Collaborating for a Cause
How Boston Scientific and IBM Research joined forces to use AI to help patients living with chronic pain
Maryam Mahdi | | 6 min read
Boston Scientific and IBM Research are collaborating to create an AI-based platform that provides doctors with personalized insights into patients’ experiences with chronic pain – measuring and analyzing those experiences in a way that would not be possible without AI. By collating data from spinal cord stimulation therapy studies, the team are developing an approach that will “objectively and continuously measure chronic pain” in a bid to move medical assessment and treatment beyond arbitrary descriptors of pain level or strength.
We caught up with Rafael Carbunaru, Vice President of Research and Development, Neuromodulation, Boston Scientific, and Jeff Rogers, Global Research Leader for Digital Health at IBM Research, to talk about the platform and the role AI will play in the future treatment landscape for chronic pain.
Why is chronic pain such a complex problem for healthcare professionals to solve?
Carbunaru: For patients living with pain, finding the right treatment is a journey. The road to effective treatment can be further complicated for many as they are not aware of – or cannot access – the services and expertise of physicians and other healthcare professionals who specialize in the field.
Pain is also a subjective experience. So, when patients do access healthcare support, they may find it difficult to accurately describe what they feel. For example, if a physician asked a patient to rate their pain on a scale of 0 to 10, they may choose a number that doesn't accurately reflect their experience. It is also important to remember that with time and changing circumstances, the same patient may rate their pain very differently, despite the fact they still have to manage the same healthcare challenges.
This complexity leads to issues for the healthcare provider who may struggle to know when to intervene or when the patient has a good medical regime. The problem is further compounded when we acknowledge that patients may only visit their doctors a handful of times a year.
If patient data could be quantitatively addressed to provide a holistic picture of the patient landscape, we would be able to provide better support for all parties involved and help reduce the burden of chronic pain in society.
What’s the story behind the partnership?
Carbunaru: For the past 18 years, we’ve been able to create therapeutics for chronic pain. Throughout this development, we remained in close contact with both patients and physicians. But we came to understand that managing pain is a unique therapeutic area. Pain is more than just a sensation; it affects many aspects of patients’ lives.
Our goal was to help the patients who are living with this huge burden and help them achieve better health. And that’s why we partnered with Jeff and his team at IBM Research. By combining Boston Scientific’s experience in neuro-engineering, clinical studies, and patient contact with IBM’s data science and AI know-how, we aimed to better understand the needs and experiences of the patients, who, on a global scale, live with pain.
We are collecting data on patients’ mood, alertness, sleep, medications, pain level, and mobility to create a holistic profile of their experience. These factors work together to produce a single index that describes the fluctuating pain patient experience that we have called Pain Patient States. Though we are not developing a therapeutic agent, by building an understanding of patient states, we will be able to support physicians in making decisions with regard to patient care and even open the possibility to automatic therapy recommendations.
Rogers: As Rafael says, the partnership is all about better defining the patient experience. But how will we do that? Boston Scientific develops devices that can be used to create flexible and personalized therapies – the challenges they face is how to continuously choose the optimal therapeutic interventions – while we use the data produced to make personalized projections of future outcomes to those interventions. The amalgamation of these two approaches is how we intend to improve patient outcomes by recommending the best projected intervention.
What impact will this collaboration have on the future of drug development?
Carbunaru: Some drugs that are used to treat pain can result in addiction – with long-term consequences for patients and their families. By creating a fuller picture of the millions of patients who live with chronic pain, we can highlight real individualized patient needs and thus point the way to new therapies and treatment approaches. And so, though our work is focused on spinal cord stimulation devices, it has the potential to be applicable to the entire pain management spectrum, including pharma. It could even help in improving therapeutic outcomes for other chronic conditions.
How has AI helped the chronic disease market to date?
Rogers: I think it has a lot of potential to help in this area, but it certainly hasn’t had a broad impact yet. In the case of our current collaboration, AI has been a major driver in understanding the correlations in the data we have collected. It’s really helping to shape ideas not only about the patient population but also individuals. We can begin to answer questions that address what a healthy Pain Patient State looks like for a particular person and create a tailored healthcare experience for them.
A patient’s physiological and psychological state are affected by pain – and AI can give us a new platform to dissect the way these aspects of human health work together.
Carbunaru: AI is still a relatively new tool in healthcare. When you introduce something new – and very powerful – to the industry, there will, of course, be questions; some questions come from regulatory bodies and others from companies themselves.
Although this may present challenges for some companies, it will definitely be a positive step forward as it will help bring new solutions to patients waiting for improved quality of life.
Is there a limit to what AI can achieve?
Rogers: It is not a silver bullet! Although AI has the massive potential to improve healthcare, like anything else, it doesn’t solve all the problems experienced by patients, physicians, and other stakeholders within the sector. At best, AI is an augmenting set of tools to help enhance traditional healthcare approaches.
The FDA and other regulatory authorities have done a great job in thinking about the future use-case of AI. But, in my view, the healthcare community and the industry haven’t clearly defined how far they can go with the technology.
Carbunaru: We’re currently living through a period where innovation is happening at breakneck speed. So, the predictions we make today may fall short of what will happen five years – or even a decade – from now. For those of us working in R&D, we have to remain optimistic about creating a better world and helping patients achieve better health outcomes regardless of where AI may take us.