Artificial Intelligence To Identify Your Brain Age using MRI Scans

Posted By : Oodles Technologies | 13-Dec-2016

Artificial Intelligence to identify your Brain Age

 

Had you ever thought of calculating your Brain Age.. ? If no, then here is the chance to determine your Brain Age within seconds with the use of a mere MRI scan. At King’s College of London, Giovanni Montana and a few of his pals have trained a machine which works on Artificial Intelligence and can easily calculate your brain age by just using some raw data from a MRI scan. It is called as Deep-Learning Artificial Intelligence machine. This clearly indicates the growth in Machine Learning and Data Mining Applications in today’s era.

 

Human intellectual abilities decline with increase in the age and the same is correlated with the anatomical changes in the brain. So, it becomes quite evident that signs of aging can be spotted through the MRI images of the brain and thus “Brain Age” can be determined.

 

This technology is considered to be exploring the cause behind dementia as the data received by calculating the difference between chronological age and brain age can reveal a lot of research-related information which could help in the treatment of the said ailment called dementia.

 

The Research Technique and the Observations Obtained

 

This technique can detect the brain age of a patient using MRI scanners within seconds and can provide an accurate data to the clinicians while the patient is still under the scanner. It is that quick.

 

The Artificial Intelligence based Technique known as Standard Deep-Learning utilised MRI scans of over 2000 healthy people between the age-group of 18-90 years. None of them was suffering from any kind of neurological condition which could influence their brain age. Therefore, one would suppose that their brain age is going to match their chronological age. Each of these scans were labelled with the chronological age of the related patients.

 

Their team utilized 80 percent of these obtained images to train a convolutional neural network for determining the age of a person in the context of their respective brain scan.To verify these observations they further used another such 200 images for that specific purpose to be 100% confirmed.

 

Finally, the team tested the said neural network on an unseen 200 images to get an idea about how well the deep learning machine can identify a brain age and up to what accuracy.

 

Comparison with the Standard Method of Determining Brain Age

 

The comparison was also made with the conventional method of determining the brain age and that required extensive image processing and breakdown of white and grey matter present in the brain from among various other things, and that followed a statistical analysis known as Gaussian process regression.

 

Readings obtained on Comparison of the two Brain Age determining Techniques

 

Both the Deep Learning and the Gaussian process regression method identified the correct chronological age of the patients tested given preprocessed data to analyze. Both the methods gave a deviation of less than five years when compared to the exact age of the patients.

 

Comparison of Readings Obtained Through the Two Methods with MRI Based Data

 

During the analysis of raw data obtained through MRI, Deep Learning proved itself far more superior in providing the accuracy as the results obtained with deep learning technique showed a mean error of 4.66 years while the observations obtained using the Gaussian process regression has shown a mean error of approximately 12 years in determining the rough age.

 

A Few Highly Considerable Merits of Deep Learning Method

 
  • It is a speedy process that takes only a few seconds when compared to the standard process which takes 24 hours of pre-processing to determine brain age of a person.

 
  • This technique can also be applied to the scans taken through different machines in different parts of the world.

 
  • This technique when applied on the twins has shown that the brain age could be linked to genetic factors but the correlation drops with age making it clear that the environmental factors affect the brain aging process to a great extent.

 

Also Read : Google DeepMind AI to Lipread Better Than a Human Professional

 

Diabetes, schizophrenia and traumatic brain injuries are a few of those conditions which are correlated to faster brain aging . So, in that way, the clinicians can get much required help in the near future for the treatment of these conditions having known the exact measures of somebody’s brain age. Apparently, Machine Learning and Data Mining technology has expanded in all directions over the past decade and that too in a positive manner. Hope that this new technology could save more people’s life and could give boost to the much hyped research work in the field of Artificial Intelligence.

 
 
 

About Author

Author Image
Oodles Technologies

Vijay is a Blogger and a Web Content Writer. He is passionate about writing be it blogging, articles writing, technical writing etc. He loves reading , watching movies and playing outdoor games.

Request for Proposal

Name is required

Comment is required

Sending message..