LesionSim - Virtual Clinical Imaging Trial Software

Use LesionSim to define image quality requirements for reliable diagnosis
Conduct Clinical Imaging Trials Quickly & Rigorously

Let's Talk
Determining Imaging Requirements through Simulation
The purpose of a medical x-ray is to visualise structures & pathologies for diagnosis, yet radiology professionals have not had access to rigorous methods for determining the adequacy of their clinical images for a particular diagnostic task.
Our vision lacks specificity when assessing image quality attributes such as noise, and its likely impact on our ability to see any subtle features needed for diagnosis. To address this, we have developed LesionSim, an application and web service for simulating realistic pathologies on clinical images and conducting web-based detection accuracy surveys. The simulated pathologies are typically designed to represent the “edge case” – that is the smallest lowest contrast version of a specific pathology that you expect to see. This task-based approach “can you see the pathology” is analogous to the ubiquitous Snellen eye chart to check whether you have 20/20 vision.
When the metrics from the quantitative image quality analysis tools are combined with the user responses to the simulations you will then have the information necessary for quantifying image quality benchmarks for consistent detection of a specific minimum edge case pathology across a fleet of imaging equipment.
Benefits
- Quantify diagnostic accuracy and minimise diagnostic related patient errors and harm
- Reduce healthcare costs to the patient and health system from diagnostic errors
- Answer the question - is our image quality adequate for the task
- Avoid cost, time, and ethical complexity, and potential harm of running imaging trials on patients
- Radiology staff engagement in managing and setting the quality of the core business
Features
- Arbitrary lesion simulation
- Lesions simulated either mathematically generated, hand-crafted or copied from real image
- Unlimited scoring questions
- Answer using scoring sliders, yes/no, picklist answers, measurement tools
- Study designer interface
- Lock studies when design completed
- Variable noise simulations
- Variable resolution simulation
- Locked or use adjustable, window and levels settings
- Image anonymization
- Predefined pseudorandom/image order presentation to minimise observer bias
- Hidden measurement values to minimise observer measurement bias
Collaborate
LesionSim runs from a central server, enabling you to set up a single Virtual Imaging Trial and invite participation from multiple locations within your organisation.
