CIRRUS Lung Screening

Aim

This software is aimed to facilitate research collaborations between a network of clinical and technical partners that is needed to address the challenge of cost effective and efficient lung CT screening. A radiologist can quickly and accurately report chest CT scans and produce a standardized report with CIRRUS Lung Screening according to Lung-RADS guidelines. AI algorithms developed within DIAG are integrated into the workstation to allow for more accurate and faster interpretation of CT scans and extraction of a rich set of biomarkers. In addition to screen for lung cancer, the workstation is also aimed to screen for two other conditions for which heavy smokers are at risk: chronic obstructive pulmonary disease (COPD) and cardiovascular diseases.

Software

This software includes computer-aided detection and volumetric measurements for both solid and subsolid nodules. The workflow can be categorized into three steps - processing, reading and reporting. During the processing stage, input DICOM data is validated first. Then, lung, airway and vessel, lobe and segments are segmented. Various CAD systems are run to detect nodules, which is then followed by a dedicated lung CT registration algorithm to register a new scan with prior scans. All nodules from the prior scans are propagated automatically to the new scan. Finally, the prior findings and the CAD marks are merged. This software also provides a functionality to propagate a newly found nodule retroactively through the previous scans. Workstation CIRRUSLungS workstation

After a user has finished reading, a structured report is automatically generated. The report provides general patient information, gives a quick overview of the annotated findings and their characteristics and optionally provides follow-up recommendations according to the BTS/Fleischner/LungRADS guidelines. The suggested follow up and other case comments can be added to the report. Finally, the user signs off the case and proceeds with the next one. Workstation CIRRUSLungS report

Usage

There are two different versions of our software. The research version runs the processing pipeline on our in-house clusters and contains experimental features. The commercial version, Veolity, uses a client-server architecture, has communication with PACS systems integrated, can use PACS as a storage database, divides worklists over multiple clients, and is CE and FDA certified. Both versions are in active use at sites in North America, Europe, Asia and Australia.

People

Colin Jacobs

Colin Jacobs

Assistant Professor

Paul Konstantin Gerke

Paul Konstantin Gerke

Research Software Engineer

DIAG Research Software Engineering

Bram van Ginneken

Bram van Ginneken

Professor, Scientific Co-Director