AI for CT scans Imaging systems to support radiology teams

Patients suspected of COVID-19 infection are tested using the 'gold standard' RT-PCR assay obtained via a nasal/throat swab. However there can be significant time delays in processing such samples. More seriously, such swab tests can misdiagnose patients, giving a supposedly clean bill of health to patients who actually do have COVID-19 disease - the so called 'false negative' problem.

This can be extremely serious for vulnerable patients with underlying medical conditions and for whom the earliest possible intervention is critical.

CT scans offer a quick (< 1hr) and efficient alternative means of diagnosing COVID-19, the radiographic data providing radiologists with evidence of disease etiology.

COVID-19 lesions in the lungs can be difficult to discern from community acquired pneumonias, other pulmonary disorders and against the diverse background of 'normal' lung conditions - particularly in the earliest stages of infection.

"This is where Artificial Intelligence (AI) could play a critical role in supporting radiology analysis, by training deep learning algorithms on a large dataset of CT scans of known status, and ensuring that all of the images to be used have been standardised across the various imaging protocols used at different CT scan facilities," says Dr Aaron Golden, School of Mathematics, Statistics and Applied Mathematics at NUI Galway.

About the project

Dr Golden aims to build an AI imaging system to support radiology teams in expedited diagnosis of early stage COVID-19 disease using CT scans.

The project will use cutting-edge AI infrastructure to standardise thousands of publicly available chest CT scans.

By using a type of machine learning algorithm called Generative Adversarial Networks, the team aim to correct for variations in imaging protocols between differing radiology facilities.

This will allow for the re-assessment of the performance of two recently published open source convolutional neural network based classifiers developed by medical researchers in China that have been previously shown to demonstrate high sensitivity/specificity in diagnosing COVID-19 in CT scans in selected patient groups with a high incidence of this disease.

Collaboration with clinicians

Collaboration with clinician radiologists will be key to this project, and Dr Golden will work closely with collaborators in University Hospital Galway.

The project team includes Dr Christoph Kleefeld (Medical Physics & Clinical Engineering, University Hospital Galway) and Dr Declan Sheppard (Clinical Director of Radiology, University Hospital Galway).

Their expertise will inform the characterisation of the full range of lesions identified in this process so as ensure expert oversight in highlighting regions-of-interest (ROI) on flagged scans.

Eventually, the project team envisage the deployment of the trained AI solution to a stand-alone desktop system to allow radiologists to import CT scans and have a predicted diagnostic assessment and highlighted ROI within minutes in the clinic.

"Our project aims to build an AI imaging system to support radiology teams in the diagnosis of COVID-19 disease using CT scans - our turn-key system will be trained on thousands of different types of archival image data and when used to examine a suspected COVID-19 patient's CT scan, will be able to screen and report 'actionable' lesions in minutes, expediting the radiology teams' ability to identify the subtle differences between COVID-19 disease and other more common lung disorders, particularly in patients for whom immediate attention is critical.", explains Dr Golden.

The project is funded by Ireland's national COVID-19 Rapid Response Research and Innovation funding, via the Health Research Board and Irish Research Council. #CovidResearchIreland

About the PI

Dr Aaron Golden is a Lecturer in the School of Mathematics, Statistics and Applied Mathematics, a member of the School's Bioinformatics and Biostatistics Research Cluster and a PI in the Ryan Institute.

Dr. Golden is a physicist by training with 24 years experience working as a data scientist. He has successfully lead diverse teams working on projects covering topics in the earth observation, astronomical and biomedical sciences, with funding support from Science Foundation Ireland, Enterprise Ireland & the Environmental Protection Agency. Between 2011-2016 he benefited from extensive clinical research experience whilst an Associate Professor of Genetics (Division of Computational Genomics) at the Albert Einstein College of Medicine in New York City.