Detection of Alzheimer : With increasing age, memory power starts to weaken. It is usually caused by damage to the brain tissue. In a broader perspective, this mental disorder is called Alzheimer’s. According to the news of Hindustan newspaper, according to the World Health Organization (WHO), about 70 percent of dementia is the biggest contributor to Alzheimer’s. At present, the number of people suffering from it is 24 million worldwide and their number is estimated to double every 20 years. Dementia is not the name of any disease, rather it is the name of many diseases or rather a group of many symptoms.
more precise and sensitive
The difficulty is that till now neither its danger has been accurately predicted nor treatment is available. In such a situation, researchers have developed a deep learning based model, in which Alzheimer’s can be predicted with 99 percent accuracy through brain images. This research findings have been published in the journal ‘Diagnostics’.
This method of predicting Alzheimer’s risk is based on the analysis of MRI images of 138 subjects, which is more accurate, sensitive and specific than the old method.
the first sign of potential danger
Researchers from Kaunas University of Technology (KTU) Multimedia Engineering Researcher Rytis Muskeliunus says that doctors around the world are emphasizing on increasing awareness to identify in the first stage of Alzheimer’s, so that affected people can get better benefits from treatment. The first sign of a possible risk of Alzheimer’s is mild forgetfulness (MCI), the stage between aging and the expected cognitive decline of dementia.
According to earlier research, functional magnetic resonance imaging (fMRI) can be used to identify areas of the brain that are associated with a potential risk of Alzheimer’s. There are no obvious symptoms in the early stages of MCI, but in some cases it can be detected by neuroimaging.
Image analysis will speed up
While it is possible to identify Alzheimer’s-related changes through manual analysis of fMRI images, this not only requires specific information, but is also time-consuming. Whereas it can be accelerated by other methods of deep learning and artificial intelligence.
Developed on the basis of deep learning, this model has been prepared in collaboration with researchers from the artificial intelligence sector of Lithuania. In this, functional MRI has been classified by improving ResNet 18 (Residual Neural Network). These images were divided into 6 different categories, ranging from MCI to reaching Alzheimer’s disease status. The accuracy of this model was 99.95 percent.