The project is based on federated learning, an AI technique that will allow connected hospitals to use an artificial intelligence model for medical image diagnosis without sharing identifiable patient information. Each hospital has a local computing node—based on Cisco UCS servers and third-generation Intel® Xeon® Scalable processors with Intel Software Guard Extensions and integrated AI acceleration—that houses the model, which learns from the center's patients' radiological images. These models, trained with local data, are sent to a central cloud server for processing and combination; all without patient data leaving the hospital and protecting the integrity of medical information. Capgemini Engineering developed this AI through Tessella, its specialized AI unit, enabling healthcare professionals to access advanced diagnostic techniques from any location. Vodafone Spain participates in this project by implementing connectivity from the hospitals to the servers provided by Intel and Cisco through a fully private and secure interconnection. This private Vodafone connectivity allows locally trained models to be securely transmitted to the central server. Gilead Sciences, a pharmaceutical company specializing in virology and a pioneer in developing an effective treatment for COVID-19, has supported this project from its inception to contribute to controlling the pandemic. The project has been funded by Intel and Cisco's Country Digital Acceleration (CDA) program, known as 'Digitaliza' in Spain. If this AI model proves successful, it could be used in the future to assist in the diagnosis of various diseases.
How can engineers improve pulse resolution and cell type discrimination in modern flow cytometry systems?
Flow cytometry is a technology that analyzes cells based on the scattering and fluorescence they produce when passing through lasers of different wavelengths of light. These signals are detected by optical sensors that produce small electrical pulses of current that...
