Pulmonary embolism happens when blood clots block one or more arteries in the lungs. PE can be dangerous if not found and treated fast. Some patients show clear signs like sudden chest pain or trouble breathing. But sometimes, pulmonary embolisms are found by accident during scans done for other reasons. So, it is very important to check these scans quickly.
Usually, radiologists look at each CT scan by hand. This takes time and depends on how busy they are and their experience. Before using AI, Atlantic Health System reported that the average wait time for a PE diagnosis was about 15 hours. This delay can cause more problems for patients and make their condition worse.
In 2020, Atlantic Health System started using an AI program to help radiologists find signs of pulmonary embolism in CT scans. The AI looks at images in two to three minutes. Doing this by hand takes longer. This quick check helps doctors focus on the scans that might show a clot right away.
Devon Klein, a radiologist at Atlantic Health, said, “Different doctors use AI in different ways, but I use it to pick the next case I should look at.” The AI sends alerts when it finds possible problems. This helps Klein and his team diagnose patients faster and start treatment sooner.
The wait time for PE diagnosis dropped from 15 hours to less than 90 minutes in outpatient tests. This is about ten times faster. In rare cases—only about 0.3% of nearly 12,000 scans—patients tested positive and got treatment quickly because of AI.
For hospital managers and IT staff, adding AI tools like those at Atlantic Health means more than just using new technology. It requires planning with many people involved, regular checking, and good data handling.
Atlantic Health created rules inside the organization to guide how AI should be used. They only use it when there is a clear medical need. This helps avoid using AI just because it is new or shiny. Suja Mathew, an executive at Atlantic Health, said, “Doctors understand possible errors and use AI with their own judgment.”
Good data is very important. Atlantic Health uses many types of data, like social and population data, to train the AI. This helps make sure AI works well for all kinds of patients and does not treat some groups unfairly.
From a management view, this means investing in safe and connected electronic health records (EHRs) that can collect and manage detailed patient information for AI to work well.
Outpatient clinics and imaging centers often deal with a large amount of data and follow-up work. AI does more than just look at images. When combined with automated workflows, AI helps smooth out processes that can cause delays.
This system cuts down the wait between diagnosis and treatment, especially for patients who might leave before results are ready.
At Atlantic Health, doctors can contact doctors who referred patients right after AI flags a positive scan. This quick communication helps give urgent care before the patient leaves. Klein said, “We now catch those patients, contact the physician, and coordinate emergency care before they leave, which is new.”
Voice technology also helps reduce paperwork for doctors. It turns conversations with patients into notes that go straight into electronic records. This lets doctors spend more time with patients and less time on computers. It also helps reduce doctor burnout and makes the whole process run smoother.
For medical managers and IT teams, using AI with workflow automation brings clear benefits. It speeds up test results, cuts down human mistakes in follow-ups, and helps use resources better. This leads to better care for patients.
While Atlantic Health’s AI example is specific, the use of AI for outpatient PE diagnosis shows larger trends in healthcare IT in the United States. Many places do not have enough radiologists, so AI tools become very useful for reading images. This is very important in emergencies and outpatient clinics where fast decisions can save lives.
Besides pulmonary embolism, Atlantic Health is testing AI for early cancer screenings, including breast, lung, pancreatic, and colon cancer. This shows they want to use AI where early detection can make a big difference.
Sunil Dadlani, an executive at Atlantic Health, said the goal is to improve patient safety and care by using technology in support of doctors, not to replace them. He said, “AI will not replace doctors but will add to human intelligence.” This shows AI works alongside healthcare workers.
Medical managers, owners, and IT staff considering AI solutions might think about these steps from Atlantic Health’s experience:
By following these steps, outpatient clinics and imaging centers in the U.S. can use AI to diagnose incidental pulmonary embolism faster. This will help healthcare providers give safer care, shorten wait times, and handle workloads better in busy settings.
AI algorithms can analyze CT scans swiftly for conditions like pulmonary embolism, triaging cases for urgent review by radiologists, leading to faster diagnoses and treatments.
AI screening has reduced diagnosis times by about one-third, enabling timely intervention for emergency cases in both inpatient and outpatient settings.
This initiative helps identify incidental pulmonary embolisms in outpatient settings, significantly reducing wait times for diagnosis to under 90 minutes from an average of 15 hours.
Clinicians are involved in every step of the process to ensure AI technologies align with their judgments and address real-world clinical problems.
The organization aims to use AI for clinical screening in oncology, focusing on early detection of various cancers, including breast, lung, pancreatic, and colon cancers.
Shiny toy syndrome refers to the risk of adopting AI technologies merely because they are available, rather than based on their relevance and effectiveness for specific clinical problems.
AI tools’ effectiveness depends on the diversity and accuracy of their training datasets. Poor representation can lead to unreliable outcomes in different populations.
The organization integrates data from multiple sources, including social vulnerability and population statistics, to create a comprehensive view of their patient demographic.
AI enhances efficiency by automating routine tasks like note-taking, which allows physicians to focus on patient care, potentially reducing burnout and improving job satisfaction.
The mission is to maximize patient safety and quality of care while leveraging technology to support clinical teams and improve overall outcomes in healthcare delivery.
Simbo is using AI agents to automate all phone related workflows. Connect now to know more.
Schedule AI Agents Demo Now© Since 2019, Simbo AI.
