Automatic Brain MRI Image Labelling

More than 100,00 MRI examinations can be labelled in less than half an hour using new automated brain MRI image labelling technology.

Researchers from King's College London have used an automated brain MRI (magnetic resonance imaging) image labelling by deriving important labels from stored radiology reports and accurately assigning them to the corresponding MRI examinations.

About Magnetic Resonance Imaging(MRI)
Encyclopedia section of medindia gives general info about Magnetic Resonance Imaging

‘Using artificial intelligence technology deep learning, brain MRI images can be labelled automatically.’

Tweet it Now

A new study published in journal European Radiology is the first study that allows researchers to label complex MRI image datasets at scale.

Usually, it will take years to manually perform labelling of more than 100,000 MRI examinations.

Deep Learning is a subfield of machine learning concerned with algorithms inspired by the structure and function of the brain called artificial neural networks.

Artificial Intelligence Makes MRI Scans Safer for Patients
Artificial intelligence (AI) may soon make magnetic resonance imaging (MRIs) safer for patients by reducing the dose of gadolinium-based contrast agent, which is administered to patients to enhance image quality of the scan.

This technology requires tens of thousands of labelled images to achieve the best possible performance in image recognition tasks.

This will be a hindrance to the development of deep learning systems for complex image datasets like MRI which is fundamental to neurological abnormality detection.

Radiomics Can Predict Patients� Response To Chemotherapy
Radiomics, a field of study that systematically extracts patterns from CT scan images and then measures these patterns using artificial intelligence, has been found useful to predict patients’ response to chemotherapy in lung cancer.

Senior author, Dr Tom Booth from the School of Biomedical Engineering & Imaging Sciences at King's College London said: "By overcoming this bottleneck, we have massively facilitated future deep learning image recognition tasks and this will almost certainly accelerate the arrival into the clinic of automated brain MRI readers. The potential for patient benefit through, ultimately, timely diagnosis, is enormous”.

Researchers believe that their validation is stronger as they evaluated their model performance on unseen radiology reports, and unseen images.

Children With Epilepsy can Retain Visual Perception After Surgery: Here�s How
Brains of children with epilepsy are able to remap following surgery to retain visual perception. This enables the children to see, despite damage to the visual cortex of the brain.

Lead Author, Dr David Wood from the School of Biomedical Engineering & Imaging Sciences said: "This study builds on recent breakthroughs in natural language processing, particularly the release of large transformer-based models such as BERT and BioBERT which have been trained on huge collections of unlabeled text such as all of English Wikipedia, and all PubMed Central abstracts and full-text articles; in the spirit of open-access science, we have also made our code and models available to other researchers to ensure that as many people benefit from this work as possible”.

To perform the deep learning image recognition tasks, they need to overcome multiple technical challenges. Once this is achieved, they must ensure that the developed models can still perform accurately across different hospitals using different scanners.

This study was possible due to broad team of experts and a huge base of supporting organizers and facilitators who helped in delivering this research.

Obtaining clean data from multiple hospitals across the UK is an important step to overcome the next challenges. They are also running an NIHR portfolio adopted study across the UK to prospectively collect brain MRI data for this purpose.

Source-Medindia