Irena Galić

Neural-network–based diagnosis from dermatoscopic images is increasingly used for clinical decision support, yet studies report performance disparities across skin tones. Fairness auditing of these models is limited by the lack of reliable skin-tone annotations in public dermatoscopy datasets. We address this gap with neural networks that predict Fitzpatrick skin type via ordinal regression and the Individual Typology Angle (ITA) via color regression, using in-person Fitzpatrick labels and colorimeter measurements as targets. We further leverage extensive pretraining on synthetic and real dermatoscopic and clinical images. The Fitzpatrick model achieves agreement comparable to human crowdsourced annotations, and ITA predictions show high concordance with colorimeter-derived ITA, substantially outperforming pixel-averaging approaches. Applying these estimators to ISIC 2020 and MILK10k, we find that fewer than 1\% of subjects belong to Fitzpatrick types V and VI. We release code and pretrained models as an open-source tool for rapid skin-tone annotation and bias auditing. This is, to our knowledge, the first dermatoscopic skin-tone estimation neural network validated against colorimeter measurements, and it supports growing evidence of clinically relevant performance gaps across skin-tone groups.

Semantic segmentation is a critical task in computer vision, enabling dense classification of image regions. This work investigates the effectiveness of the DeepLabv3+ architecture for binary semantic segmentation using annotated image data. A pretrained ResNet-101 backbone is employed to extract deep features, while Atrous Spatial Pyramid Pooling (ASPP) and a decoder module refine the segmentation outputs. The dataset provides per-image annotations indicating class presence, which are leveraged to approximate segmentation masks for training purposes. Various data augmentation techniques and training strategies were applied to support effective learning and reduce overfitting. Experimental results on the MHIST dataset show that the proposed pipeline achieves strong performance despite the lack of pixel-level annotations, with a mean Intersection-over-Union (mIoU) of 0.76 and a mean Dice coefficient of 0.84. These confirm the potential of weakly supervised segmentation using class-aware CAMs and deep pretrained encoders for structured pixel-level prediction tasks in medical imaging.

Diffusion models have recently emerged as a new paradigm to solve inverse problems in magnetic resonance imaging (MRI). This paper systematically categorizes forty representative studies that use diffusion models for MRI reconstruction. The studies are systematically divided into three dimensional categories: (1) model type distribution that includes standard, score-based diffusion, hybrid and physics-informed diffusion, latent diffusion and cold diffusion; (2) application domain distribution that includes accelerated, motion-corrected, dynamic, quantitative and cross-modal MRI; and (3) domain of reconstruction distribution that includes image-space, k-space, latent-space and hybrid approaches. In this context, diffusion-driven methods represent a significant methodological shift, evolving from purely data-driven denoising toward unified, physics-aligned reconstruction frameworks. Hybrid and cold diffusion approaches further integrate probabilistic generative priors with explicit MRI acquisition physics, which enables improved fidelity, artifact suppression and enhanced generalization across sampling regimes and anatomical domains. This study reviews current advances in the field of MRI reconstruction methods and gives future research directions, emphasizing the need for interpretable, data-efficient and multi-domain generative models that are capable of robustly addressing the diverse challenges of MRI reconstruction.

Training deep learning models for semantic segmentation requires substantial annotated datasets, which are often limited in diversity and size. This study investigates the effectiveness of using diffusion-generated synthetic data to enhance crosswalk segmentation models for assistive navigation systems serving visually impaired individuals. We hypothesized that a Stable Diffusion model fine-tuned on a small collection of 150 first-person view (FPV) images, predominantly captured in sunny conditions, could generate valuable training data representing diverse weather scenarios not present in the original dataset. To test this hypothesis, we created 1500 synthetic images with varying weather conditions using our fine-tuned Stable Diffusion model. We then trained several U-Net-based semantic segmentation models on different combinations of synthetic and original images. For evaluation, we developed a new high-quality test dataset containing 300 annotated real-world images captured across diverse weather conditions. Our results confirmed that models trained on a combination of real and synthetic images significantly outperformed those trained exclusively on the limited real-world dataset. The mixed-data model demonstrated superior generalization to weather conditions absent from the original training data, achieving higher performance metrics across all test scenarios. These findings underscore the potential of diffusion models for data augmentation when authentic datasets are constrained. The pipeline we propose offers an efficient, scalable, and adaptable approach applicable to other domains requiring synthetic data enhancement.

Celiac disease, a chronic autoimmune condition, manifests in those genetically prone to it through damage to the small intestine upon gluten consumption. This condition is estimated to affect approximately one in every hundred individuals worldwide, though it often goes undiagnosed. The early and accurate diagnosis of celiac disease (CD) is critical to preventing severe health complications, with computer-aided diagnostic approaches showing significant promise. However, there is a shortage of review literature that encapsulates the field’s current state and offers a perspective on future advancements. Therefore, this review critically assesses the literature on the role of imaging techniques, biomarker analysis, and computer models in improving CD diagnosis. We highlight the diagnostic strengths of advanced imaging and the non-invasive appeal of biomarker analyses, while also addressing ongoing challenges in standardization and integration into clinical practice. Our analysis stresses the importance of computer-aided diagnostics in fast-tracking the diagnosis of CD, highlighting the necessity for ongoing research to refine these approaches for effective implementation in clinical settings. Future research in the field will focus on standardizing CAD protocols for broader clinical use and exploring the integration of genetic and protein data to enhance early detection and personalize treatment strategies. These advancements promise significant improvements in patient outcomes and broader implications for managing autoimmune diseases.

Artificial intelligence (AI) advancements, especially deep learning, have significantly improved medical image processing and analysis in various tasks such as disease detection, classification, and anatomical structure segmentation. This work overviews fundamental concepts, state-of-the-art models, and publicly available datasets in the field of medical imaging. First, we introduce the types of learning problems commonly employed in medical image processing and then proceed to present an overview of commonly used deep learning methods, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and generative adversarial networks (GANs), with a focus on the image analysis task they are solving, including image classification, object detection/localization, segmentation, generation, and registration. Further, we highlight studies conducted in various application areas, encompassing neurology, brain imaging, retinal analysis, pulmonary imaging, digital pathology, breast imaging, cardiac imaging, bone analysis, abdominal imaging, and musculoskeletal imaging. The strengths and limitations of each method are carefully examined, and the paper identifies pertinent challenges that still require attention, such as the limited availability of annotated data, variability in medical images, and the interpretability issues. Finally, we discuss future research directions with a particular focus on developing explainable deep learning methods and integrating multi-modal data.