Accipiolx

AccipioIx is an AI-driven software designed to enhance radiology workflow by prioritizing adult non-contrast head CT scans suspicious for acute intracranial hemorrhage in acute care environments. It utilizes computer vision and deep learning algorithms to analyze images for signs of hemorrhage, providing case-level notifications to PACS or workstations for expedited clinical review.

AccipioIx is a software workflow tool developed by MaxQ-AI Ltd., designed to assist in prioritizing the clinical assessment of adult non-contrast head CT (NCCT) scans that exhibit features suggestive of acute intracranial hemorrhage (aICH) within an acute care setting. The system employs an artificial intelligence algorithm, leveraging computer vision and deep learning technologies, to analyze NCCT cases and identify suspected aICH findings. It comprises two main components: the AccipioIx Agent, which retrieves CT studies from a PACS and transfers them to the MaxQ-AI Engine, and the Engine itself, which processes the images. Following analysis, the Agent receives the case-level results from the Engine and returns them to the PACS or workstation, facilitating worklist prioritization or triage based on the identified potential for aICH. AccipioIx operates in parallel with standard care workflows and is not intended to replace clinical assessment, be used for standalone diagnostic decision-making, or to rule out hemorrhage. The AccipioIx algorithm was developed using a training set of CT cases from multiple institutions and CT manufacturers, and its operation involves algorithmic steps such as CT series identification, skull stripping, image registration, tissue segmentation, and a feature identification process that includes measures of hyperdensity compared to a predefined threshold to detect findings consistent with aICH. Performance testing, conducted through a retrospective study of 360 NCCT cases from over 30 US sites, compared AccipioIx to a ground truth established by at least two expert neuroradiologist readers. This evaluation demonstrated a system sensitivity of 92% (95% CI: 87.29-95.68%) and a specificity of 86% (95% CI: 80.18-90.81%) for identifying aICH, with an average per-case processing time of 4.1 minutes.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the AccipioIx algorithm's performance vary with different subtypes, locations, and volumes of intracranial hemorrhage?; What were the specific inclusion and exclusion criteria for the 360 cases used in the performance study, particularly concerning image quality and patient demographics?; Can the system be customized or re-trained for specific institutional needs or variations in CT scanner protocols?; How does AccipioIx handle mimics of intracranial hemorrhage or cases with subtle findings, and what is its false positive/negative profile in these scenarios?; What is the anticipated learning curve for radiological staff to effectively integrate AccipioIx's output into their existing PACS-based workflow and decision-making processes?