Scientific Research Publishing’s Post

【From U-Net to Swin-Unet Transformers: The Next-Generation Advances in Brain Tumor Segmentation with Deep Learning】 Full article: https://lnkd.in/gxurGNGB (Authored by Mushtaq Mahyoob Saleh and Bharat B. Biswal, from University of Electronic Science and Technology of China, China.) #Brain_tumor_segmentation is a vital step in diagnosis, treatment planning, and prognosis in neuro-oncology. In recent years, advancements in machine learning (ML) and #deep_learning (DL) have revolutionized segmentation accuracy. This review paper comprehensively surveys the evolution of brain tumor segmentation techniques, emphasizing the transition from conventional U-Net models to cutting-edge Swin UNET transformer architectures, discusses the impact of novel activation functions on improving gradient stability and segmentation accuracy, addresses ongoing challenges such as data heterogeneity, real-time clinical applicability, and integration barriers, and proposes future directions for developing robust, interpretable, and scalable brain tumor segmentation systems. #Vision_Transformers  Additionally, if you have any new manuscripts ready for submission, please feel free to send them directly as an attachment to the email address below. As a valued contributor, you will enjoy priority access to discounted publication fees and regular updates on the progress of your manuscript throughout the review process. Email: Hellen Wang <kelseytan.scirp@gmail.com> Hellen Wang <wqs0823@gmail.com>

DO YOU KNOW? InBrain Neuroelectronics S.L. is pioneering next-generation neural interfaces through ultra-thin, flexible graphene electrode arrays engineered for high-fidelity bidirectional communication with the brain. These devices achieve sub-millimeter resolution in both signal acquisition and stimulation, addressing limitations of conventional metal electrodes such as rigidity, biofouling, and signal degradation. By leveraging machine learning and AI pipelines, InBrain’s platform decodes complex neural dynamics and adapts stimulation in real time, enabling closed-loop therapies. Current applications span epilepsy, Parkinson’s disease, and intraoperative neuromodulation for brain tumor resections, with broader implications for adaptive neuroprosthetics and brain-computer interfaces. Founded in 2019 as a spinoff from the Catalan Institute of Nanoscience and Nanotechnology and the Catalan Institution for Research and Advanced Studies, InBrain closed a $50M Series B in October 2024, bringing its cumulative funding to $68M. Explore more at https://lnkd.in/g5rP9Ea6 .

🚀 Just published a new project on GitHub! Brain Tumor MRI Classification 🧠📊 I built a deep learning model that classifies brain MRI scans into tumor vs. non-tumor, aiming to support early detection with AI. The notebook includes data preprocessing, CNN training, evaluation, and visualization. 🔗 Check it out here: https://lnkd.in/gunjkf3X #AI #DeepLearning #MedicalAI #MachineLearning #ComputerVision

Symmetry Interactive Transformer with CNN Framework for Diagnosis of Alzheimer's Disease Using Structural MRI https://lnkd.in/e7MwFH9E Structural magnetic resonance imaging (sMRI) combined with deep learning has achieved remarkable progress in the prediction and diagnosis of Alzheimer's disease (AD). Existing studies have used CNN and transformer to build a well-performing network, but most of them are based on pretraining or ignoring the asymmetrical character caused by brain disorders. We propose an end-to-end network for the detection of disease-based asymmetric induced by left and right brain atrophy which consist of 3D CNN Encoder and Symmetry Interactive Transformer (SIT). Following the inter-equal grid block fetch operation, the corresponding left and right hemisphere features are aligned and subsequently fed into the SIT for diagnostic analysis. SIT can help the model focus more on the regions of asymmetry caused by structural changes, thus improving diagnostic performance. We evaluated our method based on the ADNI dataset, and the results show that the method achieves better diagnostic accuracy (92.5\%) compared to several CNN methods and CNNs combined with a general transformer. The visualization results show that our network pays more attention in regions of brain atrophy, especially for the asymmetric pathological characteristics induced by AD, demonstrating the interpretability and effectiveness of the method. --- Newsletter https://lnkd.in/emCkRuA More story https://lnkd.in/enY7VpM LinkedIn https://lnkd.in/ehrfPYQ6 #AINewsClips #AI #ML #ArtificialIntelligence #MachineLearning #ComputerVision

In a recent study, a team from the Shenzhen Institutes of Advanced Technology of the Chinese Academy of Sciences developed and validated a label-free, non-invasive method combining AFM with deep learning for accurate profiling of human macrophage mechanophenotypes and rapid identification of their polarization states >>> https://lnkd.in/eKqVpyxT

