If the symptoms of Alzheimer’s disease, in particular memory loss, are well known to scientists, making a diagnosis when the patient is suffering from an atypical form of this disease is still sometimes difficult today. This is why researchers of the Brain Institute have developed an automated algorithm making it possible to correlate certain specificities of the brain lesions of patients with these forms of Alzheimer’s.
The disease evolves differently depending on its form
Memory loss is often the first symptom of Alzheimer’s disease that can guide the diagnosis. Then come disorders of executive functions, orientation, then gradually settle disorders of language, movement and mood disorders. “However, the clinical presentation of the patients is very heterogeneous and different subtypes of the disease have been described, including an unusual non-genetic form which evolves very rapidly, in less than 12 years, until the death of the patient” explain the researchers .
The objective of this artificial intelligence project, baptized STRATIFIAD, is to make it possible to identify the particularities in the cerebral lesions of the atypical forms of Alzheimer’s disease already known, but also to discover new variants of the disease.
“This would be the first algorithm allowing automated classification of brain lesions in Alzheimer’s disease. This approach could complement the often manual and tedious and possibly biased analysis work carried out by expert neuropathologists.” The results will be able to guide clinicians in terms of diagnosis and allow them to adapt the treatment to the evolution of each patient.
