As a technophile, AI has always stayed around as the core of my fascination. The way it has evolved has grasped my attention prompting me to stay up to date with even minute developments. Over time one of my obsessions over this topic was whether AI would take over human lives or not?
Now imagine my surprise when I just recently stumbled across an article “AI hacks into the human brain.” My immediate thought was of “Patrick Jane” from mentalist. The guy, whose mind manipulation had always enthralled me, Oh well!
It is sufficed to say that my interest skyrocketed, and as a true science fiction believer, I went through the text. The article is a research, which is in short regarding the use of AI in the treatment of Alzheimer’s, depression as well as in recognition of the potential of a possible suicide. I’ve taken the liberty to explain the content of the research below:
Usage of AI for Alzheimer’s treatment:
You all must know that Alzheimer’s happens to be a common disease with it affecting almost 50 million affected people worldwide. In it, the victim’s brain cells start to die, and he loses his memory, the way it happened to “Allie and Noah “in “Notebook,” sad isn’t it?
The disease is said to occur due to come complex interchange between the genes and some other unknown factors. Unfortunately, the only known cure available is to slow down the progression of the disease which is why it is somewhat crucial to identify it in the early stages.
Now as it is a slow progression disease its identification during the early stages is somewhat next to impossible as explained by Jae Ho Sohn, MD, MS, a resident in the Department of Radiology and Biomedical Imaging at UC San Francisco,
“One of the difficulties with Alzheimer’s disease is that by the time all the clinical symptoms manifest and we can make a definitive diagnosis; too many neurons have died, making it essentially irreversible.”
With this being said I am sure you’re wondering where AI does come in. Well, the recent researches work on the combination of neuroimaging and machine learning. Glucose is one of the main energy components for the brain, so it is also used as a primary key for the diagnosis of Alzheimer’s.
It is a known fact that the active cells use up more glucose and as they began to die their usage of glucose decreases. What the doctors do is that they carry out PET scans where they look up various molecules, preferably glucose. The PET scans of the whole brain are carried out. Low levels of glucose are considered to be alarming as they are a sign of deceased cells.
Now as aforementioned it is a slowly progressive disease, so identification of low glucose levels is somewhat hard to identify with the naked eye, which is why an algorithm to look up low glucose levels is design.
This algorithm is trained using a large number of the sufferers PET scans along with Alzheimer’s disease Neuroimaging Initiative. The researchers used 67 magnetic resonance images out of which 38 were of Alzheimer’s patients while the rest were of a control group.
The researchers then analyzed the neural connectivity between the parts of the brain after dividing it into a small region. It was discovered that the algorithm made correct classifications of the disease when the brain regions were 2250 to 3200 cubic millimeters, same as the size of the anatomical structures linked with Alzheimer’s, e.g., the hippocampus.
Next, another analysis was conducted with 148 brain scans – 52 healthy, 48 ill and the rest of the patients suffering from mild cognitive impairment, who were predicted to develop Alzheimer’s later in life, and voila! The software provided 86% accuracy in identification of sick brains.
Prevent Potential Suicides Through AI
As sad as it is suicide happens to be one of the major causes of deaths throughout the world. Around 80,000 people are estimated to die due to suicide. With it being such a grave issue psychologist has looked up ways of suicide prevention.
Once again AI has been an awesome approach to avoid suicide. An algorithm is developed to identify potential suicide victims whose research was carried out using 34 young adults. These people were divided into two groups
- Suicidal group
- Control group
These groups had to go through a functional magnetic resonance imaging (fMRI) with which they were presented with three lists of ten words
- Suicidal words
- Happy words
- Sad words
Other brain images of shame and anger were also used.
All in all these things in combination allowed the scientists five locations that were related to six words. Lo and behold! These six words ended to be the key highlighters in the identification of suicidal thoughts.
The software was then fed this information and bam! It was able to identify 15 out of 17 suicidal patients.
Apart from these algorithms are best friends Facebook and Google actively use up AI to contribute to the prevention of suicide occurrence. Facebook carries out this process by scanning your posts.
It takes up to find out people who show patterns of suicide or suicidal thoughts through their posts. Once it comes across someone who actively portrayers suicidal thoughts it intervenes and sends the posts to human moderators. They respond by sending mental health resources to users.
Similarly, if someone searches up “ways to kill or harm yourself” on Google, it immediately responds by directing the user to a page that provides 24/7 suicide prevention lifelines as well as an online chat room to overcome the need to suicide.
Another stupendous feature offered by Google is that the autocomplete for suicidal searches is turned off. This is featured as the suicidal and self-harm searches come up under dangerous and harmful behavior. The desired results for these searches are also not displayed.
Fight Depression And Anxiety Through AI
Depression and anxiety are considered to be the two most disturbing issues faced by our society. The movie “perks of being a wallflower” paint a true picture of depression and anxiety as an all-consuming illness. The fact that it commonly occurs in young teens is even sadder.
However, AI can actively be used to as to fight against them. One method discovered by the researchers is Woebot. This tool is a friendly chat box that lightens up a depressive and anxious person through its quirky humor.
The highly effective cognitive behavioral therapy is being used, and it also provides monthly reports. Furthermore, it allows you to keep track of your moods through graphs and charts. This chat box becomes your best friend by providing a fruitfully therapeutic conversation.
Another phenomenal program is Crises Text Line which that allows the identification of suicide risks. It uses deep neural networks and natural language processing, but the best part is that its 24/7 cost-free service! Also, it efficiently facilitates 98% of the users!
All in all, I am truly all up and enthusiastic for the use of AI in neuroscience. With it contributing in the world against such major problems it is best to say AI has once again no failed to disappoint me and is truly a magnificent aspect of working on.