Combining Computational Neuroscience and Body Sensor Networks to Investigate Alzheimer’s Disease (pp. 29-38)
Authors: Newton Howard and Jeroen Bergmann
Abstract: Alzheimer’ disease is a syndrome of acquired cognitive defects that interferes with normal brain function. While its biochemical effects are well documented, the cause of Alzheimer’s disease is still not known and the best available interventions remain merely symptomatic. Different treatment modalities have been explored over the last few decades. Pharmaceutical interventions currently consist of medicines that focus on the neuropsychiatric symptoms and disease-modifying treatments. The importance of early intervention to the efficacy of treatments has led to an emphasis on detection and diagnosis in clinical research, and techniques using imaging, biomarkers and genetic information as tools for early detection have become prevalent. However, multifaceted non-invasive screening tools that incorporate computational algorithms and do not rely on imaging are not being widely developed. This paper argues that non-invasive computational methods such as BrainSpace, originally developed to explain mental processes, can be adapted to assist in the early detection and treatment of Alzheimer’s disease. BrainSpace uses cross-domain inference to combine multidisciplinary insights into Alzheimers into a single, coherent vocabulary accessible to all researchers. In addition, because temporal changes in behaviour and speech are occurring in the early stages of the disease, a Body Sensor Network (BSN) can be utilized to collect temporal information during everyday living, which can be further processed with an algorithm that allows for natural randomness. These methods show that novel non-invasive research and screening tools for Alzheimer’s disease can be devised based on measuring real-life behavior, and on incorporating insights from Artificial Intelligence as well.