we propose two feature evaluation criteria for finding differentiated eigenvalues. The eigenvalue is used as the explicit key to encrypt the generated zero-watermark by Lorenz hyperchaotic encryption, which enhances security and discriminability. The experimental results show that the proposed scheme can resist common image attacks and geometric attacks, and is distinguishable in experiments, being applicable for the copyright protection of medical images.

Medical professionals, especially those in training, often depend on visual reference materials to support an accurate diagnosis and develop pattern recognition skills. However, existing resources may lack the diversity and accessibility needed for broad and effective clinical learning. This paper presents a system designed to generate realistic medical images from textual descriptions and validate their accuracy through a classification model. A pretrained stable diffusion model was fine-tuned using Low-Rank Adaptation (LoRA) on the PathMNIST dataset, consisting of nine colorectal histopathology tissue types. The generative model was trained multiple times using different training parameter configurations, guided by domain-specific prompts to capture meaningful features. To ensure quality control, a ResNet-18 classification model was trained on the same dataset, achieving 99.76% accuracy in detecting the correct label of a colorectal histopathological medical image. Generated images were then filtered using the trained classifier and an iterative process, where inaccurate outputs were discarded and regenerated until they were correctly classified. The highest performing version of the generative model from experimentation achieved an F1 score of 0.6727, with precision and recall scores of 0.6817 and 0.7111, respectively. Some types of tissue, such as adipose tissue and lymphocytes, reached perfect classification scores, while others proved more challenging due to structural complexity. The self-validating approach created demonstrates a reliable method for synthesizing domain-specific medical images because of high accuracy in both the generation and classification portions of the system, with potential applications in both diagnostic support and clinical education. Future work includes improving prompt-specific accuracy and extending the system to other areas of medical imaging.

Cone-beam computed tomography (CBCT) is widely used in radiotherapy to image patient configuration before treatment but its image quality is lower than planning CT due to scattering, motion, and reconstruction methods. This reduces the accuracy of Hounsfield units (HU) and limits its use in adaptive radiation therapy (ART). However, synthetic CT (sCT) generation using deep learning methods for CBCT intensity correction faces challenges due to deformation. To address these issues, we propose enhancing CBCT quality using a conditional denoising diffusion probability model (CDDPM), which is trained on pseudo-CBCT created by adding pseudo-scatter to planning CT. The CDDPM transforms CBCT into high-quality sCT, improving HU accuracy while preserving anatomical configuration. The performance evaluation of the proposed sCT showed a reduction in mean absolute error (MAE) from 81.19 HU for CBCT to 24.89 HU for the sCT. Peak signal-to-noise ratio (PSNR) improved from 31.20 dB for CBCT to 33.81 dB for the sCT. The Dice and Jaccard coefficients between CBCT and sCT for the colon, prostate, and bladder ranged from 0.69 to 0.91. When compared to other deep learning models, the proposed sCT outperformed them in terms of accuracy and anatomical preservation. The dosimetry analysis for prostate cancer revealed a dose error of over 10% with CBCT but nearly 0% with the sCT. Gamma pass rates for the proposed sCT exceeded 90% for all dose criteria, indicating high agreement with CT-based dose distributions. These results show that the proposed sCT improves image quality, dosimetry accuracy, and treatment planning, advancing ART for pelvic cancer.

