Background: Multiple sclerosis is often a longitudinal disease continuum with an initial relapsing-remitting phase (RRMS) and later secondary progression (SPMS). Most currently approved therapies are not sufficiently effective in SPMS. Early detection of SPMS conversion is therefore critical for therapy selection. Important decision-making tools may include testing of partial cognitive performance and magnetic resonance imaging (MRI). Aim of the work: To demonstrate the importance of cognitive testing and MRI for the prediction and detection of SPMS conversion. Elaboration of strategies for follow-up and therapy management in practice, especially in outpatient care. Material and methods: Review based on an unsystematic literature search. Results: Standardized cognitive testing can be helpful for early SPMS diagnosis and facilitate progression assessment. Annual use of sensitive screening tests such as Symbol Digit Modalities Test (SDMT) and Brief Visual Memory Test-Revised (BVMT-R) or the Brief International Cognitive Assessment for MS (BICAMS) test battery is recommended. Persistent inflammatory activity on MRI in the first three years of disease and the presence of cortical lesions are predictive of SPMS conversion. Standardized MRI monitoring for features of progressive MS can support clinically and neurocognitively based suspicion of SPMS. Discussion: Interdisciplinary care of MS patients by clinically skilled neurologists, supported by neuropsychological testing and MRI, has a high value for SPMS prediction and diagnosis. The latter allows early conversion to appropriate therapies, as SPMS requires different interventions than RRMS. After drug switching, clinical, neuropsychological, and imaging vigilance allows stringent monitoring for neuroinflammatory and degenerative activity as well as treatment complications.

Brain tumors are a primary factor causing cancer-related deaths globally, and their classification remains a significant research challenge due to the variability in tumor intensity, size, and shape, as well as the similar appearances of different tumor types. Accurate differentiation is further complicated by these factors, making diagnosis difficult even with advanced imaging techniques such as magnetic resonance imaging (MRI). Recent techniques in artificial intelligence (AI), in particular deep learning (DL), have improved the speed and accuracy of medical image analysis, but they still face challenges like overfitting and the need for large annotated datasets. This study addresses these challenges by presenting two approaches for brain tumor classification using MRI images. The first approach involves fine-tuning transfer learning cutting-edge models, including SEResNet, ConvNeXtBase, and ResNet101V2, with global average pooling 2D and dropout layers to minimize overfitting and reduce the need for extensive preprocessing. The second approach leverages the Vision Transformer (ViT), optimized with the AdamW optimizer and extensive data augmentation. Experiments on the BT-Large-4C dataset demonstrate that SEResNet achieves the highest accuracy of 97.96%, surpassing ViT’s 95.4%. These results suggest that fine-tuning and transfer learning models are more effective at addressing the challenges of overfitting and dataset limitations, ultimately outperforming the Vision Transformer and existing state-of-the-art techniques in brain tumor classification.