DART
Research type
Research Study
Full title
The Integration and Analysis of Data Using ARtificial InTelligence to Improve Patient Outcomes with Thoracic Diseases
IRAS ID
301420
Contact name
Fergus Gleeson
Contact email
Sponsor organisation
University of Oxford / Research Governance, Ethics & Assurance team
Duration of Study in the UK
1 years, 9 months, 30 days
Research summary
In the UK, lung cancer is common with a very low 5-year survival rate as most patients are diagnosed at a late stage. Early detection on a CT scan when the cancers are small and seen as a nodule has been shown to improve survival.
DART will work with NHS England’s ambitious Lung Cancer Screening programme using CT to collect clinical, CT and histology data for research aimed at improving lung cancer diagnosis and screening using artificial intelligence, AI.
If DART is successful, using artificial intelligence we will speed up the time to diagnose lung cancer whilst also identifying incidental harmless nodules on CT. DART aims to: remove the need for other investigations such as lung biopsies, making investigations safer and quicker; help pathologists diagnose lung cancer using; help patients by providing their doctors with more information on lung and heart function; improve patient selection for lung cancer screening.
Dart aims to improve screening using AI, resulting in the avoidance of additional tests and biopsies which cause great patient anxiety, take time and are expensive.
DART will develop an AI algorithm for histology so that specimens from lung biopsies and resections can also be analysed in a similar fashion to CT scans.
Patients with lung cancer often have damaged lungs from smoking making surgery or radiation treatment unsafe. DART plans to develop an AI technique that can be used on all lung CT scans performed. As smoking can cause heart disease, patients screened for lung cancer often have heart disease. DART aims to use AI to see if we can identify this from their CT scans.
We will develop a specific risk model for Lung Cancer Screening selection, that outperforms published risk models that have been developed in academic institutions but are not used in clinical practice.
REC name
West Midlands - Black Country Research Ethics Committee
REC reference
21/WM/0278
Date of REC Opinion
7 Dec 2021
REC opinion
Favourable Opinion