Oliver Díaz

Karim Lekadir, Aasa Feragen, Abdul Joseph Fofanah et al. · eth-zurich

Despite major advances in artificial intelligence (AI) for medicine and healthcare, the deployment and adoption of AI technologies remain limited in real-world clinical practice. In recent years, concerns have been raised about the technical, clinical, ethical and legal risks associated with medical AI. To increase real world adoption, it is essential that medical AI tools are trusted and accepted by patients, clinicians, health organisations and authorities. This work describes the FUTURE-AI guideline as the first international consensus framework for guiding the development and deployment of trustworthy AI tools in healthcare. The FUTURE-AI consortium was founded in 2021 and currently comprises 118 inter-disciplinary experts from 51 countries representing all continents, including AI scientists, clinicians, ethicists, and social scientists. Over a two-year period, the consortium defined guiding principles and best practices for trustworthy AI through an iterative process comprising an in-depth literature review, a modified Delphi survey, and online consensus meetings. The FUTURE-AI framework was established based on 6 guiding principles for trustworthy AI in healthcare, i.e. Fairness, Universality, Traceability, Usability, Robustness and Explainability. Through consensus, a set of 28 best practices were defined, addressing technical, clinical, legal and socio-ethical dimensions. The recommendations cover the entire lifecycle of medical AI, from design, development and validation to regulation, deployment, and monitoring. FUTURE-AI is a risk-informed, assumption-free guideline which provides a structured approach for constructing medical AI tools that will be trusted, deployed and adopted in real-world practice. Researchers are encouraged to take the recommendations into account in proof-of-concept stages to facilitate future translation towards clinical practice of medical AI.

Richard Osuala, Grzegorz Skorupko, Noussair Lazrak et al.

Synthetic data generated by generative models can enhance the performance and capabilities of data-hungry deep learning models in medical imaging. However, there is (1) limited availability of (synthetic) datasets and (2) generative models are complex to train, which hinders their adoption in research and clinical applications. To reduce this entry barrier, we propose medigan, a one-stop shop for pretrained generative models implemented as an open-source framework-agnostic Python library. medigan allows researchers and developers to create, increase, and domain-adapt their training data in just a few lines of code. Guided by design decisions based on gathered end-user requirements, we implement medigan based on modular components for generative model (i) execution, (ii) visualisation, (iii) search & ranking, and (iv) contribution. The library's scalability and design is demonstrated by its growing number of integrated and readily-usable pretrained generative models consisting of 21 models utilising 9 different Generative Adversarial Network architectures trained on 11 datasets from 4 domains, namely, mammography, endoscopy, x-ray, and MRI. Furthermore, 3 applications of medigan are analysed in this work, which include (a) enabling community-wide sharing of restricted data, (b) investigating generative model evaluation metrics, and (c) improving clinical downstream tasks. In (b), extending on common medical image synthesis assessment and reporting standards, we show Fréchet Inception Distance variability based on image normalisation and radiology-specific feature extraction.

Thorsten Kalb, Kaisar Kushibar, Celia Cintas et al.

Addressing fairness in lesion classification from dermatological images is crucial due to variations in how skin diseases manifest across skin tones. However, the absence of skin tone labels in public datasets hinders building a fair classifier. To date, such skin tone labels have been estimated prior to fairness analysis in independent studies using the Individual Typology Angle (ITA). Briefly, ITA calculates an angle based on pixels extracted from skin images taking into account the lightness and yellow-blue tints. These angles are then categorised into skin tones that are subsequently used to analyse fairness in skin cancer classification. In this work, we review and compare four ITA-based approaches of skin tone classification on the ISIC18 dataset, a common benchmark for assessing skin cancer classification fairness in the literature. Our analyses reveal a high disagreement among previously published studies demonstrating the risks of ITA-based skin tone estimation methods. Moreover, we investigate the causes of such large discrepancy among these approaches and find that the lack of diversity in the ISIC18 dataset limits its use as a testbed for fairness analysis. Finally, we recommend further research on robust ITA estimation and diverse dataset acquisition with skin tone annotation to facilitate conclusive fairness assessments of artificial intelligence tools in dermatology. Our code is available at https://github.com/tkalbl/RevisitingSkinToneFairness.

