Search Results (105)

Background/Objectives: Breast cancer surgery can induce postural adaptations that influence spinal alignment. Although previous studies have predominantly addressed coronal deformities such as scoliosis, changes in sagittal alignment, particularly thoracic kyphosis, remain poorly characterized. An increased kyphotic angle has been associated with impaired physical function, emphasizing the importance of assessing sagittal spinal changes after surgery. Although thoracic kyphosis is traditionally measured on standing radiographs, it can also be evaluated using supine chest computed tomography (CT), routinely obtained during postoperative follow-up. This study investigated the impact of breast cancer surgery on thoracic kyphotic angle and assessed the correlation between kyphotic angles measured on supine CT and standing X-rays. Methods: This retrospective study included 185 breast cancer patients who underwent multiple chest CT and whole-spine X-rays. The thoracic kyphotic angle was defined as the Cobb angle from T4 to T12 on sagittal images. Changes in kyphotic angle among the three surgical groups were analyzed using linear mixed-effects models adjusted for age, chemotherapy, radiotherapy, and endocrine treatment. The correlation between supine CT and standing X-ray was analyzed using Pearson’s coefficient. Results: The mean follow-up interval was 42.6 ± 6.3 months. The mean angular change was minimal (0.6 ± 2.8°), and no significant effects of time, surgical group, or their interaction were observed. Kyphotic angles measured on standing X-rays strongly correlated with those obtained from supine CT (r = 0.799, p < 0.001), with standing values approximately 7.4° higher. Conclusions: Breast cancer surgery did not significantly alter the thoracic kyphotic angle. Supine CT provides a reliable and practical alternative for the assessment of postoperative kyphosis during follow-up. Full article

Medical image classification in clinical settings is frequently constrained by High-Dimensional, Low-Sample Size (HDLSS) conditions, rendering conventional Support Vector Machines (SVM) geometrically susceptible to the data piling phenomenon. This study proposes a hybrid CNN-DWD framework to address this geometrical instability by integrating multi-architecture convolutional feature extraction with Distance-Weighted Discrimination (DWD). Pre-trained ResNet50 and DenseNet121 backbones act as frozen feature extractors, generating a highly descriptive 3072-dimensional fused representation. To resolve the computational bottleneck of deploying DWD directly on massive feature spaces, a supervised Multi-Layer Perceptron (MLP) bottleneck progressively compresses this space into a 32-dimensional latent manifold. Evaluated across breast ultrasonography, breast mammography, and chest X-ray datasets under varying training allocations, the proposed architecture drastically accelerates DWD training—achieving over a 130-fold speedup. The proposed CNN-MLP-DWD framework demonstrates highly competitive diagnostic performance, achieving 93.16% accuracy on the breast ultrasonography dataset and a macro-AUC of 99.69% on the chest X-ray benchmark, comparing favorably against the evaluated baseline methods. Full article

Breast cancer is the most commonly diagnosed cancer in women, with early detection playing a critical role in clinical outcomes. Mammography remains the standard screening modality, producing X-ray images used to assess mammographic density, a key indicator of the proportion of fibroglandular tissue within the breast. The Breast Imaging-Reporting and Data System (BI-RADS) classification system is widely used to report density across four qualitative categories. High density can obscure malignancies and is independently associated with elevated breast cancer risk. Manual interpretation of mammographic density is prone to subjectivity and inter-observer variability, and supervised learning-based estimation methods trained on subjective labels may reflect this inherent subjectivity. This work proposes an unsupervised framework for quantitative breast density estimation that requires no labeled data in its core pipeline. Expert labels are used exclusively to calibrate post hoc discretization thresholds for binary classification, enabling comparison with supervised methods in the literature. The main contributions include: (i) an adaptive Region of Interest (ROI) extraction algorithm, (ii) a Convolutional Neural Network (CNN) based unsupervised segmentation pipeline tuned for mammographic density separation, (iii) a novel confidence metric for identifying unreliable segmentation outputs, (iv) a label correction mechanism for low-confidence cases, and (v) a confidence-filtered majority voting scheme for per-patient classification. The framework is evaluated on two public datasets, namely DDSM and INbreast, with segmentation performance yielding Silhouette scores exceeding 0.92. Agreement with expert labels reaches 71.43% and 79.28% for DDSM and INbreast, respectively. Image-level clustering quality assessment confirms effective unsupervised labeling, with Silhouette scores averaging 0.57 for DDSM and 0.50 for INbreast. The proposed framework provides a practical and non-subjective model for quantitative breast density estimation, with potential utility as a decision-support tool for radiologists that can be considered in clinical practice after further investigation. Full article

