By 2030, the healthcare artificial intelligence market is expected to be worth nearly $188 billion.
The Institute of Electrical and Electronics Engineers, the world’s largest non-profit engineering organization dedicated to developing technology for the benefit of humanity, is keeping an eye on AI – both the benefits and challenges of the booming technology in the future. healthcare field.
That’s why Healthcare IT news recently sat down with IEEE member Chenyang Lu. We asked him about how healthcare professionals use AI to help improve patient outcomes, the challenges of implementing AI in healthcare, and how to overcome those challenges. as to what he thinks the future of AI in healthcare will look like. He gave some very insightful answers.
Q. What is your perspective on how healthcare professionals can use AI to help improve patient outcomes?
ONE. AI will effectively become the co-pilot for our doctors and enable timely, accurate and effective treatment for each patient. AI models can make personalized predictions about a patient’s clinical outcomes, risk factors, and response to different treatments. Here are three examples of AI in healthcare that show great promise for improving patient outcomes.
First, screen for depression. According to WHO, more than 280 million people suffer from depression. Of those, more than 50% are undiagnosed and untreated. The problem of misdiagnosis stems from the time and expense involved in getting a diagnosis from a psychiatrist. A recent study shows that deep learning models can detect depression and anxiety disorders using data collected with wearable devices, opening up a new avenue to screen for depression reliably. discreetly.
This AI-based screening tool will enable clinicians to deliver selective prevention programs to individuals in a targeted and timely manner, addressing a key evidence gap in prevention Depression is defined by the United States Preventive Services Task Force.
Second, cancer care. According to a recent study, cancer patients are at high risk of clinical deterioration: 6.4% of cancer inpatients were transferred at least once to the ICU, and 2.7% of them died in the hospital. prevention. Machine learning models can generate early warnings of clinical deterioration in cancer inpatients by integrating heterogeneous data in electronic health records.
AI-generated early warnings, along with predictive associated risk factors, enable clinicians to pre-identify at-risk patients and provide early interventions to prevent the condition. condition worsens. Clinicians also face challenges in decisions about discharging patients from oncology. Long hospital stays reduce cancer patients’ ability to access hospitals. Machine learning models can be used to determine when hospitalized cancer patients are clinically stable for discharge, thereby improving the effectiveness of cancer care while ensuring patient safety. .
And third, perioperative care. Surgery poses many risks and costs for patients. Early identification of risk factors may be important for early intervention and improved outcomes. For example, pancreatectomy is the only cure for pancreatic cancer but often has a high rate of severe complications. Using data collected by a certain fitness wristband, a machine learning model can predict a patient’s risk of serious complications before surgery.
If a patient’s risk is high, they may be enrolled in prehabilitation programs to improve surgical readiness. Using EHR data, machine learning models have also been developed to identify perioperative risks and predict post-operative complications, aiming to improve patient safety and outcomes during the procedure. perioperative care.
Ask. What major challenges do you see when implementing AI in healthcare, and how can hospitals and health systems overcome these challenges?
ONE. Integrating AI models with EHR and clinical workflows is essential for implementing AI in healthcare. However, there are significant challenges in deploying AI models on current EHR platforms, as opposed to commercial cloud platforms that make building and deploying AI much easier.
Currently, we have many AI models in beta, but very few are deployed in the EHR. We are still in the early stages of AI in healthcare. Going forward, it is imperative to lower the barriers to deploying AI models in our infrastructure.
Furthermore, we needed to retool our workflows and protocols so that clinicians and AI could work together effectively. Experience in recent years shows that AI and clinicians have complementary capabilities. AI will be the co-pilot working with clinicians to collaboratively make the best decisions and treatments. Significant research is needed to develop effective human loop AI in clinical settings.
Q. What do you see as the future of AI in healthcare? What’s next and where do you predict it will go in the coming years?
ONE. We are seeing early adoption of synthetic AI to improve operational efficiency by automating clinical documentation and patient communication. Despite implementation challenges, we will see increased adoption of AI-based clinical decision support, driven by the enormous potential to improve patient outcomes and efficiency. health care results.
It is important that we generate evidence on the effectiveness and benefits of AI in healthcare in terms of patient outcomes and cost-effectiveness so that we can gradually build AI capabilities in the health system. his economy. At the same time, we need to ensure fairness, safety, security, privacy, and access to AI in healthcare through both policy and technology.
This is another area where significant research is needed to enable sustainable development of AI in healthcare.
Follow Bill’s HIT news on LinkedIn: Bill Siwicki
Email him: [email protected]
Healthcare IT News is a publication of HIMSS Media.
