Collaborating Authors

Model predicts patients at risk for Alzheimer's 2 years in advance


MIT researchers have developed a machine learning model that forecasts the cognitive decline of patients at risk for Alzheimer's disease by predicting their cognition test scores as much as two years in advance. Researchers, who will present a paper later this week at the Machine Learning for Health Care conference at the University of Michigan, contend that experiments indicate that accurate Alzheimer's predictions can be made looking six, 12, 18 and 24 months in advance. Also See: Machine learning helps to identify early signs of Alzheimer's At the Machine Learning for Health Care conference, researchers will discuss how their model could be used for clinical trials to improve the selection of candidate drugs and participating patients who are in the disease's early stages--before symptoms are evident and when treatment has the best chance of being effective. "Accurate prediction of cognitive decline from six to 24 months is critical to designing clinical trials," says Oggi Rudovic, a researcher at the MIT Media Lab. "Being able to accurately predict future cognitive changes can reduce the number of visits the participant has to make, which can be expensive and time-consuming. Apart from helping develop a useful drug, the goal is to help reduce the costs of clinical trials to make them more affordable and done on larger scales."

The five: ways to slow the onset of Alzheimer's

The Guardian

Last week, a US biotechnology company claimed to have produced the first drug with the ability to slow down the development of Alzheimer's. Biogen says it hopes to release aducanumab on to the market after it gets US Food and Drug Administration approval, which could take up to two years. Research into the drug had been abandoned but trials using higher doses of the drug are claimed to improve cognitive functions such as memory, orientation, and language. A Mediterranean diet has been suggested to reduce cognitive decline. Limiting saturated fats and simple carbohydrates benefit the cardiovascular system, having an effect on overall health.

Artificial Intelligence may Predict Alzheimer's Disease


Combining machine learning method -- a type of artificial intelligence -- with a special MRI technique may help physicians predict who is more likely to develop Alzheimer's disease, a study says. Machine learning is a type of artificial intelligence that allows computer programs to learn when exposed to new data without being programmed. "With standard diagnostic MRI, we can see advanced Alzheimer's disease, such as atrophy of the hippocampus," said principal investigator Alle Meije Wink from VU University Medical Centre in Amsterdam. "But at that point, the brain tissue is gone and there's no way to restore it. It would be helpful to detect and diagnose the disease before it's too late," Meije Wink explained.

What Is The Earliest Sign Of Alzheimer's Disease? Decline In Glucose Levels Signals Onset Of Cognitive Degeneration; Alzheimer's Prevention May Be Possible

International Business Times

Memory loss and cognitive decline are commonly thought to be the earliest signs of the neurodegenerative disorder Alzheimer's, but a new study has found declines in glucose levels in the brain come even sooner -- before the first symptoms appear. The same team also believes they have figured out a way to stop these levels from falling in the first place, a finding that could potentially prevent Alzheimer's. Although doctors have long noted the association between declining glucose levels in the brain and the onset of Alzheimer's disease, for the first time ever, a study now published online in the journal Translational Psychiatry has proved that these declining energy levels are a direct trigger for the cognitive impairments traditionally associated with the disease. According to a recent statement on the study, this may explain why diabetes, a condition in which glucose cannot enter the cells, is a known risk factor for dementia. According to the study, a protein known as p38 may be able to prevent this deprivation from occurring.

IBM takes on Alzheimer's disease with machine learning


IBM has introduced machine learning (ML) to the diagnostics field in the hopes that one day these technologies may assist in the creation of stable and effective diagnostic tests for early-onset Alzheimer's. On Monday, the tech giant said ML and artificial intelligence (AI) can be exploited to replace invasive and expensive tests for the disease. A paper documenting the research, conducted by IBM's Australian team, has been published in Scientific Reports. Alzheimer's is currently incurable and can only be treated by palliative means. Symptoms for the disease include the gradual degradation of memory, confusion, and difficulty in completing once-familiar daily tasks.