Lidia Garrucho, Kaisar Kushibar, Richard Osuala et al.

Computer-aided detection systems based on deep learning have shown good performance in breast cancer detection. However, high-density breasts show poorer detection performance since dense tissues can mask or even simulate masses. Therefore, the sensitivity of mammography for breast cancer detection can be reduced by more than 20% in dense breasts. Additionally, extremely dense cases reported an increased risk of cancer compared to low-density breasts. This study aims to improve the mass detection performance in highdensity breasts using synthetic high-density full-field digital mammograms (FFDM) as data augmentation during breast mass detection model training. To this end, a total of five cycle-consistent GAN (CycleGAN) models using three FFDM datasets were trained for low-to-high-density image translation in highresolution mammograms. The training images were split by breast density BIRADS categories, being BI-RADS A almost entirely fatty and BI-RADS D extremely dense breasts. Our results showed that the proposed data augmentation technique improved the sensitivity and precision of mass detection in models trained with small datasets and improved the domain generalization of the models trained with large databases. In addition, the clinical realism of the synthetic images was evaluated in a reader study involving two expert radiologists and one surgical oncologist.

Zuzanna Szafranowska, Richard Osuala, Bennet Breier et al.

Early detection of breast cancer in mammography screening via deep-learning based computer-aided detection systems shows promising potential in improving the curability and mortality rates of breast cancer. However, many clinical centres are restricted in the amount and heterogeneity of available data to train such models to (i) achieve promising performance and to (ii) generalise well across acquisition protocols and domains. As sharing data between centres is restricted due to patient privacy concerns, we propose a potential solution: sharing trained generative models between centres as substitute for real patient data. In this work, we use three well known mammography datasets to simulate three different centres, where one centre receives the trained generator of Generative Adversarial Networks (GANs) from the two remaining centres in order to augment the size and heterogeneity of its training dataset. We evaluate the utility of this approach on mammography patch classification on the test set of the GAN-receiving centre using two different classification models, (a) a convolutional neural network and (b) a transformer neural network. Our experiments demonstrate that shared GANs notably increase the performance of both transformer and convolutional classification models and highlight this approach as a viable alternative to inter-centre data sharing.

Richard Osuala, Smriti Joshi, Apostolia Tsirikoglou et al.

This paper presents a method for virtual contrast enhancement in breast MRI, offering a promising non-invasive alternative to traditional contrast agent-based DCE-MRI acquisition. Using a conditional generative adversarial network, we predict DCE-MRI images, including jointly-generated sequences of multiple corresponding DCE-MRI timepoints, from non-contrast-enhanced MRIs, enabling tumor localization and characterization without the associated health risks. Furthermore, we qualitatively and quantitatively evaluate the synthetic DCE-MRI images, proposing a multi-metric Scaled Aggregate Measure (SAMe), assessing their utility in a tumor segmentation downstream task, and conclude with an analysis of the temporal patterns in multi-sequence DCE-MRI generation. Our approach demonstrates promising results in generating realistic and useful DCE-MRI sequences, highlighting the potential of virtual contrast enhancement for improving breast cancer diagnosis and treatment, particularly for patients where contrast agent administration is contraindicated.

Richard Osuala, Daniel M. Lang, Preeti Verma et al.

