Home » Industry & Business » How Smart Collaboration Is Driving the AI Revolution in Health Care
Future of AI

How Smart Collaboration Is Driving the AI Revolution in Health Care

Sponsored by:
Sponsored by:
Fanny Sie

Fanny Sie

Head of Artificial Intelligence and Emerging Technology External Collaborations, Roche

Indu Navar

Indu Navar

Founder, Everything ALS

Doina Precup

Doina Precup

Canada-CIFAR AI Chair, McGill University & Mila

AI with Roche is on a mission to harness the power of AI to deliver better health outcomes to patients worldwide.

Artificial intelligence (AI) is essentially the use of computers to “think” like humans so that machines or devices can perform tasks that normally require human intelligence — such as visual perception, speech recognition, decision-making, and language translation — at a speed that human minds can only fathom. The power of AI is already being harnessed all around us, in applications from personalizing content and space exploration to traffic direction apps. Increasingly, AI is also being utilized in health care delivery.

AI and machine learning solutions have the potential to enable patients, providers, and systems to make better, faster, more informed decisions for accelerated and equitable access to health. One of the key ways that AI can impact our health care system is in gathering and making sense of all the health data we have in a fraction of the time and cost needed today. Big picture, AI can potentially augment clinicians’ knowledge, automate time-consuming processes, and help generate next-generation therapies.

Introducing AI with Roche

AI with Roche (AIR) is a centre of excellence aimed at delivering better health outcomes to people through the discovery and application of AI research. Underpinned by an open and collaborative exchange uniquely supported by Roche expertise, AIR is promoting the development of AI for health care by fostering collaborations and partnerships within Canada and beyond.

“Canada has tremendous academic expertise and talent in AI which may be leveraged to improve health solutions, for more patients, faster,” says Fanny Sie, Head of Artificial Intelligence and Emerging Technology External Collaborations for Roche. “The goal is to bring innovations to market together as an ecosystem of partners. By doing so, we aim to increase access to new and better diagnostic tools, therapeutic agents, and digital health solutions.”

Canada has tremendous academic expertise and talent in AI which may be leveraged to improve health solutions, for more patients, faster.

AI innovations in health care require support from the initial onset of the idea, all the way to a final product, in order to reach their full potential to benefit patients. The road to making health innovations a reality for patients is long and hard, but can be eased by nurturing an ecosystem of active stakeholders focused on collaboration.

Scientist doing research

The power of collaboration

AIR thrives at developing and nurturing this collaborative ecosystem. It’s about building toward a learning health system inclusive of patients, health care providers, and the AI community. The ecosystem will flourish when all parties work together to generate meaningful and higher-quality data, deploy tools to assist in transforming data-insights into action, and build safe spaces for experimentation with new innovations.

AIR supports the Canadian AI ecosystem through its foundational collaborations with the three pan-Canadian AI institutes: the Alberta Machine Intelligence Institute (Amii), the Quebec Artificial Intelligence Institute (Mila), and the Vector Institute (Vector).
By creating “a consortium of like-minded individuals and organizations,” as Sie calls it, AIR is bringing different stakeholders together to solve big challenges. It focuses on broad health challenges, from COVID-19 to amyotrophic lateral sclerosis (ALS) and other rare neurological conditions.

Doctors in a meeting

Community-driven science

“We’re aiming to accelerate solutions to market by creating these trusted and responsible environments,” says Sie. “Everyone is coordinated, accountable, transparent, and inclusive, and all the inputs and outputs are placed in a public forum where everyone can access them.”

One such coordinated initiative is AIR’s ongoing collaboration with Everything ALS, a patient-focused non-profit that aims to bring technological innovations and data science to support efforts, from care to cure, for people with ALS. The organization was founded by Indu Navar, an experienced tech entrepreneur who recently lost her husband to ALS.

“I watched my husband degrade and his disease progress for years because there’s no real method to diagnose ALS,” says Navar. “Our goal with Everything ALS is to build the data infrastructure and to analyze data collected through our patient advocacy group to identify the biomarkers for ALS.” This work will allow health care practitioners to better identify ALS and diagnose patients faster as well as to predict ALS incidence.

Virtual meeting with doctor

Supporting patients directly

“Roche is leading the way because they’re able to collaborate very effectively with patient advocacy groups,” says Navar. “Often, patients are left behind — they’re on a different island than the rest of the community and they become just receivers as opposed to collaborators. This model is very effective because as a patient advocacy group, we’re collaborating with Roche and the AI community directly. Roche brings a lot of resources and know-how and we come in with people’s stories. I always say it’s the heart and the brain — it’s emotion and analytics coming together.”

Responsible and ethical AI principles are a key part of what leads to these types of trusted relationships.

An old man speaking to a doctor

Ethical AI principles

The Roche team makes thoughtful considerations to ensure that health AI is designed and adopted in a manner that’s responsible and that leads to improved translation from research to practice.

“The AI space is dynamic and there is new legislation and policy that is expected to evolve in Canada and around the globe. For this reason it is important for Roche to not only be an active listener and responder, it is important to be a leader in this space,” says Sie. “These are always top of mind for us. There are also ethical principles of responsible AI which are paramount to the development and deployment of trusted solutions within the tech industry. We hold the Montreal Declaration for Responsible AI Development of 2018 in high regard within this domain.”

Through its internal groups and leadership, AIR contributes to the global effort for responsible rules of engagement within its collaborations, clinical implementation practices, and solution design of AI.

Group of doctors in a meeting

Safety and transparency

“In medicine, responsible AI practices are about enabling clinicians to abide by their Hippocratic Oath and humanizing the technology,” says Sie. “To do so, Roche needs to work closely with our regulatory, healthcare delivery and research environment to reassure stakeholders we are developing and implementing innovations in health care settings with the highest standards of safety and privacy prioritized at every step.”As researchers and scientists are at the forefront of innovation, establishing long-term collaborations with the academic community, rooted in responsible and ethical AI practices, can lead to powerful knowledge transfers and the improvement of AI solutions to drive new scientific discoveries.

Doina Precup, Associate Professor at McGill University and Core Academic Member at Mila, has enjoyed partnering with AIR from an academic research perspective. As an AI and machine learning researcher, she’s conducting fundamental research on reinforcement learning and working in particular on machine learning techniques to discover new medicines.

Lab scientist performing tests

Leading-edge AI research

“Discovering new medicines is a very expensive and lengthy process, so the goal of this effort is to try and leverage machine learning and data in order to make the process faster, cheaper, and more reliable,” says Precup. “It’s an interesting problem because it’s like looking for a needle in a haystack — looking for a molecule that has some specific properties — and we can do this by looking at the giant space of all possible molecules.”

Every collaborator and stakeholder holds a vital piece of the puzzle, and in working together, the health care ecosystem has the best chance of ensuring that health AI research, infrastructure, and policy evolve in a way that fosters innovation, ensures AI products are developed responsibly and ethically, and provides equitable access to solutions for all people living in Canada.

Next article