Research in the cross-modal medical image translation domain has been very productive over the past few years in tackling the scarce availability of large curated multi-modality datasets with the promising performance of GAN-based architectures. However, only a few of these studies assessed task-based related performance of these synthetic data, especially for the training of deep models. We design and compare different GAN-based frameworks for generating synthetic brain[18F]fluorodeoxyglucose (FDG) PET images from T1 weighted MRI data. We first perform standard qualitative and quantitative visual quality evaluation. Then, we explore further impact of using these fake PET data in the training of a deep unsupervised anomaly detection (UAD) model designed to detect subtle epilepsy lesions in T1 MRI and FDG PET images. We introduce novel diagnostic task-oriented quality metrics of the synthetic FDG PET data tailored to our unsupervised detection task, then use these fake data to train a use case UAD model combining a deep representation learning based on siamese autoencoders with a OC-SVM density support estimation model. This model is trained on normal subjects only and allows the detection of any variation from the pattern of the normal population. We compare the detection performance of models trained on 35 paired real MR T1 of normal subjects paired either on 35 true PET images or on 35 synthetic PET images generated from the best performing generative models. Performance analysis is conducted on 17 exams of epilepsy patients undergoing surgery. The best performing GAN-based models allow generating realistic fake PET images of control subject with SSIM and PSNR values around 0.9 and 23.8, respectively and in distribution (ID) with regard to the true control dataset. The best UAD model trained on these synthetic normative PET data allows reaching 74% sensitivity. Our results confirm that GAN-based models are the best suited for MR T1 to FDG PET translation, outperforming transformer or diffusion models. We also demonstrate the diagnostic value of these synthetic data for the training of UAD models and evaluation on clinical exams of epilepsy patients. Our code and the normative image dataset are available.

The generation of Synthetic Computed Tomography (sCT) images has become a pivotal methodology in modern clinical practice, particularly in the context of Radiotherapy (RT) treatment planning. The use of sCT enables the calculation of doses, pushing towards Magnetic Resonance Imaging (MRI) guided radiotherapy treatments. Moreover, with the introduction of MRI-Positron Emission Tomography (PET) hybrid scanners, the derivation of sCT from MRI can improve the attenuation correction of PET images. Deep learning methods for MRI-to-sCT have shown promising results, but their reliance on single-centre training dataset limits generalisation capabilities to diverse clinical settings. Moreover, creating centralised multi-centre datasets may pose privacy concerns. To address the aforementioned issues, we introduced FedSynthCT-Brain, an approach based on the Federated Learning (FL) paradigm for MRI-to-sCT in brain imaging. This is among the first applications of FL for MRI-to-sCT, employing a cross-silo horizontal FL approach that allows multiple centres to collaboratively train a U-Net-based deep learning model. We validated our method using real multicentre data from four European and American centres, simulating heterogeneous scanner types and acquisition modalities, and tested its performance on an independent dataset from a centre outside the federation. In the case of the unseen centre, the federated model achieved a median Mean Absolute Error (MAE) of 102.0 HU across 23 patients, with an interquartile range of 96.7-110.5 HU. The median (interquartile range) for the Structural Similarity Index (SSIM) and the Peak Signal to Noise Ratio (PNSR) were 0.89 (0.86-0.89) and 26.58 (25.52-27.42), respectively. The analysis of the results showed acceptable performances of the federated approach, thus highlighting the potential of FL to enhance MRI-to-sCT to improve generalisability and advancing safe and equitable clinical applications while fostering collaboration and preserving data privacy.

Accurate and timely diagnosis of brain tumors is critical for patient management and treatment planning. Magnetic resonance imaging (MRI) is a widely used modality for brain tumor detection and characterization, often aided by the administration of gadolinium-based contrast agents (GBCAs) to improve tumor visualization. Recently, deep learning models have shown remarkable success in predicting contrast-enhancement in medical images, thereby reducing the need of GBCAs and potentially minimizing patient discomfort and risks. In this paper, we present a study aimed at investigating the generalization capabilities of a neural network trained to predict full contrast in brain tumor images from noncontrast MRI scans. While initial results exhibited promising performance on a specific tumor type at a certain stage using a specific dataset, our attempts to extend this success to other tumor types and diverse patient populations yielded unexpected challenges and limitations. Through a rigorous analysis of the factor contributing to these negative results, we aim to shed light on the complexities associated with generalizing contrast enhancement prediction in medical brain tumor imaging, offering valuable insights for future research and clinical applications.