Artificial intelligence is no longer just an experimental technology in healthcare. It is already helping physicians examine medical images, identify diseases, prioritize urgent cases, perform ultrasound examinations, and make clinical decisions.
The U.S. Food and Drug Administration (FDA) maintains a list of medical devices that use artificial intelligence or machine learning and have been authorized for marketing in the United States.
Many of these technologies work behind the scenes. A patient may have an X-ray, CT scan, ultrasound, eye examination, mammogram, or colonoscopy without realizing that AI is helping healthcare professionals analyze the results.
Importantly, most AI medical devices are designed to support healthcare professionals rather than replace them.
Here are 10 notable FDA-authorized AI-enabled medical devices and technologies that demonstrate how AI is changing healthcare in the United States.
1. IDx-DR – AI for Diabetic Retinopathy Detection
Company: Digital Diagnostics
IDx-DR was an important milestone in medical artificial intelligence.
The FDA granted De Novo authorization for IDx-DR in April 2018. The device was designed to detect more-than-mild diabetic retinopathy in adults with diabetes.
Diabetic retinopathy is an eye condition caused by damage to blood vessels in the retina. It can lead to vision loss if it is not identified and managed appropriately.
The system analyzes retinal images and provides a result that helps determine whether the patient should be referred to an eye care professional.
What made IDx-DR particularly important was its ability to provide a screening decision without requiring a specialist to interpret the image first.
This demonstrated how AI could potentially expand access to screening in primary care and other settings.
For people with diabetes, easier screening could help identify eye problems earlier and encourage timely specialist care.
2. Viz LVO – Faster Stroke Detection and Care Coordination
Company: Viz.ai
Stroke treatment can be extremely time-sensitive.
A large vessel occlusion, or LVO, occurs when one of the major arteries supplying blood to the brain becomes blocked. Patients may require urgent treatment, and delays can affect outcomes.
Viz.ai developed AI-powered technology that analyzes medical imaging and helps identify suspected large vessel occlusions.
Viz LVO ContaCT is FDA-authorized radiological computer-assisted triage and notification software.
When the technology identifies a suspected LVO, it can help alert appropriate members of the clinical team.
The important word here is triage.
The technology is not meant to make every clinical decision. Instead, it can help healthcare teams identify potentially urgent cases faster and improve communication among professionals involved in stroke care.
AI-based triage systems could become increasingly valuable in emergency medicine, where every minute can matter.
3. GI Genius – AI-Assisted Colonoscopy
Company: Cosmo Artificial Intelligence / Medtronic
Colonoscopy is one of the most important tools for finding colorectal polyps and other abnormalities.
GI Genius brings artificial intelligence directly into the colonoscopy procedure.
The FDA granted De Novo authorization to GI Genius in April 2021 as a gastrointestinal lesion software detection system.
During a colonoscopy, the system analyzes the live video feed and identifies areas that may contain lesions such as polyps.
When it notices something suspicious, the technology provides a visual marker so the clinician can examine the area more carefully.
GI Genius does not replace the gastroenterologist.
Instead, it acts like an additional set of digital eyes during the examination.
This is an important example of how AI can work alongside healthcare professionals in real time rather than simply analyzing information after a procedure is finished.
4. Paige Prostate – AI in Digital Pathology
Company: Paige
Pathologists play an essential role in diagnosing cancer.
Traditionally, they examine tissue samples under a microscope. Digital pathology is changing this process by allowing tissue slides to be scanned and examined on computer screens.
AI can then help analyze those digital images.
Paige Prostate received FDA De Novo authorization in September 2021.
The software is designed to assist pathologists in identifying areas that may contain prostate cancer on digitized tissue slides.
This does not mean AI independently decides whether a patient has cancer.
A trained pathologist remains responsible for reviewing the tissue and making the diagnosis.
The value of AI is that it can help draw attention to potentially important areas within a very large digital image.
As digital pathology becomes more common in American hospitals and laboratories, AI could help pathologists manage growing workloads while maintaining careful review.
5. Caption Guidance – AI-Guided Cardiac Ultrasound
Company: GE HealthCare
Ultrasound is extremely useful, but producing high-quality ultrasound images requires training and experience.
Caption Guidance uses artificial intelligence to help healthcare professionals capture cardiac ultrasound images.
The system provides real-time guidance during image acquisition.
For example, it can help the user understand how to position and move the ultrasound probe to obtain useful views of the heart.
