Gloria Mercedes Díaz

Background: Breast cancer molecular subtypes (Luminal A, luminal B, HER2-positive, and triple-negative) are typically determined through biopsy and immunohistochemistry, procedures that are invasive and prone to interpretative variability. Artificial intelligence applied to medical imaging has emerged as a non-invasive alternative to support subtype prediction. However, methodological rigor, risk of bias and clinical translatability of existing studies remain insufficiently characterized. Methods: We conducted a systematic review in accordance with the PRISMA 2020 guidelines. Searches were performed in Scopus, Web of Science, ScienceDirect, and PubMed to identify English-language, peer-reviewed original articles published between January 1, 2020, and March 25, 2025. Two reviewers independently screened titles and abstracts using Rayyan, retrieved potentially eligible full-text articles, and assessed risk of bias and applicability using an AI-adapted QUADAS framework based on QUADAS-2. Results: From 626 identified records, 76 full-text articles were assessed for eligibility. The AI-adapted QUADAS framework showed that 67 of these 76 articles were excluded because they presented a high or unclear risk of bias in at least one evaluated domain. Consequently, only 9 studies were judged to have a low risk of bias and were included in the qualitative synthesis. The included evidence was dominated by studies using dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), with additional contributions from ultrasound-based approaches. The methodological strategies ranged from radiomics-based machine learning models to deep learning architectures. Discussion: Integrating radiomic signatures with clinical predictors may improve performance in selected clinical scenarios, and peri-tumoral information alongside delayed DCE-MRI phases can contribute complementary diagnostic value. Nevertheless, the small number of low-risk studies highlights the need for standardized acquisition and preprocessing reporting, patient-level data splitting, robust external validation, and broader data availability to strengthen the clinical translation and generalizability of AI-based molecular subtyping.

Background and objective: Contrast-enhanced spectral mammography (CESM) is a new imaging modality that integrates digital mammography and the use of iodinated contrast agents in a dual-energy acquisition protocol, facilitating the differentiation between benign and malignant breast lesions. However, its clinical use is constrained by the potential risks associated with iodinated contrast agents and the increased radiation exposure. Method: The proposed strategy adapts a self-calibrated pixel attention (SCPA) mechanism to enhance the ability of conditional image-to-image translation models to synthesize contrast uptake in breast lesions. To assess the generalizability and feasibility of our approach under realistic data variations, we used two independent CESM datasets. The performance of the model was evaluated through quantitative and qualitative analyses, considering both whole-image performance and contrast-uptake regions. The results were compared with state-of-the-art conditional generative models. Furthermore, the clinical reliability of the images synthesized by the proposed strategy was assessed through a qualitative analysis performed by an expert breast radiologist. Results: The proposed strategy outperformed state-of-the-art approaches in the synthesis of both whole images and contrast-enhanced regions. In particular, the SCPA-conditional generative adversarial network exhibited superior preservation of the morphology of the lesion while maintaining the overall quality of the generated images. Moreover, the clinical feasibility evaluation showed acceptable realism and technical quality suitable for radiologist interpretation, although in some cases, residual synthesis errors were observed that could affect patient management. Conclusion: The integration of attention mechanisms substantially improves the synthesis of contrast uptake in regions of interest (ROIs) within CESM studies with limited and heterogeneous datasets. These findings demonstrate the feasibility of generating contrast enhancement from non-contrast mammography images, highlighting the potential of this approach to reduce the need for contrast administration while preserving clinically relevant diagnostic information. This work should therefore be regarded as a feasibility study that establishes a clinically aligned ROI-focused framework for CESM virtual contrast synthesis and provides a foundation for future work.

La inteligencia artificial, esa tecnología que pareciera haber emergido de repente hace apenas unos años, pero que lleva varias décadas de desarrollo, se ha convertido en uno de los temas de mayor interés a nivel mundial. Una consulta en la muy conocida herramienta de Google Trends (https://trends.google.com), nos muestra cómo la búsqueda del concepto alcanzó un máximo en el mes de mayo de 2023, tan sólo un mes después del lanzamiento de GPT-4. Este modelo de inteligencia artificial generativa es capaz de simular conversaciones humanas, en respuesta a preguntas específicas, y permite incluso crear documentos completos con una legibilidad similar a la de los textos creados por humanos. Ese mismo mes, también fue el lanzamiento de la actualización conocida como GPT-4o, la cual mostró al mundo la capacidad para la interpretación y generación realista de texto, imágenes, vídeos y voz, en más de 50 idiomas.

La evaluación de la investigación es actualmente uno de los temas de mayor relevancia y controversia en la ciencia y la academia, especialmente en el ámbito universitario. Esta evaluación se utiliza comúnmente para clasificar instituciones, grupos, productos académicos como las revistas científicas, y al personal docente investigador. Los indicadores de calidad e impacto de la investigación influyen decisivamente en el ascenso en carreras académicas, selección de beneficiarios de becas e incentivos, culminación de programas de alto nivel (como maestrías y doctorados), asignación de recursos en instituciones y centros de investigación, y en la definición de políticas públicas a nivel institucional y gubernamental [1], [2].

Objective: Currently, mammography and ultrasonography are the most used imaging techniques for breast cancer screening. However, these examinations report many indeterminate studies with a low probability of being malignant, i.e., BIRADS 3 and 4A. This prospective study aims to evaluate the value of breast magnetic resonance imaging (MRI) to clarify the BIRADS categorization of indeterminate mammography or ultrasonography studies. Methods: MRI studies acquired prospectively from 105 patients previously classified as BIRADS 3 or 4A were analyzed independently by four radiologists with different experience levels. Interobserver agreement was determined by the first-order agreement coefficient (AC1), and divergent results were re-analyzed for consensus. The possible correlation between the MRI and the mammography/ultrasound findings was evaluated, and each study was independently classified in one of the five BIRADS categories (BIRADS 1 to 5). In lesions categorized as BIRADS 4 or 5 at MRI, histopathological diagnosis was established by image-guided biopsy; while short-term follow-up was performed in lesions rated as BIRADS 3. Results: Breast MRI was useful in diagnosing three invasive ductal carcinomas, upgraded from BIRADS 4A to BIRADS 5. It also allowed excluding malignancy in 86 patients (81.9%), avoiding 22 unnecessary biopsies and 64 short-term follow-ups. The MRI showed good diagnostic performance with the area under roc curve, sensitivity, specificity, PPV, and NPV of 0.995, 100%, 83.5%, 10.5%, and 100%, respectively. Conclusions: MRI showed to be useful as a problem-solving tool to clarify indeterminate findings in breast cancer screening and avoiding unnecessary short-follow-ups and percutaneous biopsies.