Trace Id is missing
May 25, 2023

San Raffaele University and Research Hospital transforms clinical research and delivers precision medicine using the Microsoft Intelligent Data Platform

San Raffaele University and Research Hospital is one of the largest of its kind in Italy, as well as a pioneer of cutting-edge translational research and technology adoption at both a national and international level. During the pandemic, the organization saw an opportunity to start using artificial intelligence to better triage incoming symptomatic patients and identify their future reaction to the virus. Fast forward three years, the pilot project has been turned into an intelligent data solution powered by AI – which San Raffaele is going to use to transform clinical research, improve the end-to-end patient journey inside the hospital and accelerate its own move towards precision medicine.

San Raffaele University and Research Hospital

“When you operate in a hospital that’s almost as big as a city, you simply can’t afford to run it without data.”

Carlo Tacchetti, Professor of Human Anatomy at Vita-Salute San Raffaele University, is describing the transformational role of data and technology in helping his organization provide outstanding care to 1.5 million people every year. 

Vita-Salute San Raffaele University is the academic brainpower behind its namesake hospital – the Scientific Institute of Research, Hospitalization and Care (IRCCS) Ospedale San Raffaele – which itself is one of the most distinguished university hospitals and scientific institutes in the country. 

“San Raffaele University and Research Hospital has one key goal,” says Tacchetti. “To treat the highest number of patients possible using cutting-edge and modern solutions”.

"And at the same time, to rationalize the use of resources allocated by the Italian public healthcare system and drive faster diagnosis that contributes to innovative clinical research.”

To support this, the organization is relying on data and AI. Together with Microsoft Industry Solutions (formerly Microsoft Consulting) and its partner Porini, the hospital has embarked on a monumental journey to develop a state-of-the-art platform for data analysis and support in experimental treatments. A platform that’s helping to treat diabetes, lung cancer and more – all while improving patient experience and empowering physicians. 

“The healthcare sector is unlike any other industry, especially when it comes to its relationship with technology,” he says. “There are such high levels of accountability and responsibility, that you can’t just rely on artificial intelligence to make a diagnosis or to perform parts of a clinicians’ job”.

“That’s where the greatness of this project truly lies. We’re not just implementing technology. We’re doing it in a way that is conceived exclusively for patients and clinicians.”  

Crisis leads to opportunity

San Raffaele University and Research Hospital is a global pioneer of digital and cloud solutions implementation in healthcare – responsible for 15% of the entire scientific production of the Italian research hospitals network. With more than 50,000 annual hospitalizations and more than 1100 ongoing clinical trials, the organization is a highly fertile ground for both innovation and experimentation with new technologies.*

“At San Raffaele, we want to empower our scientists, physicians and students to discover new and important therapeutic approaches against life threatening diseases,” comments Dr. Anna Flavia d’Amelio Einaudi, Managing Director at Vita-Salute San Raffaele University. 

Much of that recently focused on the use of AI and cloud technology, due to their enormous potential in data analysis and the generation of insights they enable. “These have so much to offer to healthcare,” says Antonio Esposito, Professor of Radiology at Vita-Salute San Raffaele University.  “We’re increasingly seeing them being used for imaging, genomics, precision medicine and much more. 

“Their involvement has been such that medical research is no longer just a case of ‘bench to bedside’, but rather ‘cloud to bedside’”.

This was something that came keenly into focus when the pandemic reached northern Italy in 2020. “We could divide AI adoption at our hospital between before-Covid and after Covid,” he says. 

“Then, in February 2020, our hospital started opening its doors to more than 20,000 Covid patients and we quickly plunged into a state of emergency. We needed to find fast, deployable new solutions to contain the crisis.” 

More specifically, the hospital needed to be quicker at triaging incoming patients, understanding how severely the disease would develop in the following days and weeks, and identifying the most appropriate form of care for them – from home-based, remotely monitored self-isolation to ICU admission.   

“We wanted to be able to forecast what type of support the patient would need based on a few objective parameters collected on arrival, and we wanted to do it in no more than 30 minutes,” he says.

“It was the start of our partnership with Microsoft and Porini – and the birth of our AI project.”

From crisis management to a new data strategy

Microsoft and Porini were quick to realize the huge benefits that a data-driven approach could have, even beyond the pandemic.

As part of the program, the three entities had managed to collect data from more than 1,500 patients during the first pandemic wave across 16 Italian hospitals. They then created an algorithm capable of segregating patients in three risk classes (low, intermediate, and high risk). This soon proved capable of predicting individual risks of mortality.   

“As we progressed, it became clear we were creating a new solution with the potential to transform the way we diagnose and manage patients at our organization,” says Esposito. 

“We had a great and reliable platform that was allowing us to generate new and sometimes even unexpected insights – and if used in production, to reduce Covid patient hospitalizations by 27%. 

“This persuaded us to extend our pilot project beyond the Covid experience, investing into a data platform that would allow us to make even more of the data we had.”

Inside the platform

San Raffaele’s solution is a secure end-to-end data platform for data collection, collaboration, management, analysis and more. It features an enterprise platform tasked with integrating pseudo-anonymized or anonymized patient data; a data science development environment to perform statistical analysis, develop AI models and answer clinical questions; and a user portal to help researchers to search for, manage and organize clinical research and related AI models.

As result of a tight collaboration between Porini, Microsoft’s Industry Solutions and Cloud for Healthcare teams and the hospital, the platform is designed to help San Raffaele tap into the most up-to-date, industry-specific technologies and competencies that Microsoft has to offer. 