Contrast agents in dynamic contrast enhanced magnetic resonance imaging allow to localize tumors and observe their contrast kinetics, which is essential for cancer characterization and respective treatment decision-making. However, contrast agent administration is not only associated with adverse health risks, but also restricted for patients during pregnancy, and for those with kidney malfunction, or other adverse reactions. With contrast uptake as key biomarker for lesion malignancy, cancer recurrence risk, and treatment response, it becomes pivotal to reduce the dependency on intravenous contrast agent administration. To this end, we propose a multi-conditional latent diffusion model capable of acquisition time-conditioned image synthesis of DCE-MRI temporal sequences. To evaluate medical image synthesis, we additionally propose and validate the Fréchet radiomics distance as an image quality measure based on biomarker variability between synthetic and real imaging data. Our results demonstrate our method's ability to generate realistic multi-sequence fat-saturated breast DCE-MRI and uncover the emerging potential of deep learning based contrast kinetics simulation. We publicly share our accessible codebase at https://github.com/RichardObi/ccnet and provide a user-friendly library for Fréchet radiomics distance calculation at https://pypi.org/project/frd-score.

Smriti Joshi, Richard Osuala, Lidia Garrucho et al.

Test-time adaptation enables a trained model to adjust to a new domain during inference, making it particularly valuable in clinical settings where such on-the-fly adaptation is required. However, existing techniques depend on large target domain datasets, which are often impractical and unavailable in medical scenarios that demand per-patient, real-time inference. Moreover, current methods commonly focus on two-dimensional images, failing to leverage the volumetric richness of medical imaging data. Bridging this gap, we propose a Patch-Based Multi-View Co-Training method for Single Image Test-Time adaptation. Our method enforces feature and prediction consistency through uncertainty-guided self-training, enabling effective volumetric segmentation in the target domain with only a single test-time image. Validated on three publicly available breast magnetic resonance imaging datasets for tumor segmentation, our method achieves performance close to the upper bound supervised benchmark while also outperforming all existing state-of-the-art methods, on average by a Dice Similarity Coefficient of 3.75%. We publicly share our accessible codebase, readily integrable with the popular nnUNet framework, at https://github.com/smriti-joshi/muvi.git.

Sebastian Ibarra, Javier del Riego, Alessandro Catanese et al.

Dynamic contrast-enhanced (DCE) MRI is essential for breast cancer diagnosis and treatment. However, its reliance on contrast agents introduces safety concerns, contraindications, increased cost, and workflow complexity. To this end, we present pre-contrast conditioned denoising diffusion probabilistic models to synthesize DCE-MRI, introducing, evaluating, and comparing a total of 22 generative model variants in both single-breast and full breast settings. Towards enhancing lesion fidelity, we introduce both tumor-aware loss functions and explicit tumor segmentation mask conditioning. Using a public multicenter dataset and comparing to respective pre-contrast baselines, we observe that subtraction image-based models consistently outperform post-contrast-based models across five complementary evaluation metrics. Apart from assessing the entire image, we also separately evaluate the region of interest, where both tumor-aware losses and segmentation mask inputs improve evaluation metrics. The latter notably enhance qualitative results capturing contrast uptake, albeit assuming access to tumor localization inputs that are not guaranteed to be available in screening settings. A reader study involving 2 radiologists and 4 MRI technologists confirms the high realism of the synthetic images, indicating an emerging clinical potential of generative contrast-enhancement. We share our codebase at https://github.com/sebastibar/conditional-diffusion-breast-MRI.

Grzegorz Skorupko, Fotios Avgoustidis, Carlos Martín-Isla et al.

The nnU-Net framework has played a crucial role in medical image segmentation and has become the gold standard in multitudes of applications targeting different diseases, organs, and modalities. However, so far it has been used primarily in a centralized approach where the collected data is stored in the same location where nnU-Net is trained. This centralized approach has various limitations, such as potential leakage of sensitive patient information and violation of patient privacy. Federated learning has emerged as a key approach for training segmentation models in a decentralized manner, enabling collaborative development while prioritising patient privacy. In this paper, we propose FednnU-Net, a plug-and-play, federated learning extension of the nnU-Net framework. To this end, we contribute two federated methodologies to unlock decentralized training of nnU-Net, namely, Federated Fingerprint Extraction (FFE) and Asymmetric Federated Averaging (AsymFedAvg). We conduct a comprehensive set of experiments demonstrating high and consistent performance of our methods for breast, cardiac and fetal segmentation based on a multi-modal collection of 6 datasets representing samples from 18 different institutions. To democratize research as well as real-world deployments of decentralized training in clinical centres, we publicly share our framework at https://github.com/faildeny/FednnUNet .

