AI-Driven Prediction for Biomedical Data Analysis using Image Processing, Graph Signal Processing and Computational Biology

This workshop focuses on the pivotal role of artificial intelligence (AI) in the fields of a pioneering integration of computational biology, bioinformatics, medical image processing, and graph signal processing (GSP) for healthcare services.

Participants in this workshop are expected to present their research how AI based processing generalizes traditional signal processing to irregular graph structures, enabling advanced analysis of biological networks derived from genomics, proteomics, and multimodal medical imaging data.

Methodological Focus & Applications

The fusion of AI techniques such as deep learning, reinforcement learning, NLP-based models extract insights from biomedical literature and genomic sequences. This holistic approach bridges fostering innovations in multiomics integration and 3D biomolecular structure prediction on geometric data. Key areas of exploration include:

• Generative Frameworks: Models such as diffusion models or GANs not only synthesize realistic patient data and molecular structures, but also tackle real-world challenges in precision medicine i.e. from single-cell sequencing diversity mapping to anomaly detection in neuroimaging graphs.
• Predictive Analytics: Through interactive sessions, we will explore the capabilities of AI algorithms in processing and predicting treatment outcomes and devising the most efficacious treatment combinations, underscoring the contribution of AI to more accurate and individualized patient care.
• Emerging Integrations: NLP-enhanced graph-based fusion of MRI/CT images with genomic signals for robust diagnostics, generative model-driven augmentation of sparse datasets, quantum-enhanced simulations of biological systems, and ethical AI frameworks for biotech.

(We invite original and unpublished papers in the following areas, but are not restricted to them)

• Computational Biology: AI-driven protein folding and dynamics simulation; Systems biology modeling with graph neural networks; Quantum computing for molecular interaction prediction.
• Bioinformatics: Multi-omics data fusion and analysis pipelines; Single-cell RNA-seq graph clustering; Ethical AI for genomic privacy preservation.
• Medical Image Processing: Multimodal MRI/CT segmentation with CNNs; Transfer learning for low-data diagnostic imaging; Anomaly detection in radiological graph signals.
• Graph Signal Processing: Graph Fourier transforms for biological networks; Polynomial graph filters for denoising; Sampling and reconstruction on irregular domains.
• NLP-Based Models: BioBERT for literature mining and hypothesis generation; Sequence-to-graph NLP for genomic annotation; Transformer models for biomedical knowledge graphs.
• Generative Models: GANs for synthetic medical image augmentation; Diffusion models for molecular structure generation; Variational autoencoders for patient data synthesis.