This is different from AI systems that simply analyze an image after it has been created.
Caption Guidance assists during the actual process of collecting the medical image.
The FDA classified the technology as an AI-guided image acquisition and optimization device.
This type of technology could make ultrasound easier to perform in more healthcare settings, particularly when highly experienced ultrasound professionals are not immediately available.
However, qualified medical professionals remain responsible for interpreting the images and making clinical decisions.
6. HeartFlow FFRct – AI and Advanced Coronary Artery Analysis
Company: HeartFlow
Heart disease remains one of the biggest health challenges in the United States.
When physicians suspect coronary artery disease, they need to determine whether narrowing of the coronary arteries is reducing blood flow to the heart.
HeartFlow FFRct provides a noninvasive means of obtaining additional information from coronary CT imaging.
The technology uses advanced computational methods to create information about blood flow and the functional significance of coronary artery disease.
The FDA granted De Novo authorization to HeartFlow FFRct in 2014, with later versions receiving additional FDA clearances.
The technology can help physicians better determine whether a coronary blockage is affecting blood flow, without immediately relying on an invasive procedure to obtain the same information.
This demonstrates an important direction for medical AI: turning existing medical images into additional clinical information.
Instead of simply looking at what an artery looks like, advanced software can help healthcare professionals understand how the disease may be affecting function.
7. OsteoDetect – AI for Wrist Fracture Detection
Company: Imagen Technologies
Fractures are among the most common reasons patients receive X-rays.
But identifying every fracture can sometimes be challenging, particularly when abnormalities are small or difficult to see.
OsteoDetect was developed to help. The FDA granted De Novo authorization to OsteoDetect in May 2018.
The software uses machine learning to analyze wrist X-rays in adults and identify signs of distal radius fractures.
It can highlight suspicious areas for healthcare professionals to review. This makes OsteoDetect an example of computer-assisted detection and diagnosis in radiology.
The radiologist or other qualified healthcare professional remains responsible for the final interpretation. AI simply provides additional information that may help during the review process. Similar technology is now being developed and used across many areas of medical imaging.
8. EchoGo Heart Failure – AI-Assisted Heart Failure Assessment
Company: Ultromics
Heart failure can be difficult to diagnose, particularly in certain forms where the heart’s pumping strength may appear relatively normal.
EchoGo Heart Failure uses machine learning to analyze echocardiographic data and provide additional information that may help clinicians identify heart failure with preserved ejection fraction(HFpEF).
The FDA cleared EchoGo Heart Failure as a clinical decision-support technology.
The system analyzes an ultrasound image of the heart and provides information to the clinician reviewing the patient’s cardiovascular condition.
The important point is that the AI output is adjunctive. Healthcare decisions should not be based solely on the software’s result.
Instead, clinicians combine the AI-generated information with symptoms, medical history, imaging, laboratory results, and other relevant clinical information.
This type of AI could become increasingly important as healthcare organizations look for ways to identify complex cardiovascular conditions earlier.
9. BrainScope – AI-Assisted Brain Injury Assessment
Company: BrainScope
Evaluating possible traumatic brain injuries can be challenging.
BrainScope combines EEG technology with advanced data analysis to help healthcare professionals assess patients with suspected brain injuries.
Different versions of BrainScope’s technology have received FDA clearances.
The system analyzes the brain’s electrical activity and provides useful insights to support healthcare professionals during patient assessment.
This technology does not replace CT imaging, neurological examinations, or physician judgment.
Instead, it can provide another source of information.
BrainScope shows how AI can quickly analyze complex brain signals and provide useful information to support clinical evaluation.
As healthcare moves toward more data-driven decision-making, similar technologies may emerge for neurological and other complex conditions.
10. Rapid ASPECTS – AI for Stroke Imaging
Company: RapidAI
Rapid ASPECTS is another example of artificial intelligence being applied to stroke care.
The technology analyzes brain CT images and provides an automated ASPECTS assessment.
ASPECTS stands for Alberta Stroke Program Early CT Score. It is a method healthcare professionals use to assess certain changes associated with ischemic stroke on CT imaging.
Automating parts of this process can help provide clinicians with additional information when evaluating stroke patients.
The FDA cleared Rapid ASPECTS as computer-assisted diagnostic software.
As with other medical AI technologies, its purpose is to support healthcare professionals rather than independently determine the patient’s complete treatment plan.
Stroke care is particularly well suited to AI because speed matters.
