AI roundup: robot companions, automated admin tasks, and brain scans

Artificial intelligence (AI) may not replace physicians (yet), but the technology is making its way into almost every level of healthcare.

Already, AI has shown the potential to help physicians determine the appropriate imaging tests for patients and predict a hospitalized patient’s length of stay. In the face of ongoing clinician shortages and burnout, information technology consulting company Accenture estimates that about 70% of healthcare professionals’ tasks could be reinvented by automation or technology augmentation such as AI.

We’ve rounded up the latest AI healthcare developments you may have missed.

AI companion: Some older New Yorkers are turning to AI companion robots amid the country’s loneliness and isolation epidemic. The New York State Office for the Aging (NYSOFA) partnered with healthcare software startup Intuition Robotics to bring the AI-powered care companion ElliQ to 800+ older New Yorkers, and participants had a 95% reduction in loneliness during the pilot study over the past year, NYSOFA announced on Aug. 1.

ElliQ works by initiating conversations with users, suggesting activities like recording a digital memoir, and engaging in cognitive games. The majority of these interactions are designed to increase the mental, social, and physical well-being of older adults, according to the announcement. Each ElliQ companion subscription costs $39.99 per month after a one-time $249.99 enrollment fee.

“The pandemic encouraged us to open our eyes and think of innovative ways to deliver care to our clients. Partnering with an AI platform like ElliQ fits our strategy perfectly and I’m looking forward to getting ElliQ devices into the homes of many more older New Yorkers,” Becky Preve, executive director of the Association on Aging in New York, said in a statement.

Clinical documentation: Amazon Web Services (AWS) rolled out AWS HealthScribe, which uses generative AI to help providers with clinical documentation for electronic health records, the company announced July 26.

“Documentation is a particularly time-consuming effort for healthcare professionals, which is why we are excited to leverage the power of generative AI in AWS HealthScribe and reduce that burden,” Bratin Saha, VP of AWS machine learning and artificial intelligence services, said in a statement.

The World Health Organization raised concerns earlier this year about patient privacy risks when AI is incorporated in healthcare. However, AWS HealthScribe doesn’t retain any customer data after processing a transcript of the patient-provider conversation, and according to the company, that the information gleaned will not be used to train AWS HealthScribe.

AWS HealthScribe is available for clinicians working in general medicine and orthopedics without needing to manage the tech or train their own large language models, per Amazon.

Administrative tasks: Information technology company Cognizant is using Google Cloud’s generative AI capabilities to help reduce administrative workloads, it announced Wednesday. For example, the generative AI can automate the prior authorization process and file appeals when insurance denies a claim.

“Together, Cognizant and Google Cloud are moving beyond highly publicized content generation applications of generative AI to build enterprise healthcare solutions that drive significant cost optimization, business efficiencies, and better experience,” Cognizant CEO Ravi Kumar S said in a statement.

Administrative expenses account for about 15%–30% of healthcare spending, according to a 2022 Health Affairs brief.

AI meets brain scans: NYU Langone Health’s neurosurgery department is using AI to help clinicians understand brain cancer, according to the health system’s recent quarterly report.

The project, called NYUMets, brings the power of AI tools to the world’s largest public database of metastatic brain tumor images, allowing clinicians to map the disease’s progression and treatment responses.

“Our goal is to encourage computational scientists to start thinking about cancer the way clinicians do—not as an independent entity, but as a disease someone lives with over time,” NYU neurosurgeon Eric Oermann said in a statement.