It was created based on security by design principles and the Zero Trust security model with the aim to improve clinicians’ diagnosis and prognosis capabilities, as well as their ability to classify patients and provide personalized therapeutic choices. Ultimately paving way for San Raffaele’s shift towards precision medicine. 

To do this, the solution uses various Microsoft Intelligent Data Platform capabilities, such as Azure Health Data Services, Azure Cognitive Services, Azure Infrastructure, Azure Analytics Services and Azure Machine Learning. It also relies on Text Analytics for Health to extract and label medical information from unstructured texts such as doctors’ notes and electronic health records.

“Our Intelligent Data Platform ensures that all available patient data can be collected from a variety of different sources,” says Eng. Marco Denti, Head of Software Factory and Innovation ICT at Gruppo San Donato (owner of San Raffaele Hospital).

“It is then standardized using FHIR, validated, stored and made available to researchers and clinicians as single source of truth for their analysis and development of new Intelligent solutions, which will answer compelling clinical questions.”
https://www.hsr.it/

The perfect change to reform an outdated ecosystem

There is even more to the platform than its data management and analysis capabilities. 

“For years, we had been relying on a highly heterogenous and challenging hospital information system,” says Marco Denti. “We had hundreds of different applications running in silos that inevitably made the data spread across databases unstandardized.

“This project was a great chance for us to innovate our ecosystem, break the different siloed applications and bring all our data into one single location – ultimately allowing us to better support research activities within the group.”

For Antonello Bianchi, Healthcare and Life Science Solutions Lead at Porini, it was also an opportunity to explore an entirely new approach to AI. “Everything about this solution is customized and tailored to the needs of the hospital and research teams that use it,” he says. 

“This means building models that specifically address the needs of the clinical research, but also providing insights on clinicians day-to-day activities – helping them to keep track of their operational flows.”

Echoing his words, Stefano Brusamolino, Practice Lead Healthcare and Life Science in Porini adds: “Throughout the entire development phase, we made sure that we sat down with the clinicians and built it with the data and features that they provided, following their inputs and creating something that was flexible, compatible with their needs and easy to use.

“In practical terms, that meant making sure that any researcher and doctor using the platform could do so even without being a data scientist – and allow them to perform their work without requiring any particularly technical capabilities. This is where Microsoft Industry Solutions proved especially useful helping them to develop their own Tech Capabilities to allow researchers full conscious autonomy on using AI to improve their effectiveness while reducing time and resources."

“This was the real secret behind our solution’s success.”

Building a legacy for future generations

A prototype of the platform is running on testing use cases to support pilot clinical trials on the treatment of lung and kidney cancer, diabetes, cardiovascular disease and more.   

“Our first attempt at using the platform was lung cancer and the immunotherapy for it,” says Carlo Tacchetti. “We started using the platform to foresee whether a patient eligible for immunotherapy would respond or not to treatment, so that we could provide them with alternative therapies and save them precious time and the costs of undergoing this type of treatment in the first place.”

The number of use cases has evolved since then, with Esposito mentioning particularly important work being done on subjects at risk of myocardial infarction.

“These are subjects that can benefit from medical and behavioral preventive strategies,” he says. “However, models based on traditional cardiovascular risk factors often struggle to predict myocardial infarction during a medium-term follow-up.” 

That’s where he says AI can help, allowing the inclusion of new valuable data in individual risk assessment processes.

With plans to expand to many more areas in the pipeline, Denti declares himself grateful for the newly found flexibility and adaptability that the platform offers. 

“This project has truly transformed our mindset and that of our technicians, clinicians and researchers,” says Denti. “Knowing that we have a platform that is flexible, dynamic, and capable of supporting all our needs is game-changing for us.” 

It’s also a clear signal that the platform is destined to grow, and its legacy needs to be guaranteed. “All we want is for this dream, which we’ve managed to turn into reality, to live on for the years and decades to come,” says Tacchetti. 

“But we can’t do this just on our own forever. That’s why we’re recruiting fresh new minds and launching new university courses and programs designed to forge new skillsets and attract more generations of innovators to our group.” 

Echoing his words, d’Amelio Einaudi concludes: “We feel the responsibility of paving the way of future healthcare not only for our patients, but for other institutions and the entire healthcare ecosystem, and of preparing the healthcare professionals of the future, today”.

“By partnering with Microsoft, we have access to the most crucial technology and valuable resources to achieve that and more.”

 *The information provided in the statement, can be found on the official website of San Raffaele Hospital: 

DISCLAIMER
AI tools and technologies, including GPT-4 and other LLMs, can make mistakes and do not always provide accurate or complete information. Customers/partners need to thoroughly test and evaluate whether the AI tool is fit for purpose and identify and mitigate any risks or harms to end users associated with its use.

Microsoft products and services (1) are not designed, intended or made available as a medical device, and (2) are not designed or intended to be a substitute for professional medical advice, diagnosis, treatment, or judgment and should not be used to replace or as a substitute for professional medical advice, diagnosis, treatment, or judgment. Customers/partners are responsible for ensuring solutions including GPT-4 and other LLMs comply with applicable laws and regulations.

“We feel the responsibility of paving the way of future healthcare not only for our patients, but for other institutions and the entire healthcare ecosystem. By partnering with Microsoft, we have access to most crucial technology and valuable resources to achieve that and more.”

Dr. Anna Flavia d’Amelio Einaudi, Managing Director, Vita-Salute San Raffaele University

Take the next step

Fuel innovation with Microsoft

Talk to an expert about custom solutions

Let us help you create customized solutions and achieve your unique business goals.

Drive results with proven solutions

Achieve more with the products and solutions that helped our customers reach their goals.

Follow Microsoft