top of page
  • bryanstlouis

5 FDA Approved Medical Devices That You Didn't Know Used Artificial Intelligence

By Quality Means Business

Healthcare organizations are using artificial intelligence (AI), defined by the US Food and Drug Administration as “the science and engineering of constructing intelligent machines,” for an increasing range of clinical, administrative, and research applications. Software incorporating artificial intelligence (AI), notably the subtype of AI known as machine learning (ML), has become an increasingly significant component of an expanding number of medical devices as technology progresses in many facets of health care. Machine learning’s to produce fresh and useful insights from enormous amounts of data created each day, during the delivery of healthcare is one of its most enticing advantages. The most important question that many accentuate on is “What is AI, and how is it used in t health care?”


Artificial Intelligence allows us to measure a computer’s capacity to carry out operations that resemble human behavior, such as learning and problem-solving. It can be utilized for many different things, including automating processes, recognizing patterns in data, and combining data from many sources. For example, radiology and ophthalmology are two professions that rely on image analysis, as well as wearable sensor products that collect/analyze data to diagnose diseases and anticipate the beginning of other health disorders are already using AI technologies.

AI programs can also predict patient outcomes based on data from electronic health records by determining which patients are more likely to develop disease or prognosticate who should be monitored more closely within a family. Based on factors such as vital signs and test results from electronic health records, one such model identifies patients in the emergency room who may be at increased risk of developing sepsis. When compared to other risk-assessment tools, AI’s ability to better predict which discharged patients are likely to be readmitted following their release seems to be far superior. Other healthcare systems are likely to follow suit in developing their own models as technology becomes more widely available. Well-established federal regulations implement efforts to facilitate data exchange between electronic health record systems and mobile applications, a process known as interoperability.

As a resource for information about these devices and the FDA’s involvement in this field, the FDA has made a list of intelligence medical devices advertised in the United States. Here are 5 FDA Approved Medical Devices that utilize AI.

1. EarliPoint System


The device classification name is Pediatric autism spectrum disorder diagnosis aid. A pediatric autism spectrum disorder diagnosis aid is a prescription device that is intended for use as an aid in the diagnosing of autism spectrum disorder in pediatric patients. The device collects data based on the clinical presentation of a patient. Then, an analysis algorithm is applied to the collected data. The device may be stand-alone or implemented as a software application on a smartphone or other general-purpose computing platform (FDA, US Food & Drug Administration).


The FDA approved the marketing of a device to aid in the diagnosis of autism spectrum disorder today (ASD). The Pediatric autism spectrum disorder diagnosis aid is a machine learning-based software designed to assist healthcare providers in diagnosing ASD in children aged 18 months to 5 years old who exhibit potential symptoms of the disorder. Alzheimer’s disease, Parkinson’s disease, major depression, epilepsy, spinal cord injury, and traumatic brain injury are just a few of the neurological disorders and conditions that neurological devices can help diagnose, prevent, and treat. Neurological devices can help restore hearing and sight and increase function in people who have limb loss or congenital limb differences. Neurodiagnostic, neuro-interventional, and neurostimulation devices are examples of neurological devices.

2. Atrial Fibrillation History Feature


The device classification name is Photoplethysmograph Analysis Software for Over-The-Counter Use. A photoplethysmograph analysis software device for over-the-counter use analyzes photoplethysmograph data and provides information for identifying irregular heart rhythms. This device is not intended to provide a diagnosis (FDA, US Food & Drug Administration).


Heart failure is a chronic condition that usually worsens over time and can lead to death from end-stage heart failure. End-stage heart failure patients experience debilitating shortness of breath, exhaustion, painful swelling in the legs and feet, and fluid buildup in the lungs and abdomen.


Apple received 510(k) clearance for a new feature for its smartwatch that displays an estimate of how frequently their heart rhythm exhibits signs of atrial fibrillation (AFib). In 2018, the Food and Drug Administration approved Apple Watch for use in detecting irregular heart rhythms. The new feature is intended to assist AFib patients in tracking the effect of lifestyle factors on the frequency of their irregular heart rhythm.


3. BrainInsight

The device classification name is Automated Radiological Image Processing Software. To provide automated radiological image processing and analysis tools. Software implementing artificial intelligence, including nonadaptive machine learning algorithms trained with clinical and/or artificial data. In these devices, the algorithm training data typically impacts device performance. Adaptive AI algorithms are not within the scope of this product code (FDA, US Food & Drug Administration).


Quantitative imaging is defined by the Radiological Society of North America as the extraction of quantifiable features from medical images for the assessment of normal or the severity, degree of change, or status of a disease, injury, or chronic condition relative to normal. Obtaining quantitative imaging biomarkers, such as midline shift, is difficult in neurocritical care due to limited access to neuroimaging.