Background/Objectives: Machine learning approaches are widely used in modern medical diagnostics, including cancer detection. The results can be significantly improved by aggregating individual measurements, and appropriate aggregation methods should be established. Methods: We applied various measurement aggregation strategies both before and after machine learning modeling to two datasets of X-ray diffraction images: human breast biopsy samples and canine claw samples. Two classifiers, Random Forest and Logistic Regression, were used to determine classification metrics: the area under the receiver operating characteristic curve (ROC-AUC) and balanced accuracy. Results: We found that all aggregation types improve classification metrics, with aggregation after modeling yielding better performance. Depending on the dataset and approach, either classifier can produce better results. For human breast samples, Random Forest with the logit aggregation strategy provides an ROC-AUC exceeding 0.9. For the canine dataset, both Random Forest with the logit aggregation strategy and Logistic Regression with the median of cancer probabilities achieve an ROC-AUC of about 0.85. Conclusions: We examined several simple, straightforward aggregation methods for patient diagnosis based on multiple measurements per patient and achieved significant improvements in classification metrics. Full article

The development of realistic breast phantoms is critical for the evaluation of imaging systems and quantitative image analysis methods. In this work, breast samples derived from the same digital model were produced using 3D printing technology and evaluated for structural similarity and reproducibility. Four independently manufactured phantoms were imaged using mammography and breast tomosynthesis. Radiomic features were extracted from regions of interest in order to assess inter-phantom variability. The results showed very good agreement between the four printed phantoms. Most first-order and GLCM radiomic features exhibited very low inter-phantom variability, indicating consistent structural and intensity characteristics. Neighborhood-based texture features showed slightly higher variability, reflecting their sensitivity to local structural differences. Fractal and power spectrum analyses also confirmed the high structural similarity of the phantoms. These results indicate that the proposed manufacturing approach can produce reproducible breast imaging phantoms suitable for mammography and tomosynthesis imaging studies, with potential applications in imaging system evaluation and radiomic research. Full article

X-ray imaging of dense breasts and breast implants often suffers from reduced lesion visibility because strong attenuation lowers contrast, while conventional rhodium (Rh) K-edge filtering suppresses part of the high-energy spectral tail. This study presents a Monte Carlo framework for spectroscopic mammography using a voxelated 1 mm thick cadmium telluride (CdTe) sensor and a first-order detector interaction model to evaluate energy-dependent image quality. The model reproduces fluorescence and inter-voxel energy redistribution in CdTe, but not the full detector chain, and remains idealized with respect to charge transport, carrier collection, threshold dispersion, and pile-up. Energy-resolved simulations in the 10–50 keV range were used to compute spectroscopic contrast-to-noise ratio (CNR) curves and to form integrated-spectrum (IS) images for four tested spectra. For the dense-breast calcium hydroxyapatite (HA) speck detection task considered here, and under the present simulation assumptions, replacing the standard 28 kVp + 50 μm Rh spectrum with 28 kVp + 1 mm Al increased the simulated IS image CNR by 23.11%, with an approximately 5% increase in estimated primary-incident air kerma at the phantom entrance plane. Preliminary experimental implant-phantom images were included as a qualitative feasibility check, showing a trend consistent with simulations. Within the limits of this task-specific simulation, the results suggest that preserving the transmitted high-energy tail can improve HA speck visibility for the present 1 mm CdTe photon-counting detector, with the 28 kVp + 1 mm Al spectrum outperforming the other tested cases. Full article