Lidia Garrucho, Kaisar Kushibar, Claire-Anne Reidel et al.

Artificial Intelligence (AI) research in breast cancer Magnetic Resonance Imaging (MRI) faces challenges due to limited expert-labeled segmentations. To address this, we present a multicenter dataset of 1506 pre-treatment T1-weighted dynamic contrast-enhanced MRI cases, including expert annotations of primary tumors and non-mass-enhanced regions. The dataset integrates imaging data from four collections in The Cancer Imaging Archive (TCIA), where only 163 cases with expert segmentations were initially available. To facilitate the annotation process, a deep learning model was trained to produce preliminary segmentations for the remaining cases. These were subsequently corrected and verified by 16 breast cancer experts (averaging 9 years of experience), creating a fully annotated dataset. Additionally, the dataset includes 49 harmonized clinical and demographic variables, as well as pre-trained weights for a baseline nnU-Net model trained on the annotated data. This resource addresses a critical gap in publicly available breast cancer datasets, enabling the development, validation, and benchmarking of advanced deep learning models, thus driving progress in breast cancer diagnostics, treatment response prediction, and personalized care.

Lidia Garrucho, Kaisar Kushibar, Socayna Jouide et al.

Computer-aided detection systems based on deep learning have shown great potential in breast cancer detection. However, the lack of domain generalization of artificial neural networks is an important obstacle to their deployment in changing clinical environments. In this work, we explore the domain generalization of deep learning methods for mass detection in digital mammography and analyze in-depth the sources of domain shift in a large-scale multi-center setting. To this end, we compare the performance of eight state-of-the-art detection methods, including Transformer-based models, trained in a single domain and tested in five unseen domains. Moreover, a single-source mass detection training pipeline is designed to improve the domain generalization without requiring images from the new domain. The results show that our workflow generalizes better than state-of-the-art transfer learning-based approaches in four out of five domains while reducing the domain shift caused by the different acquisition protocols and scanner manufacturers. Subsequently, an extensive analysis is performed to identify the covariate shifts with bigger effects on the detection performance, such as due to differences in patient age, breast density, mass size, and mass malignancy. Ultimately, this comprehensive study provides key insights and best practices for future research on domain generalization in deep learning-based breast cancer detection.

Richard Osuala, Kaisar Kushibar, Lidia Garrucho et al.

Despite technological and medical advances, the detection, interpretation, and treatment of cancer based on imaging data continue to pose significant challenges. These include inter-observer variability, class imbalance, dataset shifts, inter- and intra-tumour heterogeneity, malignancy determination, and treatment effect uncertainty. Given the recent advancements in Generative Adversarial Networks (GANs), data synthesis, and adversarial training, we assess the potential of these technologies to address a number of key challenges of cancer imaging. We categorise these challenges into (a) data scarcity and imbalance, (b) data access and privacy, (c) data annotation and segmentation, (d) cancer detection and diagnosis, and (e) tumour profiling, treatment planning and monitoring. Based on our analysis of 164 publications that apply adversarial training techniques in the context of cancer imaging, we highlight multiple underexplored solutions with research potential. We further contribute the Synthesis Study Trustworthiness Test (SynTRUST), a meta-analysis framework for assessing the validation rigour of medical image synthesis studies. SynTRUST is based on 26 concrete measures of thoroughness, reproducibility, usefulness, scalability, and tenability. Based on SynTRUST, we analyse 16 of the most promising cancer imaging challenge solutions and observe a high validation rigour in general, but also several desirable improvements. With this work, we strive to bridge the gap between the needs of the clinical cancer imaging community and the current and prospective research on data synthesis and adversarial networks in the artificial intelligence community.

