Mater hospital trials use of Artificial Intelligence in radiology department

Technology has led to rapid notification of suspected pathologies and the freeing up of time

Since the introduction of AI this year, more than 15,600 patient scans have been analysed, leading to more than 700 pathologies being flagged within just two to three minutes of a scan being completed.
Since the introduction of AI this year, more than 15,600 patient scans have been analysed, leading to more than 700 pathologies being flagged within just two to three minutes of a scan being completed.

The Mater Hospital in Dublin has reported that it has successfully used artificial intelligence to screen for illnesses and speed up emergency treatment.

Staff in its radiology department have become the first in an Irish hospital to embrace AI technology which, it said, has led to rapid notification of suspected pathologies and the freeing up of time.

Since its introduction this year, more than 15,600 patient scans have been analysed, leading to more than 700 pathologies being flagged within just two to three minutes of a scan being completed.

It also said 500 intracranial haemorrhages and 200 pulmonary emboli were identified. In outpatient care, 50 pulmonary emboli were detected by scans which were performed to assess other medical conditions.

READ SOME MORE

“The overall accuracy of the software in real Irish clinical practice has been deemed to be highly sensitive and specific, with an accuracy rate of more than 90 per cent,” the hospital said, announcing its results on Wednesday.

The new technology, including AI assistance, has been applied to CT scans that cover head and spinal areas.

The Aidoc screening platform runs constantly in the background to analyse medical imaging data, detect urgent issues and highlight priority cases which are brought to the immediate attention of consultants.

“Our experiences have underscored the tangible benefits of AI, notably in expediting critical diagnoses and reducing turnaround times by rapidly flagging anomalies detected in scans,” said consultant radiologist Professor Peter MacMahon.

“This allows radiologists to prioritise the reading of these results and verify their accuracy before deciding upon a further course of action.”

Between April and August, the software underwent a process of rigorous scrutiny, verification and testing before being fully deployed.

The hospital noted, however, that the software is used to “augment rather than replace” the capabilities of radiologists, ensuring human scrutiny, verification and expertise remains at the centre of operations.

Further trials of specialised AI tools focusing on analysis of bone and chest X-rays in the emergency department are planned next year.

  • Sign up for push alerts and have the best news, analysis and comment delivered directly to your phone
  • Find The Irish Times on WhatsApp and stay up to date
  • Our In The News podcast is now published daily – Find the latest episode here
Mark Hilliard

Mark Hilliard

Mark Hilliard is a reporter with The Irish Times