By selecting "Allow All, " you are consenting to the storage of cookies on your device for improved site navigation, analysis of site usage, and support in offering unrestricted access to scientific content. More information. A recent study from the University of Gothenburg reveals that a new method can yield comparable information from brain images obtained via computed tomography (CT) as those obtained through magnetic resonance imaging (MRI). This method has the potential to enhance diagnostic support, especially in primary care settings, for various brain disorders, including dementia. Even though CT is a widely available and cost-effective imaging technology in healthcare, it is considered inferior to MRI in capturing subtle structural changes in the brain and changes in the ventricular system. Currently, specialized departments in larger hospitals are required for certain advanced imaging technologies. However, the newly developed software, which utilizes deep learning as a form of artificial intelligence (AI), can provide diagnostic assistance to radiologists and other professionals in interpreting CT images. By training the software to transfer interpretations from MRI images to CT images of the same brains, the researchers have been able to generate diagnostically valuable data from routine CT scans that are comparable to specialized MRI scans. Professor Michael Schöll, who led the study, envisions that this straightforward and efficient method can provide additional information from routine examinations conducted in primary care settings as well as specialized investigations in healthcare. Initially, the method has been employed to support the diagnosis of dementia, but it is also expected to have other applications in neuroradiology. This clinical application of AI-based algorithms has been well-validated and has the potential to serve as a rapid and reliable decision-making support system that reduces the number of false negatives. The researchers anticipate that this solution can enhance diagnostics in primary care, thereby improving patient flow to specialist care.

The lead author of the study, Meera Srikrishna, a postdoctoral researcher at the University of Gothenburg, describes this as a significant advancement in imaging diagnosis. The software enables the measurement of the size of various brain structures or regions similar to advanced analysis of MRI images, despite the relatively lower image quality of CT scans. The software was trained using images from 1, 117 individuals who underwent both CT and MRI imaging, including healthy older individuals and patients with different forms of dementia. Another area of investigation for the team is normal pressure hydrocephalus (NPH), where early results indicate that the method can aid in diagnosis and monitoring treatment effects. NPH predominantly affects older individuals, causing an accumulation of fluid in the cerebral ventricular system and resulting in neurological symptoms. About 2% of individuals over the age of 65 are affected by this condition, but it often goes undiagnosed or is mistakenly attributed to other diseases due to its challenging diagnosis process. Furthermore, safely evaluating the efficacy of shunt surgery to alleviate brain fluid build-up can also be difficult. The team believes that their method can have a significant impact on improving the care of NPH patients. The software has been developed over several years and is currently being further developed in collaboration with clinics in Sweden, the UK, and the US, as well as a company, to ensure its approval and implementation in healthcare settings. In a separate interview conducted at SLAS EU 2023 in Brussels, Belgium, Laura Kuett, a Ph. D. student in Systems Biology, discusses the potential of 3D Imaging Mass Cytometry. Thermo Fisher Scientific also shares insights on ensuring the high quality of raw materials in drug production and accelerating drug discovery through analytical R&D in two other interviews.