Jonas Teuwen, Nikita Moriakov, Christian Fedon et al.

The two-dimensional nature of mammography makes estimation of the overall breast density challenging, and estimation of the true patient-specific radiation dose impossible. Digital breast tomosynthesis (DBT), a pseudo-3D technique, is now commonly used in breast cancer screening and diagnostics. Still, the severely limited 3rd dimension information in DBT has not been used, until now, to estimate the true breast density or the patient-specific dose. This study proposes a reconstruction algorithm for DBT based on deep learning specifically optimized for these tasks. The algorithm, which we name DBToR, is based on unrolling a proximal-dual optimization method. The proximal operators are replaced with convolutional neural networks and prior knowledge is included in the model. This extends previous work on a deep learning-based reconstruction model by providing both the primal and the dual blocks with breast thickness information, which is available in DBT. Training and testing of the model were performed using virtual patient phantoms from two different sources. Reconstruction performance, and accuracy in estimation of breast density and radiation dose, were estimated, showing high accuracy (density <+/-3%; dose <+/-20%) without bias, significantly improving on the current state-of-the-art. This work also lays the groundwork for developing a deep learning-based reconstruction algorithm for the task of image interpretation by radiologists.

Basel Alyafi, Oliver Diaz, Joan C Vilanova et al.

Medical image synthesis has gained a great focus recently, especially after the introduction of Generative Adversarial Networks (GANs). GANs have been used widely to provide anatomically-plausible and diverse samples for augmentation and other applications, including segmentation and super resolution. In our previous work, Deep Convolutional GANs were used to generate synthetic mammogram lesions, masses mainly, that could enhance the classification performance in imbalanced datasets. In this new work, a deeper investigation was carried out to explore other aspects of the generated images evaluation, i.e., realism, feature space distribution, and observers studies. t-Stochastic Neighbor Embedding (t-SNE) was used to reduce the dimensionality of real and fake images to enable 2D visualisations. Additionally, two expert radiologists performed a realism-evaluation study. Visualisations showed that the generated images have a similar feature distribution of the real ones, avoiding outliers. Moreover, Receiver Operating Characteristic (ROC) curve showed that the radiologists could not, in many cases, distinguish between synthetic and real lesions, giving 48% and 61% accuracies in a balanced sample set.

Basel Alyafi, Oliver Diaz, Robert Marti

Early detection of breast cancer has a major contribution to curability, and using mammographic images, this can be achieved non-invasively. Supervised deep learning, the dominant CADe tool currently, has played a great role in object detection in computer vision, but it suffers from a limiting property: the need of a large amount of labelled data. This becomes stricter when it comes to medical datasets which require high-cost and time-consuming annotations. Furthermore, medical datasets are usually imbalanced, a condition that often hinders classifiers performance. The aim of this paper is to learn the distribution of the minority class to synthesise new samples in order to improve lesion detection in mammography. Deep Convolutional Generative Adversarial Networks (DCGANs) can efficiently generate breast masses. They are trained on increasing-size subsets of one mammographic dataset and used to generate diverse and realistic breast masses. The effect of including the generated images and/or applying horizontal and vertical flipping is tested in an environment where a 1:10 imbalanced dataset of masses and normal tissue patches is classified by a fully-convolutional network. A maximum of ~ 0:09 improvement of F1 score is reported by using DCGANs along with flipping augmentation over using the original images. We show that DCGANs can be used for synthesising photo-realistic breast mass patches with considerable diversity. It is demonstrated that appending synthetic images in this environment, along with flipping, outperforms the traditional augmentation method of flipping solely, offering faster improvements as a function of the training set size.