I went for my annual checkup with my PCP this past week where she recommended “all the preventative things.” As the list was wrapped up, she asked, “Any questions?” I quickly responded, “Yes, do the providers you are referring me to use Artificial Intelligence (AI) assisted imaging or tools?” She was noticeably caught off guard by my question and shared that she did not know. A rather informative conversation ensued as she asked many questions about AI and how I’ve seen AI used within healthcare. The conversation concluded as she asked me if I would be willing to share with her what I found out about her referrals.
My expectation is to receive the best level of care available, which includes providers who are assisted by AI tools. I am looking for reassurance that my providers will be practicing at a standard that has been proven to capture unexpected findings, standardize assessments, expedite provider response and help clinical teams prioritize and triage findings. AI in Healthcare (AIH) is no longer the novelty that a few cutting-edge organizations leverage – it’s an expectation.
AI-Assisted Care Pathways – The Next Age
We are rounding the corner of the next age in AIH where novelty is abandoned and the expectation is set. Clinical care teams must act now and consider how AI should be woven into their care pathways – from prevention to acute care incidents, during rehabilitation and chronic disease management and through end-of-life care. This next age will see standards and best practices where AIH is intentionally threaded through clinical care pathways to standardize care and improve patient outcomes.
Setting an Expectation
The expectation that AI is woven into acute stroke care pathways is a standard that the AHA specifically addressed last year when they released their scientific statement outlining the ideal foundational requirements for stroke programs – an effort to address heterogeneity found among equally certified programs. These foundational requirements aim to reduce variability, strengthen stroke care services and improve patient outcomes. These requirements also address public expectations to receive the highest level of care possible at a stroke-certified site. Of note, this statement identified that AI-assisted neuroimaging for CT and CTA was a foundational aspect for all levels of certified stroke centers.
Let’s review the AHA scientific statement highlights related to AI neuroimaging – which includes imaging of the arch, neck and head – to facilitate expedited team processes and identification of stroke pathology:
- Artificial intelligence-supported neuroimaging available 24/7 at ASRH, PSC, TSC, and CSC’s
- Noncontrast CT available 24/7 at ASRH, PSC, TSC, and CSC’s
- CTA available 24/7 at PSC, TSC, and CSC’s + interpreted within 20 min of completion
- CTA imaging completed immediately after CT
- CTA interpretation times similar to CT
- CT Perfusion available 24/7 at TSC and CSC’s
- MRI available 24/7 at TSC and CSC’s
- Transcranial doppler services available 24/7 at CSC’s
Why would the AHA set an expectation about AI neuroimaging for acute stroke patient assessment and decision-making? To answer this, we need to examine the available evidence – namely around diagnostics in acute stroke care. Last year, the Neurology team at Johns Hopkins School of Medicine addressed the burden of diagnostic errors on patient outcomes. Of note, this team reported a 17.5% diagnostic error rate and a 9.8% serious misdiagnosis-related harm rate in stroke patients! This is just one example of such data that has driven organizations like the AHA to set practice expectations that include AI technology.
Adopting New Expectations – One Year Later
One year later, the AHA released another statement addressing AI in stroke care. In this article, the AHA acknowledges that AI has the “potential” to impact stroke care, but raises concerns that this may not be right away and encourages health organizations and leaders to take action. These concerns were directly related to the ability of care teams to develop best practices to advance the use of AI in patient care. This statement addresses the need for additional research and standardization of AI best practices. Specific to the discussion around AI neuroimaging is the standardization of platforms and how to implement AIH into care pathways.
Aidoc has developed the leading AI healthcare platform that directly addresses the call for standardization, with a fully automated 24/7 monitoring and analysis system that integrates within native workflow and IT infrastructures while coordinating care team response. This technology is supported by evidence-based research showing precise approaches to patient care. Furthermore, Aidoc’s AI Operating System (aiOS™) end-to-end platform and customer delivery teams lead the charge in helping care teams overcome the challenges of AI adoption with a single, unified AI platform coupled with well-developed customer delivery programs focused on AI technology implementation, adoption, and long-term success – directly addressing the urgent needs outlined in the latest AHA statement regarding AI in cardiovascular care.
