AI tool helps remote monitoring clinics explore patient data
A new artificial intelligence (AI) tool allows cardiac care teams to ask straightforward questions to quickly search remote monitoring data and get answers in seconds. It works like an internet search engine and does not require reports or waiting.
Electrophysiology (EP) remote monitoring platform company Octagos launched a new AI capability layer on its system called Ask Atlas. It provides care teams with a faster, more intuitive way to understand what is happening across their patient population, workflows and clinical operations. It eliminated the need for traditional data tables, manual chart review and custom analytics requests.
"Remote monitoring has given clinicians more data than ever before, but access to data is not the same as understanding it," Shanti Bansal, MD, founder and CEO of Octagos and a practicing electrophysiologist, said in a statement. "Ask Atlas gives care teams a way to ask the questions that matter in real time. That kind of direct access to their own data changes how a care team can understand, manage, and act on its own information."
Atlas AI assists in the rapid interpretation of cardiac device transmissions by automatically pulling clinically relevant findings and structuring reports for efficient review. Human IBHRE-certified specialists provide clinical oversight and expertise. The system offers greater than 99% specificity, sensitivity and accuracy, which the company says significantly outperforms human experts alone.[1]
The company said Ask Atlas now allows natural language data access to device clinic operations, allowing clinicians and administrators to query data in the Octagos platform. This includes complex questions about the entire remote monitoring dataset, including device transmissions, EHR data, clinical context, follow-up activity, workflow status, and billing information.
Sample questions the system can answer include "which ICD patients have an EF <30% and are not on GDMT?," "which patients received ICD shocks in the past 30 days without a scheduled follow-up?," and "which remote monitoring events qualified for billing but have not been e-signed?"

