AI Could Bring These Big Breakthroughs to Healthcare In the Next Year

May 2, 2019 - 7 minutes read

Recently, Boston-based Partners HealthCare ranked AI-enhanced technologies by which will have the greatest impact on medicine and healthcare in the next year. The report, titled the 2019 “Disruptive Dozen”, was compiled after more than 60 interviews took place.

Erica Shenoy, M.D., is associate chief of Massachusetts General Hospital’s Infection Control Unit. She says the 12 winners of the study have significant potential to affect clinics and patients. They also have products or ideas that will be selling by 2020.

Shenoy elaborated, “They have to be technologies that are pretty close to making it to market. The idea here is that these are high probability of deployment within the next couple of years.”

Major Players to Keep an Eye On

We’ve grouped the winners into several categories:


Ranked #3, clinical imaging will help make diagnosing easier and more accurate for physicians. As computer vision, machine learning, and AI become more attuned to medical images, clinical operations will change for the better. AI can help triage images for the physician’s attention, and a deep learning algorithm can subsequently mark what it thinks looks abnormal or concerning.

Researchers in Washington are using deep learning to diagnose malaria. The group’s software detects and counts malaria parasites, and it’s pretty good at it, too: it has a 90% accuracy and specificity rate. That’s as good as a physician. This tool, ranked #4, comes with an affordable, automated digital microscope for medical labs. For tests in Peru and Thailand, the results have been promising and significant. The software is in commercial development.


Medical imaging is a field that will lead major change in many medical specialties; this area of healthcare was ranked #1 for its potential by the Partners HealthCare group. The obvious example is radiology, but OB/GYNs are using medical imaging for mammograms now, too. AI further transforms medical imaging by calculating and assigning risk scores for breast cancer. Researchers in Massachusetts are using machine learning to improve the breast cancer screening process. Their tool is already being used in a large hospital, and it’s helping physicians calculate breast density using mammogram data.

Ranked #6, eye health practitioners stand to see improvement more quickly than many other medical specialties. The Food and Drug Administration recently approved an AI system that detects diabetic retinopathy. Researchers in the U.K. are creating AI tools that aren’t so specialized; they want to look for more than 50 common eye conditions using a single image!

Mental Health

Mental and brain health is very important to all patients. Fortunately, monitoring and analysis for these are getting more automated. Boston-based researchers added details into 30 TB of EEG data before feeding this data to a deep learning algorithm. The software can detect seizures automatically, no matter what the underlying condition or cause may be. This tool will be used in a hospital soon, and it’s ranked #5 on the list.

When patients experience a stroke, they get confused and cannot articulate what’s happening to them. Massachusetts MedTech developers are using AI to detect for bleeding in the brain. The software was trained on 2,000 CT images of heads. Most interestingly, it can localize the source of bleeding and calculate how much brain tissue has been affected. The team will continue expanding their tool, which was ranked at #9.

Another research group in Massachusetts is developing an app for patients that have drug, alcohol, or opioid addiction. This is ranked at #12. The app serves as virtual group therapy for the patient, and it’s been shown to be highly effective in helping patients recover and prevent relapse.

Societal Health

Suicide is as much a mental problem as it is a societal one. Researchers across the U.S. and at various social media companies teamed up to apply natural language processing and machine learning to suicide risk prediction. This tool, ranked #2 for its potential, alerts caregivers and physicians when a child in their care is thinking about committing suicide using predictive analytics.

Ranked #10, an AI tool that alerts physicians when it thinks a patient’s injuries could be caused by violence from their intimate partner. The tool uses an integrated approach to holistically analyze clinical records and radiological images to calculate the possibility of domestic abuse.

Operations and Administration

Sharing data between healthcare providers and specialists is the only way to provide a seamless patient experience. Wouldn’t it be nice to periodically download your entire medical history from one website or app? Well, the Fast Healthcare Interoperability Resources (FHIR), ranked #7, are aiming to become the standard for sharing medical and healthcare information. FHIR helps create an integrated experience for patients.

Healthcare administration is notoriously complex and difficult to navigate. Medical coding and billing haven’t caught up with technology, but a Boston startup is aiming to solve this issue with machine learning. The algorithm reads doctors’ notes from health records to map the appropriate diagnostic and procedure codes. Ranked at #8, this tool is going to save healthcare administrators a lot of time and effort.

Ranked #11, the power of adding voice technology to healthcare is poised to become stronger in the next five years. AI-enabled voice technology uses speech recognition and natural language processing. Using voice technology, doctors can deliver better care, save time, and spend more quality time with the patient. Voice technology can also automate data entry.

Endless Potential

As data becomes more foundational to healthcare, AI will become an integral tool for figuring out optimizations and improvements in every aspect of the field.

Which of these 12 breakthroughs is your favorite? Which do you think has the most potential? Let us know in the comments!

Tags: , , , , , , , , , , , , , , , , , , , , ,