Artificial Intelligence (AI) is the future of medical care. It is the key to diagnosing, caring for, and developing treatments for individuals with dementia. In the past year itself we have come a long way in terms of developing technologies that can aid medical professionals treating individuals with different neurodegenerative disorders. Here is what AI has accomplished in regards to the respective neurodegenerative disorders:
Alzheimer’s and Frontotemporal Dementia
Alzheimer’s and Frontotemporal dementia, along with other neurodegenerative disorders cannot be diagnosed with 100 percent certainty during an individual’s lifetime. Medical professionals have to wait until post-mortem to diagnose with clarity the certain type of neurodegenerative disorder the individual was dealing with. Thanks to the development of AI we no longer have to wait that long. The University College London’s(UCL) Institute of Healthcare Engineering was able to use machine learning to develop an algorithm that can diagnose with clarity the disorder a patient is suffering from. The algorithm is known as SuStaIn (Subtype and Stage Inference). Researchers at UCL fed SuStaIn with MRI (magnetic resonance imaging) of the brains of patients dealing with Alzheimer’s and Frontotemporal dementia. Now, SuStaIn is able to display to medical professionals the specific parts of the brain where neurodegeneration (death of neurons) is occurring and classify the distinct subgroups of both Alzheimer’s and Frontotemporal dementia. Classifying these neurodegenerative disorders into subgroups allows for medical professionals to provide proper medication to individuals. This is due to the fact that treatments for one subgroup does not essentially work for another one.
Alzheimer’s and Frontotemporal dementia both have three distinct subgroups which can be identified by SuStaIn. The three subgroups for Alzheimer’s include inflammatory, non-inflammatory, and cortical. Patients suffering from the inflammatory subgroup have an imbalance in their albumin (protein that carries medicines and hormones across the body) and globulin (protein that fights infections and is responsible of the transport of nutrients) ratios. Patients dealing with the non-inflammatory subgroup have an increase in beta-amyloid (protein) markers as compared to the other subgroups. Finally, patients suffering from the cortical subgroup don’t have a gene related with zinc deficiency present. The three subgroups for Frontotemporal dementia include behavioral (causes personality changes), semantic, and progressive(aka non fluent aphasia) both which cause language deficiencies.
Professor Paul Bentley and Professor Daniel Rueckert of the University of Edinburgh have led a team of scientists to develop a type of software that can detect the early signs of vascular dementia. The software is able to detect small vessel disease (SVD). In SVD, walls of small vessels in the heart are damaged due to the buildup of plaque and is one of the precursors of heart disease such as ischemic strokes (blood being blocked from entering the brain due to the narrowing/blockage of arteries). Multiple strokes can lead to vascular dementia, and therefore identifying SVD is crucial in steps for prevention. Without the software, medical professionals have to diagnose SVD by looking for white matter (tissue in the brain that contains nerve fibers and transports signals between brain cells) changes. It is strenuous to look for these changes using the naked eye. The new software allows for a precise and accurate diagnosis of the disease. Researchers are using deep learning to train the software. Currently, while comparing the software against medical professionals in the UK, it is accurate 85 percent of the time. The goal is that this software can be used worldwide for the diagnosis of SVD. This can help doctors prescribe the proper treatment options for the patient and prevent vascular dementia (second most common type of neurodegenerative disorder after Alzheimer’s).