About this Research Topic
With increasing life expectancy and declining fertility, both the developed and developing countries have witnessed a rapidly growing proportion of persons older than 65 years in their population. As a result of the aging population, age-related neurodegenerative disorders, such as dementia (e.g., Alzheimer's disease [AD], Parkinson's disease dementia [PDD], cerebrovascular disease [CVD], and Lewy body disease [LBD]) and mild cognitive impairment (MCI) which can progress to dementia, are becoming serious public health problems. Numerous studies have been conducted to explore risk factors and to better understand the underlying pathogenesis. It is now widely believed that such age-related neurodegenerative disorders are multifactorial involving genetic, epigenetic, environmental and life style factors. Despite decades of considerable efforts and resources devoted to research and drug development, few pharmaceutical treatment has been developed that are clinically proven to be effective in reversing the deterioration or slowing down the progression of such neurodegenerative disorders. It is therefore of great importance to identify subjects of high risk and/or accurately predict the timing of onset of the disorders so that preventive interventions, such as cognitive training, increased physical activity and blood pressure management in persons with hypertension, can be applied in a timely manner to delay onset or decelerate disease progression in old age.
A variety of statistical methods have been employed in for modeling trajectory or risk of neurodegenerative diseases, such as change-point model, time-varying effect model (TVEM), group-based trajectory model, Cox proportional hazard model, logistic regression, and machine learning methods such as support vector machine, survival forests, conditional survival and supervised learning using time windows. Recently, with rapid advancement in artificial intelligence, deep learning has also been used. Meanwhile, different risk factors, biomarkers, environmental and life style factors, and neuropsychological features have been utilized used for prediction, including:
- demographic information such as age, gender, education and ethnicity, genetic data such as APOE genotypes,
- cognitive performance such as global cognition and cognitive domains,
- vascular risk factors such as hypertension, diabetes and heart disease,
- imaging data such as hippocampal volume, and precuneus and medial temporal cortical thickness,
- life style factors such as cigarette smoking, alcohol consumption and sleeping,
- and environmental factors such as pesticide exposure.
While a few of these approaches were reported to have good predictive ability, some were built with limited sample size, some used information less commonly collected, while others need further validation. More studies leveraging recent advances in statistics and computer science that have better performance are greatly needed, especially those that can accurately predict neurodegenerative disorders years before their onset.
In this Research Topic, we welcome scientists from various fields of research to report their findings in employing different approaches, such as novel statistical modelling, machine learning and deep learning, or using different risk factors or biomarkers, in an attempt to predict more accurately the trajectories, or the risk or time of onset of common age-related neurodegenerative disorders such as MCI, and dementia/AD dementia.
A variety of statistical methods have been employed in for modeling trajectory or risk of neurodegenerative diseases, such as change-point model, time-varying effect model (TVEM), group-based trajectory model, Cox proportional hazard model, logistic regression, and machine learning methods such as support vector machine, survival forests, conditional survival and supervised learning using time windows. Recently, with rapid advancement in artificial intelligence, deep learning has also been used. Meanwhile, different risk factors, biomarkers, environmental and life style factors, and neuropsychological features have been utilized used for prediction, including:
- demographic information such as age, gender, education and ethnicity, genetic data such as APOE genotypes,
- cognitive performance such as global cognition and cognitive domains,
- vascular risk factors such as hypertension, diabetes and heart disease,
- imaging data such as hippocampal volume, and precuneus and medial temporal cortical thickness,
- life style factors such as cigarette smoking, alcohol consumption and sleeping,
- and environmental factors such as pesticide exposure.
While a few of these approaches were reported to have good predictive ability, some were built with limited sample size, some used information less commonly collected, while others need further validation. More studies leveraging recent advances in statistics and computer science that have better performance are greatly needed, especially those that can accurately predict neurodegenerative disorders years before their onset.
In this Research Topic, we welcome scientists from various fields of research to report their findings in employing different approaches, such as novel statistical modelling, machine learning and deep learning, or using different risk factors or biomarkers, in an attempt to predict more accurately the trajectories, or the risk or time of onset of common age-related neurodegenerative disorders such as MCI, and dementia/AD dementia.
Keywords: Prediction, Neurodegenerative Diseases, Statistical Modelling, Machine Learning, Risk
Important Note: All contributions to this Research Topic must be within the scope of the section and journal to which they are submitted, as defined in their mission statements. Frontiers reserves the right to guide an out-of-scope manuscript to a more suitable section or journal at any stage of peer review.