AI systems that rapidly analyze imaging and communicate useful information may help clinical teams make time-sensitive decisions more efficiently.
Why Is Radiology Leading Healthcare AI?
Many FDA-authorized AI medical devices are designed to analyze medical images, such as X-rays, CT scans, MRIs, and ultrasounds.
There is a practical reason for this. X-rays, CT scans, MRIs, mammograms, ultrasounds, and other imaging technologies produce large amounts of digital data.
AI systems are particularly good at identifying patterns within images.
For example, algorithms can be trained to look for:
- Possible tumors
- Bone fractures
- Stroke-related abnormalities
- Lung conditions
- Cardiovascular problems
- Eye disease
- Suspicious lesions
AI can also help prioritize medical images that may contain urgent findings.
Instead of simply replacing the radiologist, the technology can act as a second layer of analysis.
What Does FDA Authorization Actually Mean?
The phrase “FDA-approved AI” is often used loosely in news articles and marketing materials.
However, not every FDA-authorized medical device is technically “approved.”
Depending on the device and its risk level, medical technologies may reach the U.S. market through different regulatory pathways.
Two common pathways for AI medical devices are 510(k) clearance and De Novo authorization.
The FDA maintains an AI-Enabled Medical Device List to improve transparency around medical devices that incorporate AI.
According to the FDA, devices on the list have met applicable premarket requirements, including review of their safety and effectiveness for their intended use.
FDA authorization does not mean an AI system is perfect.
Healthcare professionals still need to use the technology as intended and understand its limitations.
Will AI Medical Devices Replace Doctors?
For most current medical AI applications, replacement is the wrong way to think about the technology.
The more realistic model is AI plus healthcare professionals.
- A radiologist may use AI to help identify suspicious findings.
- A gastroenterologist may receive AI assistance during a colonoscopy.
- A pathologist may use AI to highlight areas of a digital tissue slide.
- A cardiologist may receive additional information generated from medical imaging.
- A stroke team may receive an automated alert when imaging suggests an urgent condition.
In all of these situations, human expertise remains essential.
Doctors understand the patient’s symptoms, medical history, other conditions, medications, personal circumstances, and complete clinical picture.
What Healthcare Leaders Should Consider Before Adopting AI Devices
FDA authorization is an important starting point, but hospitals should still carefully evaluate AI technology before adopting it.
Healthcare leaders should ask several questions.
- Does the technology solve a real clinical problem?
- How well does it perform across different patient populations?
- Will it fit into existing clinical workflows?
- How will healthcare professionals review AI-generated results?
- How is patient information protected?
- What happens when the AI result is wrong?
- How will the organization measure whether the technology actually improves care?
Hospitals should also consider training.
Even a well-designed AI system may provide little value if healthcare professionals do not understand when or how to use it.
The Future of FDA-Authorized AI Medical Devices
AI medical technology is developing quickly.
Future devices may provide more sophisticated support for cancer detection, cardiovascular disease, neurological conditions, surgery, remote monitoring, pathology, and preventive healthcare.
Generative AI could eventually become part of some regulated medical technologies as well.
However, healthcare AI will require continued oversight. Algorithms can perform differently when patient populations, clinical settings, equipment, or data change.
The FDA has therefore been developing policies for how AI-enabled medical devices can be evaluated and updated throughout their life cycle.
For healthcare organizations, this means AI adoption should not be viewed as a one-time technology purchase.
Performance, security, clinical outcomes, and patient safety need to be monitored continuously.
Conclusion
FDA-authorized AI medical devices are already changing healthcare in the United States.
IDx-DR demonstrates how AI can expand diabetic eye screening. Viz LVO and Rapid ASPECTS show its potential in time-sensitive stroke care. GI Genius brings computer vision into colonoscopy, while Paige Prostate applies AI to cancer pathology.
These AI technologies are helping healthcare professionals improve ultrasound imaging, assess heart conditions, detect fractures, evaluate heart failure, and identify possible brain injuries.
These technologies represent only part of a much larger shift. The future of healthcare will likely involve doctors and artificial intelligence working together to improve patient care. It will increasingly involve doctors working with AI.
When used responsibly, AI medical devices can help healthcare professionals process information more quickly, identify key findings, reduce repetitive tasks, and make better use of the growing volume of medical data available to them.
The real measure of success, however, will not be how many AI hospitals adopt.
It will be whether these technologies help healthcare professionals deliver safer, faster, more accurate, and more accessible care to patients across America.
