The diagnosis of hematological diseases, such as leukemia and multiple myeloma, heavily relies on the analysis of bone marrow samples, a process that is still performed manually in most hospitals. This procedure is highly time-consuming and requires trained specialists to visually count the cells, a labor-intensive task prone to variations in results depending on the observer’s expertise. The need to optimize this process and reduce variability makes it an ideal candidate for automation.
To address this, a multidisciplinary team from the Center for Biomedical Research in Network (CIBER), in collaboration with the company Spotlab and researchers from the Technical University of Madrid (UPM), Complutense University of Madrid (UCM), and several hospitals led by Hospital Universitario 12 de Octubre, has developed an artificial intelligence system that automates cell counting in bone marrow samples. This system, based on a deep learning algorithm, automatically identifies and counts different types of cells, providing fast and accurate analysis that meets specialists’ needs.
The solution, clinically led by Joaquín Martínez of Hospital 12 de Octubre and CIBERONC, uses AI to capture and analyze images of samples without the need for costly equipment, instead relying on smartphones. This feature not only significantly reduces implementation costs but also facilitates the tool’s usage in any hospital, allowing for a global expansion of this advanced diagnostic technology.
David Bermejo-Peláez, a researcher at CIBER-BBN and Spotlab, explains that the system can digitize images without scanners or complex devices, using smartphones to process samples. This makes the system highly scalable and adaptable to healthcare centers of any level, improving access to advanced diagnostics worldwide.
The benefits of this system extend beyond speeding up analysis. According to María Jesús Ledesma, a researcher at UPM and CIBER-BBN, the algorithm reduces observer variability, providing consistent results and enhancing diagnostic reliability. Additionally, María Linares, a researcher at UCM, highlights that the system increases precision and efficiency in diagnosing hematological diseases, such as leukemia and multiple myeloma, optimizing hematologists’ work and enabling more accurate diagnoses.
This study, funded by the European Union and published in the journal Microscopy and Microanalysis, represents an advancement in the integration of artificial intelligence in clinical practice. The project is also part of the Strategic Program for Economic Recovery and Transformation (PERTE) for Cutting-Edge Health, managed by the Carlos III Health Institute (ISCIII) and the Center for Technological Development and Innovation (CDTI). This research aims to further improve diagnostic accuracy, treatment selection, and prognosis for patients with hematological diseases worldwide.
Article written by MedicalXpress
05/11/2024
Source:
Salud a Diario