Welcome to the third and final part of our series on how looking at data in a different way can unlock more value from healthcare AI!
In our last entry, we examined the detrimental effects that bias can have on AI-driven care. We also delved into one of the main bottlenecks to medical AI innovation: a lack of access to large-scale data. In case you missed it, you can read the second part here.
For this post, we’ll take a look at why improving healthcare AI is dependent on changing our perspective on data collection. We’ll also dive into the most important metrics to use for measuring success in medical AI endeavors. There’s a lot to cover, so let’s jump right in!
Changing Our Preconceptions of Data Collection
Bias in AI systems is already a big problem. When you introduce this issue into healthcare, it can lead to some catastrophic and possibly fatal effects. As we discussed in part 2 of this series, giving AI developers access to more diverse data can go a long way towards preventing bias. But to make this happen, some wholesale changes to the way healthcare providers and patients view data sharing are required.
Calum MacRae, MD, PhD, Chief of Cardiovascular Medicine at Brigham & Women’s Hospital, compares this to the de facto data sharing that is commonplace across other industries: “When you use Uber or Google, you are contributing data to the model and helping those companies refine their algorithms and improve the ability for you and all other consumers to use the model going forward.”
Of course, this isn’t to say that healthcare should approach data security with the same cavalier attitudes as these tech titans. Rather, MacRae indicates that consumers tend to frown upon healthcare organizations using data while other companies never face the same scrutiny.
“…the reality is that there is this disproportionate sense of indignation or anger when we think about a healthcare organization using our data for research,” he explains. “And we haven’t really done anything to explain why that should be different.” Sharing data more freely can certainly lead to improvements in healthcare for all patients. But it’s difficult for them to see it this way. MacRae says, “There is a sense that people hold onto their health data because they don’t get anything in return.”
MacRae was far from the only presenter at the 2019 World Medical Innovation Forum to echo these sentiments. Alistair Erskine, MD, MBA, Chief Digital Health Officer at Partners HealthCare, stresses that part of the problem is due to there being no universal data model for healthcare organizations to follow. If standards with rock-solid security were put in place, medical developers would be able to accelerate progress without sacrificing patient privacy.
Besides this, MacRae emphasizes that the attitudes of healthcare industry stakeholders are also to blame. “I believe that one of the core professional mandates for physicians is to lead us there in a way that is responsible and truly beneficial to the patient. If we don’t do that, we are compromising our long-term vision for how healthcare should evolve.”
Defining Success for AI in Healthcare
For many other industries employing AI, measuring success is straightforward. Whether it’s a tech developer in San Francisco or a bank in New York City, financial return on investment is usually the key indicator. But in healthcare, this should not be the case.
Instead, the most important metric to use for measuring healthcare AI is whether or not it’s helping to guide providers and patients in the right direction. This holds true whether it’s in the context of identifying risks and informing patients about them or empowering providers to make the right decisions when it comes to treatment plans.
Erskine equates this with being able to connect the dots in a data repository of patients with similar conditions. “As a clinician, sometimes I think that I make decisions based on my heart or my gut: things that don’t have evidence that I can point to,” he says.
“But the truth is that the evidence does exist, and I’m not making these decisions based on nothing. It’s just that the evidence is buried somewhere so deep in the data that I can’t find it within the time I have to treat that person. Artificial intelligence can do that for me.”
Constance Lehman, MD, PhD, Radiology Professor at Harvard Medical School and Chief of Massachusetts General Hospital’s Breast Imaging division, thinks that AI’s greatest potential lies in education and awareness. “No matter what clinics I work in, we see women who are so surprised when we tell them they have breast cancer,” she explains. “This is the year we are going to change that. This is the year we can start to use artificial intelligence to inform women of their risk with a level of accuracy that we have never had before.”
A Shift in Perspective Can Save Lives
With AI, healthcare providers and patients have the power to change outcomes for the better. Today’s technology and data already permit this possibility. We just have to adapt the right perspective to make it a reality.
So far, it’s been difficult for the medical industry to embrace AI with open arms. Reframing the role this technology plays could improve results drastically. It could also substantially reduce the chances of bias creeping into AI systems. But to do so, we’ll not only need to reassess how data is collected, but how we give innovators and practitioners access to it.
At first glance, these endeavors can seem arduous. But if we succeed, many more people would benefit in the form of a brighter, healthier future.Tags: AI, AI App Developer, AI app developer San Francisco, AI in MedTech, app developers san francisco, artificial intelligence app development, eHealth app developer San Francisco, eHealth app development San Francisco, health app developers, healthcare, MedTech and AI, MedTech app, MedTech app developer, MedTech app developer San Francisco, MedTech app developers San Francisco, mHealth app developer San Francisco, mobile app developer San Francisco