This week, several technology companies, including Infinitus, Ambience Healthcare, and Anthropic, are launching innovative artificial intelligence tools aimed at improving trust and usability for patients, healthcare providers, and revenue cycle managers.

Hallucination-Free AI for Patient Engagement

Infinitus has introduced the first hallucination-free voice AI agents designed specifically for patient engagement. By restricting the responses of these AI agents, the company aims to enhance trust and compliance.

• These agents provide patients with 24/7 access to healthcare expertise.
• They help keep providers informed and streamline payer reimbursements.
• Potential benefits include improved healthcare literacy, medication adherence, and escalation of reported side effects to care teams.

According to Ankit Jain, CEO and co-founder of Infinitus, “The most important questions arise after hours when care teams are unavailable.” The AI agents also facilitate automated calls to simplify care coordination and clinical documentation, reducing delays in care.

Infinitus employs a proprietary method called Discrete Action Space, which limits the AI’s responses to a vetted set of phrases that comply with clinical and regulatory standards, minimizing the risk of hallucination often seen in traditional AI systems. These agents are HIPAA- and SOC 2-compliant, validating information in real-time based on specific data sources.

To further enhance accuracy, the agents monitor for anomalies and inconsistencies, generating alerts for human oversight. Jain emphasizes that “autonomous self-evaluation is essential for building trust and accountability in AI technology.” Infinitus claims its AI tools have facilitated over 100 million minutes of healthcare conversations for more than a million patients.

Ambience Healthcare’s Azure Integration

Ambience Healthcare has announced that its AI platform for health systems is now available on the Microsoft Azure Marketplace, streamlining deployment. The platform integrates with electronic health records and offers ambient listening across over 100 specialties.

• It analyzes conversations to identify precise ICD-10 and CPT codes.
• Generates patient summaries and referral letters to enhance revenue cycle management.

Jake Zborowski, General Manager of the Microsoft Azure Platform, stated, “Azure Marketplace and trusted partners like Ambience Healthcare enable customers to achieve more with less.”

Claude AI Introduces Voice Mode

Anthropic is preparing to launch a voice mode for Claude, its AI agent that competes with OpenAI’s ChatGPT. Users may soon be able to interact with this previously text-only tool via voice.

In January, Stanford Medicine reported that Claude was assisting physicians in informing patients about their test results. The in-house tool developed by Stanford Healthcare utilized Anthropic’s Claude 3.5 Sonnet LLM to create more understandable test result descriptions, thereby reducing the time physicians spend explaining results.

Dr. Christopher Sharp, Chief Medical Information Officer at Stanford Medicine, remarked, “I appreciate not having to start from scratch, as the draft is already in patient-friendly language.”

Feedback from a pilot test involving 10 primary care physicians led to further refinements, with a second cohort of 24 physicians testing the tool for an additional two months.

For more information, contact Andrea Fox, Senior Editor at Healthcare IT News.

Healthcare IT News is a HIMSS Media publication.