Background: Breast cancer is a heterogeneous disease with distinct molecular subtypes, each requiring tailored therapeutic strategies. Accurate classification of these subtypes is crucial for optimizing treatment and improving patient outcomes. While immunohistochemistry remains the gold standard for subtyping, it is invasive and may not fully capture tumor heterogeneity. Artificial Intelligence (AI), particularly Deep Learning (DL), offers a promising non-invasive alternative by analyzing medical imaging data. Methods: In this study, we propose a multimodal DL model that integrates mammography images with clinical metadata to classify breast lesions into five categories: benign, luminal A, luminal B, HER2-enriched, and triple-negative. Using the publicly available Chinese Mammography Database (CMMD), our model was trained and evaluated on a dataset of 4056 images from 1775 patients. Results: The proposed multimodal approach significantly outperformed a unimodal model based solely on mammography images, achieving an AUC of 88.87% for multiclass classification of these five categories, compared to 61.3% AUC for the unimodal model. Conclusions: These findings highlight the potential of multimodal AI-driven approaches for non-invasive breast cancer subtype classification, paving the way for improved diagnostic precision and personalized treatment strategies.

The accurate segmentation of stroke lesions is crucial for the diagnosis and treatment of stroke patients, as it provides spatial information about affected brain regions and the extent of damage. While conventional manual techniques are time-consuming and prone to errors, advanced deep learning models have shown promising results in medical image segmentation. Recently, several complex architectures, such as vision Transformers and attention-based convolutional neural networks (CNNs), have been introduced for this task. However, the question remains whether such high-level designs are necessary to achieve the best results for all segmentation cases. In this paper, we evaluated the performance of four types of deep models for stroke segmentation: (1) a pure Transformer-based architecture (DAE-Former), (2) two advanced CNN-based models (LKA and DLKA) with attention mechanisms, (3) a hybrid model that incorporates CNNs with Transformers (FCT), and (4) the well-known self-adaptive nnU-Net framework. We examined their performance on two publicly available datasets, ISLES 2022 and ATLAS v2.0, and found that the nnU-Net, with its relatively simple design, achieved the best results among all the models tested. Furthermore, we investigated the impact of an imbalanced distribution of the number of unconnected components in each slice, as a representation of common variability in stroke segmentation. Our findings reveal a potential robustness issue of Transformers to such variability, which may explain their unexpected weak performance. Additionally, the success of nnU-Net underscores the significant impact of pre- and post-processing techniques in enhancing segmentation results, rather than solely focusing on architectural designs. These findings suggest that proposed complex architectures may be task-specific and simpler models with appropriate pre-/post-processing pipeline can be equally or more effective in generalization across different tasks in medical image segmentation.

Extraction and separation of divalent Mn and Zn ions from acidic sulfate solutions containing 0.05 M of each metal ion have been carried out using 0.04 M Na-CYANEX 272 in kerosene. The influences of the following parameters were studied: shaking time, equilibrium pH, extractant concentration, diluents type, and temperature. The extraction percentage of both metal ions increased with increasing equilibrium pH in the range 2.0–4.0 and 2.75–6.0 for Zn(II) and Mn(II), respectively. The separation factor, S_(Zn/Mn), increases with the increase in pH to reach a maximum value of about 24 × 10³ at pH 4.0. Extraction of both metal ions increased with increasing extractant concentration. Aliphatic diluents were preferred compared to aromatic diluents. The thermodynamic parameters (ΔH, ΔS, and ΔG) for Mn(II) and Zn(II) were calculated. The extraction mechanism of both metals by Na-CYANEX 272, stripping from loaded extractant and recycling of extractant used in the extraction process were also investigated.

Objectives: Medical imaging is indispensable for diagnosis, with abdominal imaging playing a pivotal role in generating medical reports and informing clinical decision-making. Recent works in artificial intelligence (AI), particularly in multimodal approaches such as vision-language models, have demonstrated significant potential to enhance medical image analysis by seamlessly integrating visual and textual data. While 2D imaging has been the main focus of many studies, the enhanced spatial detail and volumetric consistency offered by 3D images, such as CT scans, remain relatively underexplored. This gap underscores the need for innovative approaches to unlock the potential of 3D imaging in clinical workflows. Methods: In this study, we utilized a multimodal AI pipeline, Phi3-V, to address 2 key challenges in abdominal imaging: generating clinically coherent medical reports from 3D CT images and performing visual question answering based on these images. Results: Our optimized model attained an average GREEN score of 0.409 for medical report generation and an accuracy of 79% for multiple-choice visual question answering on the validation cases. Conclusions: These findings demonstrate the potential of multimodal AI in advancing the analysis of 3D medical imaging, paving the way for more robust and efficient applications in healthcare. Advances in knowledge: This study advances the use of multimodal AI for 3D CT imaging, achieving improvements in medical report generation and visual question answering.