The Spanish Society of Hematology and Hemotherapy (SEHH), together with Spotlab and the biopharmaceutical company GSK, have presented the MedulAI project in an institutional signing ceremony before hematologists, healthcare managers and innovation managers from the 14 participating hospitals throughout Spain, among which is the Biological Hematology laboratory of the Clinical Laboratory Service of Parc Taulí.
MedulAI is a pioneering initiative in Spain that seeks to improve hematological diagnosis through Artificial Intelligence (AI) applied to bone marrow cytological analysis, which allows for reducing analysis times, improving the statistical robustness of the analysis and reducing the risk of diagnostic errors.
Beyond the presentation of the project, the conference served to formalize the launch of MedulAI and open a space for reflection on the main healthcare challenges facing Hematology of the future and the role that digitalization and artificial intelligence will play in the configuration of the hospital of the future. The meeting highlighted the need to promote a collaborative ecosystem that is capable of accelerating technological and clinical innovation and improving care for hematology patients. In this context, MedulAI is one of the initiatives promoted through GSK Gate2Health, GSK's open innovation program aimed at supporting digital transformation projects in the healthcare field. The participation of 14 centers from seven autonomous communities together with the impetus of the SEHH puts on the table the national nature of MedulAI and the growing interest of the hematology community in the incorporation of digital tools and Artificial Intelligence in clinical practice.
The 14 hospitals that will participate in the deployment and implementation of the project are: the Ramón y Cajal University Hospital, the La Paz University Hospital, the Jiménez Díaz Foundation University Hospital, the Infanta Leonor University Hospital and Hematoclin Médic (collaborating center with the Quirón Madrid University Hospital) (Community of Madrid); the Verge de la Rosada University Hospital and the Reina Sofia University Hospital (Andalusia); the La Nostra Senyora de Sonsoles Hospital in Ávila (Castile and León); the La Nostra Senyora de la Candelaria University Hospital (Canary Islands); the General University Hospital of Elche (Valencian Community); the Catalan Institute of Oncology (ICO) Badalona, the Parc Taulí University Hospital and the Hospital de la Mar (Catalonia) and the Son Llàtzer University Hospital (Balearic Islands).
MedulAI is a clinically validated AI-based system designed to digitize and analyze bone marrow samples. The solution allows conventional microscopes to be adapted to become smart digital microscopes capable of capturing high-quality hematological images and analyzing them using advanced AI algorithms. Its goal is to help specialists improve the accuracy of diagnosing blood diseases such as multiple myeloma, reduce interobserver variability, and significantly optimize analysis times.
The system applies deep learning and computer vision algorithms to digitized images of bone marrow aspirates, automatically identifying and classifying the cells present. This process currently requires the detailed review of thousands of cells by highly specialized professionals, a complex and demanding task that requires time and experience. These tools are designed to provide greater speed, consistency and traceability. MedulAI does not replace the clinical judgment of the hematologist in any case. The tool has been conceived as a support system that provides additional analysis capabilities and facilitates the review of samples, always keeping the specialist ultimately responsible for the diagnostic interpretation. The incorporation of new digital tools only makes sense when it contributes to reinforcing the work of professionals and improving patient care. MedulAI was born precisely with this vocation of supporting hematologists.
The tool used in the project has demonstrated the ability to reduce the time spent on morphological analysis by up to 80%, and standardize the diagnostic process independently of the center or professional by up to 62,5%. In addition, automation allows the analysis of much larger cell volumes than conventional manual procedures. This improves the statistical robustness of the analysis and reduces the risk of diagnostic errors in hematological pathologies such as leukemias, myelodysplastic syndromes, multiple myeloma, lymphomas, myeloproliferative neoplasms or complex bone marrow disorders.
Access to hematological digitization
One of the differentiating aspects of MedulAI is its ability to standardize access to hematological digitization in centers with different levels of technological capacity, favoring a simpler and more scalable adoption. The system adapts to existing microscopes, avoiding large investments in equipment and facilitating its implementation in hospitals of any size and technological capacity. The collaborative platform also allows samples to be analyzed from anywhere, shared between specialists, creating digital repositories for clinical sessions and standardizing analysis processes, among other applications.
In addition, the system has been validated by specialists in Hematology and developed in collaboration with leading hospitals and scientific societies. Its development is based on a solid scientific base that already accumulates more than seven million digitized cells, more than 40.000 analyzed images, more than 1.000 patients included, multicenter validation in leading hospitals and 17 scientific publications in specialized journals and conferences.