Medical image segmentation represents a fundamental task in medical image analysis, serving as a critical component for accurate diagnosis, treatment planning, and disease monitoring. The emergence of Denoising Diffusion Probabilistic Models (DDPMs) has revolutionized the landscape of generative modeling and recently gained significant attention in medical image analysis. This comprehensive review examines the current state of the art in diffusion models for medical image segmentation, covering theoretical foundations, methodological innovations, computational efficiency strategies, and clinical applications. We analyze recent advances in latent diffusion frameworks, transformer-based architectures, and ambiguous segmentation modeling while addressing the practical challenges of implementing these models in clinical environments. The review encompasses applications across multiple medical imaging modalities including Magnetic Resonance Imaging (MRI), Computed Tomography (CT), ultrasound, and X-ray imaging, providing insights into performance achievements and identifying future research directions. Through systematic analysis of publications mostly from 2019 to 2025, we demonstrate that diffusion models have achieved remarkable progress in addressing fundamental challenges including data scarcity, inter-observer variability, and uncertainty quantification. Notable achievements include inference time being reduced from 91.23 s to 0.34 s for echocardiogram segmentation (LDSeg, Echo dataset), DSC scores up to 0.96 for knee cartilage MRI segmentation, and a +13.87% DSC improvement over baseline methods for breast ultrasound segmentation. This review serves as a comprehensive resource for researchers and clinicians interested in leveraging diffusion models for medical image segmentation, providing a roadmap for future research and clinical translation. Full article

Breast ultrasound imaging is widely used for the detection and characterization of breast abnormalities; however, generating detailed and consistent radiological reports remains a labor-intensive and subjective process. Recent advances in deep learning have demonstrated the potential of automated report generation systems to support clinical workflows, yet most existing approaches focus on chest X-ray imaging and rely on convolutional–recurrent architectures with limited capacity to model long-range dependencies and complex clinical semantics. In this work, we propose a multimodal Transformer-based framework for automatic breast ultrasound report generation that integrates visual and textual information through cross-attention mechanisms. The proposed architecture employs a Vision Transformer (ViT) to extract rich spatial and morphological features from ultrasound images. For textual embedding, pretrained language models (BERT, BioBERT, and GPT-2) are implemented in various encoder–decoder configurations to leverage both general linguistic knowledge and domain-specific biomedical semantics. A multimodal Transformer decoder is implemented to autoregressively generate diagnostic reports by jointly attending to visual features and contextualized textual embeddings. We conducted an extensive quantitative evaluation using standard report generation metrics, including BLEU, ROUGE-L, METEOR, and CIDEr, to assess lexical accuracy, semantic alignment, and clinical relevance. Experimental results demonstrate that BioBERT-based models consistently outperform general domain counterparts in clinical specificity, while GPT-2-based decoders improve linguistic fluency. Full article

Background/Objectives: Full-spine radiography is frequently performed from childhood to adulthood, raising concerns about radiation-induced breast cancer risk. To assess such probabilistic risks as cancer, accurate estimation of equivalent and effective organ doses is essential. The purpose of this study is to investigate X-ray imaging conditions for radiation reduction based on breast organ dose and to evaluate the accuracy of simulation software for dose calculation. Methods: Breast organ doses from full-spine radiography were calculated using the Monte Carlo-based dose calculation software PCXMC. Breast organ doses were estimated under various technical conditions of full-spine radiography (tube voltage, distance, grid presence, and beam projection). Dose reduction methods were explored, and variations in dose and error due to phantom characteristics and photon history number were evaluated. Results: Among the X-ray conditions, the greatest radiation reduction effect was achieved by changing the imaging direction. Changing from the anteroposterior to posteroanterior direction reduced doses by approximately 76.7% to 89.1% (127.8–326.7 μGy) in children and 80.4% to 91.1% (411.3–911.1 μGy) in adults. In addition, the study highlighted how phantom characteristics and the number of photon histories influence estimated doses and calculation error, with approximately 2 × 10⁶ photon histories recommended to achieve a standard error ≤ 2%. Conclusions: Modifying radiographic conditions is effective for reducing breast radiation exposure in patients with scoliosis. Furthermore, to ensure the accuracy of dose calculation software, the number of photon histories must be adjusted under certain conditions and used while verifying the standard error. This study demonstrates how technical modifications, projection selection, and phantom characteristics influence breast radiation exposure, thereby supporting the need for patient-tailored imaging strategies that minimize radiation risk while maintaining diagnostic validity. The findings may be useful in informing radiographic protocols and the development of safer imaging guidelines for both pediatric and adult patients undergoing spinal examinations. Full article

Background: Deep learning (DL)-based medical image classification is becoming increasingly reliable, enabling physicians to make faster and more accurate decisions in diagnosis and treatment. A plethora of algorithms have been developed to classify and analyze various types of medical images. Among them, Convolutional Neural Networks (CNNs) have proven highly effective, particularly in medical image analysis and disease detection. Methods: To further enhance these capabilities, this research introduces MediVision, a hybrid DL-based model that integrates a vision backbone based on CNNs for feature extraction, capturing detailed patterns and structures essential for precise classification. These features are then processed through Long Short-Term Memory (LSTM), which identifies sequential dependencies to better recognize disease progression. An attention mechanism is then incorporated that selectively focuses on salient features detected by the LSTM, improving the model’s ability to highlight critical abnormalities. Additionally, MediVision utilizes a skip connection, merging attention outputs with LSTM outputs along with Grad-CAM heatmap to visualize the most important regions of the analyzed medical image and further enhance feature representation and classification accuracy. Results: Tested on ten diverse medical image datasets (including, Alzheimer’s disease, breast ultrasound, blood cell, chest X-ray, chest CT scans, diabetic retinopathy, kidney diseases, bone fracture multi-region, retinal OCT, and brain tumor), MediVision consistently achieved classification accuracies above 95%, with a peak of 98%. Conclusions: The proposed MediVision model offers a robust and effective framework for medical image classification, improving interpretability, reliability, and automated disease diagnosis. To support research reproducibility, the codes and datasets used in this study have been publicly made available through an open-access repository. Full article

Microwave Breast Imaging (MWBI) is an emerging imaging modality aiming to detect breast lesions, which are dielectrically contrasted against the background healthy tissue, in the microwave frequency spectrum. MWBI holds potential to outperform X-ray mammography’s low sensitivity in young and dense breasts, thus supporting timelier detection of interval cancers, as a supplemental screening or diagnostic imaging method. The specificity of MWBI remains unknown, however, as management of false positives has not been systematically addressed yet. An earlier First-In-Human clinical investigation on 24 symptomatic patients provided proof-of-concept for the Wavelia MWBI sectorized multi-static radar imaging technology, which generates clinically meaningful 3D images of the breast, performs semi-automated detection of breast lesions and extracts diagnostic features to distinguish malignant from benign lesions. This paper focuses on a set of technological upgrades, accessories and data processing modules, designed and implemented in the 2nd generation prototype of Wavelia, to handle the diversity in breast geometry, tissue consistency and deformability, in a larger clinical investigation reporting on the bilateral MWBI scan of 62 patients. The presented add-on modules contribute to enhanced quality of scan and a more valid reference reporting space for the MWBI imaging outputs, with a direct positive impact on overall specificity. Full article

Intraoperative radiotherapy (IORT) is a radiation technique that allows for the delivery of a high radiation dose to the target while preserving the surrounding structures, which can be displaced during the surgical procedure. An important limitation of this technique is the lack of real-time image guidance, which is one of the main achievements of modern radiation therapy because it allows for treatment optimization. IORT can be delivered by low-energy X-rays or by accelerated electrons. The present review describes the most relevant clinical applications for IORT and discusses the potential advantages of using artificial intelligence (AI) to overcome some of the current limitations of IORT. In recent decades, IORT has proven to be an effective treatment in several cancer types. In breast cancer, IORT can be used to deliver a single dose of radiation (partial breast irradiation) or as a boost in high-risk patients. In locally advanced rectal cancer, a single dose to the tumor bed can improve local control and prevent pelvic relapse in primary and recurrent tumors. In sarcomas, IORT enables the delivery of high doses, achieving good functional outcomes with low toxicity in tumors located in the retroperitoneum and extremities. In pancreatic cancer, IORT shows promising results in borderline resectable and unresectable cases. Ongoing technological advances are addressing current challenges in imaging and radiation planning, paving the way for personalized, image-guided IORT. Recent innovations such as CT- and MRI-equipped hybrid operating theaters allow for real-time imaging, which could be used for AI-assisted segmentation and planning. Moreover, the implementation of AI in terms of machine learning, deep learning, and radiomics can improve the interpretation of imaging, predict treatment outcomes, and optimize workflow efficiency. Full article