Enhancing Corpus Callosum Segmentation in Fetal MRI via Pathology-Informed Domain Randomization https://lnkd.in/ezb5rUQ2 Accurate fetal brain segmentation is crucial for extracting biomarkers and assessing neurodevelopment, especially in conditions such as corpus callosum dysgenesis (CCD), which can induce drastic anatomical changes. However, the rarity of CCD severely limits annotated data, hindering the generalization of deep learning models. To address this, we propose a pathology-informed domain randomization strategy that embeds prior knowledge of CCD manifestations into a synthetic data generation pipeline. By simulating diverse brain alterations from healthy data alone, our approach enables robust segmentation without requiring pathological annotations. --- Newsletter https://lnkd.in/emCkRuA More story https://lnkd.in/enY7VpM LinkedIn https://lnkd.in/ehrfPYQ6 #AINewsClips #AI #ML #ArtificialIntelligence #MachineLearning #ComputerVision

The thalamus, being a crucial region for sensory processing and multi-sensory integration with both bottom-up and top-down connectivity, is in a powerful anatomical and functional position to guide cortical processes. While past thalamic research has focused mostly on pure sensory processing (i.e., their relay role), the importance of non-relay functions (e.g., cognition) has recently been emphasized, but much less understood. We are especially interested in the anterior thalamic nuclei (ATN) and its cell types, which have been implicated in long-term memory formation, attention, and spatial navigation. In this talk, I will describe our work aimed at developing tools to genetically access ATN cell types with high specificity, investigate their role in contextual learning and generalization, and how computational models along with in vivo calcium imaging is helping us uncover similarities vs. differences between the ATN and the broader cognitive network (i.e., cortex and hippocampus). We believe that these complementary approaches have the potential to reveal novel cell type- and functionally-distinct subnetworks within the ATN, which underlie cognitive functions.

We are pleased to share that our latest article has been published in CAUCHY: Journal of Pure and Applied Mathematics, titled An Explainable Deep Learning Approach for Brain Tumor Detection Using MobileNet and Grad-CAM Visualization. This study explores deep learning approaches for brain tumor detection using MRI images, addressing one of the key challenges in medical AI: the lack of interpretability. By combining MobileNet with Grad-CAM visualization, the proposed framework not only ensures efficiency and speed but also provides clinical transparency through visual explanations of the model’s predictions. When compared to other architectures, MobileNet demonstrated superior performance, offering a better balance of computational efficiency and interpretability. This research highlights the potential of lightweight models in supporting real-world clinical applications where accuracy, speed, and trust are equally important. This work is a collaboration with Amalan Fadil Gaib, Safrizal Ardana Ardiyansa, Eric Julianto, Ngurah Bagus, and Ando Zamhariro Royan. We are grateful for the contributions of everyone involved in making this publication possible. 📖 Read the full article here: https://lnkd.in/gcy5psPq

As federal funding for basic research faces steep cuts, UC scientists are pushing the boundaries of brain-computer interfaces: restoring speech after ALS, easing Parkinson’s symptoms, and improving bionic vision with AI (that's us 👋🏻 at UC Santa Barbara). This is what public science makes possible: decades of federally funded research laying the groundwork for life-changing technologies. 🧠 Read the full article from UC News: https://lnkd.in/g9CeAwVJ #Research #Neuroscience #AI #BCI #BionicVision #NeuroTeach

Self-Supervised Cross-Encoder for Neurodegenerative Disease Diagnosis https://lnkd.in/e86fhTFg Deep learning has shown significant potential in diagnosing neurodegenerative diseases from MRI data. However, most existing methods rely heavily on large volumes of labeled data and often yield representations that lack interpretability. To address both challenges, we propose a novel self-supervised cross-encoder framework that leverages the temporal continuity in longitudinal MRI scans for supervision. This framework disentangles learned representations into two components: a static representation, constrained by contrastive learning, which captures stable anatomical features; and a dynamic representation, guided by input-gradient regularization, which reflects temporal changes and can be effectively fine-tuned for downstream classification tasks. Experimental results on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset demonstrate that our method achieves superior classification accuracy and improved interpretability. Furthermore, the learned representations exhibit strong zero-shot generalization on the Open Access Series of Imaging Studies (OASIS) dataset and cross-task generalization on the Parkinson Progression Marker Initiative (PPMI) dataset. The code for the proposed method will be made publicly available. --- Newsletter https://lnkd.in/emCkRuA More story https://lnkd.in/enY7VpM LinkedIn https://lnkd.in/ehrfPYQ6 #AINewsClips #AI #ML #ArtificialIntelligence #MachineLearning #ComputerVision