BrainInsight has the potential to significantly improve workflow in the neurocritical care setting. BrainInsight provides automatic midline shift measurements equivalent to manual measurements from expert neuroradiologists, according to study results from intensive care units. This enabling technology has the potential to reduce the demand for acute stroke follow-up imaging on neuroradiological specialists. Nonspecialists, particularly those reading images in emergencies, can benefit from BrainInsight’s automated midline shift measurement.


4. Viz SDH

The device classification name is Radiological Computer-Assisted Triage and Notification Software. Radiological computer-assisted triage and notification software is an image-processing device intended to aid in the prioritization and triage of time-sensitive patient detection and diagnosis based on the analysis of medical images acquired from radiological signal acquisition systems. The device identifies or prioritizes time-sensitive imaging for review by prespecified clinical users based on software-based image analysis but does not provide information from the image analysis other than triage and notification (FDA, US Food & Drug Administration).


SAN FRANCISCO - 27 JULY 2022 - Viz.ai, the leading AI-powered disease detection and intelligent care coordination platform announced today that Viz Subdural has received FDA 510(k) clearance (SDH). The Viz SDH algorithm employs artificial intelligence to detect subdural hemorrhage, allowing physicians to effectively triage patients and provide optimal care.


Subdural Hematoma (SDH) is expected to become the most common neurosurgical diagnosis by 20301, with multiple global clinical trials looking into promising new treatments. Acute and chronic subdural hemorrhages, on the other hand, necessitate different types of intervention via different clinical pathways, with some requiring immediate attention from the care team. Viz SDH is the only SDH-specific AI-powered detection and care coordination platform that can detect acute and chronic subdural bleeds and alert the care team to mobilize if immediate intervention is required.

5. Columbo

The device classification name is Automated Radiological Image Processing Software. Columbo is software used to analyze lumbar spine pictures from MRIs. Columbo is designed to speed up the reading process, offer information on potential diseases and abnormality detection, and shorten the recording process for MRI readings of the lumbar spine. According to the Spine Reporting Survey we conducted, radiologists believe that a thorough L-Spine MRI examination in cases without any complications should take an average of 17.2 minutes, and a thorough examination should take an average of 30 minutes in cases that are more complicated. However, they only have an average of 13.6 minutes to read an MRI in a real clinical setting, at the cost of up to 30% in errors and omissions, to provide automated radiological image processing and analysis tools. Software implementing artificial intelligence, including nonadaptive machine learning algorithms trained with clinical and/or artificial data. In these devices, the algorithm training data typically impacts device performance. Adaptive AI algorithms are not within the scope of this product code (FDA, US Food & Drug Administration).


The main advantages of Columbo are its low cost, ease of use, and optimal cost-result ratio. Columbo has the potential to reduce the economic burden imposed on the business because the low cost of the service will make it available to a larger portion of the population, assisting with the implementation of preventative measures that will reduce the number of chronically disabled employees.


The key innovation of Columbo is its unique machine learning algorithm, which is capable of reading MRI images and is based on fully convolutional neural networks combined with medical domain knowledge, dropout regularization, and packet normalization to assist qualified clinicians in providing a complete diagnosis. To improve resolution, a U-Net-like architecture is used.


Endnotes


  1. “BrainInsight™ Automated AI Tools: SWOOP® Portable MRI™.” Automated AI Tools: Swoop® Portable MRI™, https://hyperfine.io/swoop/braininsight-overview.

  2. Center for Devices and Radiological Health. “Artificial Intelligence and Machine Learning (AI/Ml)-Enabled Medical D.” U.S. Food and Drug Administration, FDA, 10 Nov. 2022, https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-aiml-enabled-medical-devices#resources.

  3. Center for Devices and Radiological Health. “Medtronic HeartWare Ventricular Assist Device (HVAD) System.” U.S. Food and Drug Administration, FDA, 20 Jan. 2023, https://www.fda.gov/medical-devices/cardiovascular-devices/medtronic-heartware-ventricular-assist-device-hvad-system.

  4. Center for Devices and Radiological Health. “Neurological Devices.” U.S. Food and Drug Administration, FDA, 8 Apr. 2021, https://www.fda.gov/medical-devices/products-and-medical-procedures/neurological-devices.

  5. “Columbo Is Software for Analysis of Lumbar Spine Images Obtained with MRI.” CoLumbo, https://columbo.me/.

  6. “Viz.ai Receives FDA 510(k) Clearance for Viz Subdural (SDH).” Viz.ai Receives FDA 510(k) Clearance for Viz SUBDURAL (SDH), https://www.viz.ai/press-release/viz-ai-receives-fda-510-k-clearance-for-viz-subdural-sdh.


52 views0 comments
bottom of page