Fujifilm develops AI tech for predicting Alzheimer’s development


Share post:

Fujifilm Company has developed AI expertise that may predict whether or not sufferers with gentle cognitive impairment will progress to Alzheimer’s illness inside two years.

In keeping with a press assertion, the AI Know-how for AD Development Prediction has been discovered to have excessive generalisability, predicting sufferers who would progress to AD with excessive accuracy, even for topics from totally different cohorts. 


The AI predictive expertise was constructed utilizing Fujifilm’s superior picture recognition applied sciences. Utilizing each AI and deep tech, atrophy patterns as seen on 3D MRI mind pictures have been extracted and calculated from the hippocampus and the anterior temporal lobe, each areas which are strongly correlated with the development of AD. From these patterns, AD development is predicted.

The expertise was skilled utilizing affected person information from the North American-Alzheimer’s Illness Neuroimaging Initiative (NA-ADNI), the world’s largest AD analysis mission. The research additionally used different scientific data equivalent to cognitive check scores.

Utilizing the info of North American and Japanese cohort affected person teams from the NA-ADNI and J-ADNI initiatives, the expertise was discovered to have a prediction accuracy of 88% and 84%, respectively. 

Findings from Fujifilm’s research with Japan’s Nationwide Middle of Neurology and Psychiatry (NCNP) have been printed within the journal npj Digital Drugs.


It’s projected that the variety of sufferers with dementia around the globe will develop to roughly 139 million by 2050 from about 55 million in the present day. Rising circumstances of Alzheimer’s illness, the commonest reason for dementia, are additionally predicted to proceed. 

Regardless of advances in AD remedy lately, many scientific trials haven’t been profitable given the low share of sufferers who progress from gentle cognitive impairment to AD inside two years and plenty of of such sufferers stay unchanged even when positioned beneath placebo. 

Fujifilm and NCNP intention to deal with these limits and contribute to precisely consider the efficacy of drug candidates for treating AD.

Shifting ahead, the companions search to foretell the pace of sufferers’ development to AD and examine the potential of bettering the success fee of their scientific trials by excluding affected person contributors who don’t progress to AD and lowering the hole within the distribution of development pace between management and remedy teams. Additionally they plan to conduct extra scientific trials quickly.

Furthermore, the organisations look to use the AI expertise’s algorithms to the mind pictures and scientific information of various psychological and neural illnesses, anticipating that it’s going to additionally have the ability to predict prognoses of psychological illnesses and responses to remedy. 


Two years in the past, Alzheimer’s Analysis UK began an identical initiative to enhance the detection of neurodegenerative illnesses like AD. It sought to develop and trial a diagnostic wearable system which may detect illness markers.

Final 12 months in November, US-based BrainCheck scored $10 million in Sequence B funding, proceeds from which have been deliberate for use to speed up the adoption of its cell and web-based platform for assessing cognitive decline and screening dementia.

Supply hyperlink


Please enter your comment!
Please enter your name here


Related articles