According to Vignesh Shetty, senior vice president and general manager of Edison AI and Platforms for GE Healthcare (GEHC), the new rules and requirements add to the complexity of healthcare – something that intelligence Artificial intelligence can solve.
Healthcare IT News asked Shetty about the progress made on a 360-patient view since he last spoke to us about exciting advances in data-driven insights, and how algorithms achieve challenging data exchanges.
Q. Regarding the 360 patient view, what progress has been made towards improving health outcomes?
ONE. From MRI scanners used by doctors to detect tumors to portable X-ray devices in the ER or ICU used to image the lungs of COVID patients, doctors and patients are enjoying benefit from artificial intelligence integrated in medical devices.
The goal is to have AI tools ready when and where they are needed to contribute to faster diagnosis and, ideally, better patient outcomes.
One example of AI’s tangible impact in healthcare is GEHC’s Intensive Care Unit, an industry-first set of AI algorithms embedded in a mobile X-ray device. Built in partnership with the University of California San Francisco (UCSF) using GEHC’s Edison platform, AI algorithms help radiologists prioritize and flag a suspected case of pneumothorax , a common but sometimes life-threatening type of atelectasis, to help minimize the risk of complications. Period.
A recent review at Cleveland University Hospital found that while using the Intensive Care Unit, reporting times were reduced by 78% – from 3 hours 22 minutes to 44 minutes – for urgent tests. positive for pneumothorax.
Another innovation among GEHC’s AI-powered devices is AIRTM Recon DL technology, a pioneering, deep learning-based reconstruction algorithm for MRI exams that can reduce scan times by up to 50 percent. % without loss of image quality.
Chicago area-based Duly Health and Care, using this technology, has reduced wait times allowing them to see more patients sooner.
One of the doctors said the reduction in the amount of time they spent at the table – less than 10 minutes compared to up to 45 minutes – was a significant improvement for patients, especially those with back pain.
Q. Interoperability is slow because data standards are not strictly defined. How is this changing?
ONE. We believe that the adoption of the Cures Act, which enforces the FHIR representation of healthcare data in the form of a model, and interoperability supported through loose coupling by API representation, will help promote drive interoperability between traditional EHR systems and new-age innovation workloads that will run on traditional platforms.
These complement international and national initiatives with similar goals. For example, the US Office of the National Coordinator for Health Information Technology (ONC) has issued a number of interoperability standards and guidelines for the healthcare industry that will help reduce significantly reduce current challenges.
Q. Without a standard for genomic data, how can AI help solve this 1:1 out-of-negotiation data exchange for these rich data sets?
ONE. AI algorithms can be used to process and analyze genomic data and extract related information, such as gene variants or mutations, which can be organized and formatted in a standard manner .
AI accelerates the goal of precision medicine by building predictive models or other analytical tools that can help understand genomic data and extract insights.
In addition to the traditional standards that have prevailed in the world of genomics, the HL7 working groups have assembled a standardized representation of genomic data through five resources for FHIR version 4.
FHIR V5, which is under active development, represents even better genetics, so we expect this trend to accelerate.
Q. As information blocking rules go into effect, how can AI help increase data sharing and interoperability with data privacy needs?
ONE. With information blocking enforcement, we’ve found that the salience of insights becomes the primary source of value. Investments that seem unnecessary in a time of data scarcity become foresight in a context of abundance.
For example, when a lot of data flows through API automation, it will be impossible for humans to identify leaks on data pipelines.
In the case of structured data, machine learning algorithms can identify protected health information features to always comply with a privacy perspective. For unstructured data, we can use a combination of deep learning algorithms and natural language processing to structure them and solve interoperability.
Andrea Fox is the senior editor of Healthcare IT News.
Email: [email protected]
Healthcare IT News is a publication of HIMSS.