Background/Objectives: The Wavelia Microwave Breast Imaging (MWBI) technology aims to increase sensitivity in dense breasts, where X-ray mammography is of limited value. Its potential contribution to the reduction in the false positives in breast cancer diagnosis, by developing MWBI image descriptors supporting malignant-to-benign lesion discrimination, is also being investigated. After a First-In-Human (FiH) study with interesting findings on a small dataset of 24 symptomatic breast lesions, an upgraded 2nd prototype of Wavelia was manufactured and tested on a larger and more diverse dataset, including 62 patients and a balanced distribution of malignant and benign symptomatic breast lesions. Methods: A set of technological and methodological evolutions, outlined in this article, was implemented in Wavelia#2 to handle the diversity in larger patient datasets. Multi-modal MWBI imaging is employed to parameterize the interaction mechanisms between the microwaves and the imaged breast at varying geometrical and tissue consistency conditions. MWBI Region-Of-Interest (ROI) extraction and characterization based on multidimensional radiomic feature vectors is implemented to expand the malignant-to-benign lesion diagnostics potential of MWBI compared to the limited scope of the FiH study with Wavelia#1, which employed three specific preselected features. Results: This study demonstrates significant diagnostic accuracy of multiple texture-based and intensity-based features to discriminate between malignant and benign breast lesions with Wavelia#2 MWBI. A phenomenological qualitative assessment of the false positive rate on healthy breasts is also presented for the MWBI technology for the first time. Conclusions: The analysis contributes to the rationalization of the MWBI imaging and image analysis outputs towards standardization, objective interpretability, and ultimate clinical acceptance. Full article

Breast cancer is one of the leading causes of death among women of all ages and backgrounds globally. In recent years, the growing deficit of expert radiologists—particularly in underdeveloped countries—alongside a surge in the number of images for analysis, has negatively affected the ability to secure timely and precise diagnostic results in breast cancer screening. AI technologies offer powerful tools that allow for the effective diagnosis and survival forecasting, reducing the dependency on human cognitive input. Towards this aim, this research introduces a deep meta-learning framework for swift analysis of mammography images—combining a Siamese network model with a triplet-based loss function—to facilitate automatic screening (recognition) of potentially suspicious breast cancer cases. Three pre-trained deep CNN architectures, namely GoogLeNet, ResNet50, and MobileNetV3, are fine-tuned and scrutinized for their effectiveness in transforming input mammograms to a suitable embedding space. The proposed framework undergoes a comprehensive evaluation through a rigorous series of experiments, utilizing two different, publicly accessible, and widely used datasets of digital X-ray mammograms: INbreast and CBIS-DDSM. The experimental results demonstrate the framework’s strong performance in differentiating between tumorous and normal images, even with a very limited number of training samples, on both datasets. Full article

Background/Objectives: Advanced breast cancer patients may develop bone metastases, leading to severe pain, fractures, and mortality. Current treatments have limited efficacy, necessitating targeted therapy approaches. Bone metastatic breast cancer cells secrete PTHrP that promotes tumor growth and bone degradation. Previous PTH/PTHrP antagonists failed clinically due to short half-life and insufficient bone targeting. The present study evaluated the following two novel bone-targeted PTH antagonists fused to a collagen-binding domain (CBD) for enhanced bone localization: PTH(7-33)-CBD and [W2]PTH(1-33)-CBD. Methods: Intra-tibial and intra-iliac breast tumor models in mice were used to evaluate drug efficacy in reducing tumor burden and bone destruction. Results: Bioluminescent imaging, X-ray, and micro-CT analysis revealed that both drugs significantly reduced tumor burden and osteolysis compared to control, with [W2]PTH(1-33)-CBD additionally improving trabecular bone structure. Drug efficacy was confirmed in both intra-tibial and intra-iliac breast tumor models. Conclusions: These findings identify CBD-fused PTH/PTHrP antagonists as a promising therapeutic strategy for breast cancer bone metastases. Full article