The idea of receiving medical treatment from a machine instead of a doctor is hardly a novel idea, yet we’re only just starting to see exactly what’s possible.
Surgical robots have been in use for over 30 years, and lab workers have long relied on the help of simple automated assistants. However, these systems have historically been limited to performing small sets of predetermined tasks.
This is set to change thanks to rapid advancements in Artificial Intelligence (AI) technologies. Researchers around the world are exploring how AI can be used to diagnose illnesses, discover treatments, and even care for patients in ways beyond the means of medical professionals. If successful, those innovations have the potential to save lives and lower healthcare costs in the United States by $150 billion annually within the next six years.
Uncovering the unseen
When doctors have difficulty diagnosing health issues, the problem isn’t necessarily a lack of information. Sometimes they have too much information, where comprehending all the data to find patterns and clues becomes a near-insurmountable task.
That’s why researchers at the Houston Methodist Research Institute have turned to AI to help make sense of their data, having developed software that’s capable of reviewing mammograms 30 times faster than humans, enabling speedier treatment for those who desperately need it. The system’s 99 percent accuracy rate is also a crucial improvement, considering half of the United States’ annual 12.1 million mammogram readings yield false results, leading to 20 percent of biopsies being unnecessary.
AI succeeds at these sorts of tasks because it can identify patterns in large volumes of data and quickly perform trial and error tests, which is why the approach is being replicated in other areas of medicine. For example, a team at the Fralin Biomedical Research Institute at Virginia Tech is utilizing machine learning in its effort to develop a data-driven approach to diagnosing mental illness based on patients’ MRI scans. Meanwhile, others are using AI to aid in the drug discovery process because it can monitor individual changes in thousands of cells simultaneously, allowing scientists to more easily see the impact of their treatments.
Transforming the experience for patients
AI isn’t just offering help behind the scenes: Mabu, a robotic “personal healthcare companion,” is offering patients round-the-clock assistance while making them feel comfortable through its humanoid appearance and penchant for conversation.
It may sound counterintuitive to have a robotic carer act like a person – after all, why not have a human do it? The problem is that it’s a time-consuming job, meaning patients who need regular attention may only receive occasional visits. Because Mabu is a live-in robot, it can keep track of patients’ medication regimens, fill out the written logs that often get ignored, and communicate data back to your doctors.
Another project, Text2Move, is using AI to check in with patients without needing to actually place robots within households. Instead, the service uses a combination of machine learning and location data to deliver contextual advice via text message with the aim of improving lifestyles. For instance, a patient might receive a suggestion to go for a walk if they’ve been in their home all day, or they may be reminded to apply sunscreen if they’re outdoors and the UV index is high.
While many of these initiatives are still early in their development, it’s clear that AI is redefining what’s possible in the health space. And that means better tools for medical professionals — and ultimately better treatment for patients.