Oxipit to Showcase AI X-ray Longitudinal Comparison at RSNA

“In radiology a lot of X-ray reporting specifically deals with comparison of how a condition progresses over time. Operating ‘under-the-hood’, ChestEye automatically compares the latest radiograph with a previous chest X-ray of the same patient and produces a concise report, which includes a description of the observed changes. This feature enables further gains in radiologist productivity and improves the performance of other ChestEye suite components as well," noted CEO of Oxipit Gediminas Peksys.

For example, ChestEye may detect lung nodules in both the previous and the current image. With Dynamics, ChestEye compares the size of the nodules and describes the progress of the condition in an automatically generated report. In the words of Mr Peksys, this allows for more actionable treatment decisions and brings ChestEye reporting closer towards reports written by a radiologist.

The feature also improves the performance of ChestEye Queue — a triage solution for patient management - as it takes into account the progress of patient condition.

“Dynamics further builds upon the vanguard capabilities of Oxipit ChestEye deep learning models, which enables us to develop new solutions for a more efficient radiology workflow,” noted Peksys.

At the RSNA conference Oxipit will also present ChestEye Quality and ChestEye Negative solutions.

ChestEye Quality is an automated clinical audit service for retrospective X-ray report analysis. Working in the background, the product analyses chest X-ray images as well as accompanying reports. ChestEye Quality then automatically highlights cases with the highest likelihood of important discrepancies.

“Current quality assurance procedures only select a small random sample of reports for secondary specialist analysis. ChestEye Quality enables to double-check all retrospective patient cases. Contrary to a periodic clinical audit, The ChestEye Quality solution operates in real-time, enabling to identify discrepancies that may affect patient treatment,” said Peksys.

ChestEye Negative produces preliminary reports for chest X-ray images with no abnormality. This enables radiologists to focus on the images with pathology. The product significantly assists the radiologist, especially in clinical environments with a low fraction of abnormal studies, such as large scale health screening projects and out-patient clinics.

For more information: www.oxipit.ai

Study: AI Found to Reduce Bias in Radiology Reports