SEARCH > News > TECNALIA’ role on synthetic image generation and personalized multi-modal tools

TECNALIA’ role on synthetic image generation and personalized multi-modal tools

25 of August of 2025

image

TECNALIA is the largest centre of applied research and technological development in Spain, a benchmark in Europe and a member of the Basque Research and Technology Alliance. It collaborates with companies and institutions to improve their competitiveness, people's quality of life and achieve sustainable growth, thanks to a team of more than 1,500 people committed to building a better world through technological research and innovation.

Our Vision

To empower industries and communities through transformative solutions that address global challenges. From smart cities to eco-friendly mobility, our work spans critical areas like industrial automation, digital evolution, energy sustainability, health innovation, urban resilience, and the circular economy.

Digital Leadership

In a world defined by connectivity, we focus on three pillars to unlock tomorrow’s opportunities:

  • Intelligence: Harnessing AI, quantum computing, and data-driven insights.
  • Security: Building trust through blockchain, cybersecurity, and robust systems.
  • Value Creation: Redefining business models with smart technologies and IoT ecosystems.

This approach enables us to pioneer a data-centric economy, offering tailored solutions for manufacturers and integrators. Our expertise lies in:

  • Quantum Computing & AI: Pioneering next-gen algorithms and cognitive systems.
  • Industry 4.0 Tools: Robotics, smart maintenance, and worker-centric digital workflows.
  • Connected Technologies: From IoT networks to high-performance computing for Big Data.
  • Sustainable Innovation: Sensors, computer vision, and DevOps for agile, green systems.

TECNALIA’s Contribution to the Project: Advancing AI-Driven Diagnostic Tools

TECNALIA contributes expertise in artificial intelligence (AI) and computer vision to the project, focusing on the synthetic generation of histopathological images. This imaging modality, which provides macroscopic insights into lesions, serves as a cross-cutting element across different use cases, though TECNALIA has deeper experience in the gastrointestinal domain. Pathological images complement other clinical imaging modalities—such as MRI, CT, or colonoscopy—by offering macroscopic views of lesions at varying levels of detail. These modalities, combined with heterogeneous clinical data (e.g., clinical records, omics), enable the development of multimodal diagnostic and prognostic tools that integrate diverse information sources for a comprehensive understanding of diseases. TECNALIA leads the “Personalized Multimodal Data Generation” task, focusing on synthetic multimodal data generation tailored to personalized and group specific scenarios. This work is critical to advancing personalized and equitable healthcare, as it addresses the need for diagnostic and prognostic tools that adapt to diverse, underrepresented clinical contexts.

In addition to synthetic image generation, TECNALIA is evaluating the impact of using synthetic datasets to enhance the accuracy of diagnostic and prognostic models. This includes benchmarking synthetic vs. real data to refine algorithms for tasks like lesion diagnosis or detection. Furthermore, the team is developing multimodal interpretability and explainability strategies to improve transparency in generative, diagnosis and prognosis models to be included as part of future CADx (Computer-Aided Diagnosis) systems. These strategies are critical for ensuring clinicians understand and trust AI-driven decisions.

By combining synthetic data generation, multimodal integration, and explainable AI, TECNALIA is advancing the next generation of diagnostic tools. Multimodal systems are poised to revolutionize medicine by bridging gaps between imaging, clinical data, and patient outcomes, enabling more accurate, personalized, and interpretable healthcare solutions. As a leader in this field, TECNALIA’s work not only addresses current limitations in data availability and model interpretability but also sets a foundation for scalable, future-ready diagnostic frameworks.

Tecnalia SEARCH team: