A new method developed by researchers at the University of Gothenburg can provide as much information from computed tomography (CT) brain images as magnetic resonance imaging (MRI) images. CT is a more affordable and accessible imaging technology compared to MRI, but it is considered inferior in capturing subtle structural changes in the brain. The new method uses deep learning, a form of artificial intelligence, to transfer interpretations from MRI images to CT images of the same brains. This method has the potential to improve diagnostic support in primary care, particularly for conditions like dementia. It can also be used in neuroradiology and has applications for other brain diseases such as normal pressure hydrocephalus.
The researchers believe that their developed software can be a reliable form of decision-making support, reducing the number of false negatives and optimizing patient flow to specialist care. The software has been trained on a dataset of 1,117 people who underwent both CT and MRI imaging. The initial study mainly involved healthy older individuals and patients with dementia. The team is also investigating the application of this method for normal pressure hydrocephalus diagnosis and monitoring the effects of treatment. The development of the software is continuing in collaboration with clinics in Sweden, the UK, and the US, along with a company, for it to be approved and transferred to healthcare.
Overall, this new method has the potential to revolutionize imaging diagnosis by generating useful data from routine CT scans. It can provide similar measurements and analysis capabilities as advanced MRI imaging, allowing for the segmentation and measurement of different structures or regions of the brain. In addition to dementia, the method can be applied to other brain disorders, such as normal pressure hydrocephalus. By improving diagnostics in primary care, this new method aims to enhance patient care and treatment.
