Adhd research

[cjnolet]

I’m beginning to get really interested in machine learning and data science research to discover some new way of being able to detect the disorder.

It looks like much current research has been focused on fMRI images combining the spatial-temporal dimensions of the data in order to make predictions. (I believe this approach has become so prevalent since it’s nowhere near as invasive as, say, having to draw blood to study DNA.

I did find one such DNA study and I’m hoping to get a hold of the dataset through my school so I can play around with ways to improve upon current methods and maybe find something useful.

Any of you guys been diving into recent innovations in the biotech field? Specifically I’ve found some papers that use the same dataset on the fMRI data. the dataset contains individuals with HIV, Alzheimer’s, and ADHD, as well as “normal” individuals who are not afflicted with either of these. The most recent paper, published in February of this year, only achieves 70% accuracy on the ADHD dataset.

Would be neat to be a part of the solution here, being an ADHDer myself. Definitely a project I can hyperfocus on...

[Halem1982]

Can’t dna be taken by swabbing the inside of our mouth? That’s how my dr took mine when I requested a dna test that showed which medications would be most effective for me. I’m also a stem cell and bone marrow donor if my dna is matched with anyone with a blood cancer, and they swabbed the inside of my mouth too to put me in their system.

You’re right though, there has to be a better way to diagnose and treat adhd along with other mental health disorders, because most dr’s aren’t qualified enough to make these life altering decisions for us. I hope you do find a better way!! Maybe going to the annual CHADD or ADDA conventions would put you in front of the right people to collaborate with?

[cjnolet in reply to Halem1982]

Ah, you are right about swabs. I was trying to think of a reasonable justification for why there seems to be so much more effort put into the fMRI than into DNA and the research I was looking at used blood samples. I'll have to dive much deeper for sure. I don't know a whole lot about the domain of neuroscience/genetics yet but my soon-to-be advisor does mostly health-related research. He knows I have ADHD and recently told me to research a very specific machine learning topic. While I was looking up papers on that topic, I happened to noticed this series of published ADHD papers. Makes me wonder if he did that on purpose

Have you been to either of those conventions? Do they normally have doctors/researchers there?

[Halem1982 in reply to cjnolet]

I’ve been to both on multiple occasions. I feel blessed to have been able to meet some of the most respected people in the field at these conventions. I pretty much stalked Dr. Hallowell to get an appt with him when I was going through a relapse on meth. He was so nice and he even gave me his cell phone number and told me to call him at 3am if I needed to. Then I went to see him in Boston a few months later. There are so many wonderful people there. Writers, therapists, psychiatrists, coaches, scientists, patients, spouses or parents of patients. It all had a true grassroots vibe when I went because literally everyone I met except for one group of people who were trying to sell some sort of app for substance abuse, were all there to do whatever they could to help others. It seemed like the majority of the professionals at the event weren’t the average professionals we have to use as our dr’s at home. They all seemed to have a deeper connection with the disorder. They have it themselves or they’re a parent of a child with it or their spouse has it. I love going to both events and I plan on going to the CHADD conference in November.

[HadEnuf in reply to cjnolet]

There are a few known alleles associated with a part of the ADHD population; but because the DX is associated with several, distinct, overlapping sets of traits that produce similar outcomes, there's no genotype that applies consistently enough.

If the selection pressure arguments hold as much water as they appear to, it's entirely possible there are several genotypes that, through convergent evolution, produce similar-enough neurotypes to appear to be a single syndrome.

[Halem1982 in reply to HadEnuf]

HadEnuf, I always love reading your responses to see if I understand what the heck you’re talking about. It’s like watching jeopardy. Lol. How do you understand all of this stuff so well? I wish I had a cliff’s notes version of your responses😂😂😂😂

I’ve never been very good at reading. Menus even annoy me sometimes. Lol

[HadEnuf in reply to Halem1982]

Honestly, it doesn't seem like all that much, to me...I spend a lot of time around people I think are brilliant, though. (That's supposed to come off like a Holiday Inn Express commercial, BTW.)

All I really am is insatiably curious—probably explains my affinity for teh kittehs.

[cjnolet in reply to HadEnuf]

Sounds like a life time of tireless stimulation chasing has given you quite an amazing understanding of the world around us.

[HadEnuf in reply to cjnolet]

I don't know about tireless!

There's a lot of others' homework behind it: I just happened to notice and remember instead of filtering it out.

Sometimes it's a curse, sometimes it's a superpower, same as the opposite tendencies are for others: humanity needs both, in the long run.

[Ami00 in reply to Halem1982]

Ive got a library of hyperfocused infromation that I can recall instantly. 5-8 track mind simultanously following some short term interest. Mostly of no use because I lose interest to ever do anything with the info. One subject after another - after months of bingewatching netflix or something. SOrt of gave up some years ago that I can make a positive contribution to someone or something. Endless started projects that will never see the light of day. Self esteem is non-existent. Fighting like hell to save my second attempt at being a father and husband.

[HadEnuf]

I've always wanted to find something that's a strong indicator of how one _processes_ and _responds_ to a readily-controlled stimulus; and there have been some inroads along those lines, recently—I recall some study that found an extremely strong correlation between a response timing and ADHD diagnosis.

This might actually suffer fewer confounding factors that anatomic scans of any kind.

The type of tests used to probe for visual field loss in patients deemed vulnerable to glaucoma might actually be a good starting point, given the tendency, with ADHD, to notice stimulus peripheral to one's focus. Perhaps change the objective to be more center-of-field focused and introduce extra ”blips” well outside the target area?

(The classic example in my own life would be recalling a series of flash cards drawn in front of me—except for one drawn in the presence of a background disturbance, like someone entering the room, even behind me.)

[cjnolet in reply to HadEnuf]

Wow. That just made me realize something. One of my research interests is to be able to parse videos of crowds at clubs and be able to contextualize the dynamics of the crowd in order to, eventually, introduce my own stimulus and influence the crowd (kind of like an automated dj system with a feedback loop).

This problem of parsing the sensor data; and more specifically, video, isnt much different than the field of traffic analysis. The feedback portion that you mentioned also sounds very similar in theory to being able to the objective of predicting the next song to play; as it requires contextualizing and inferencing based on continuous feedback. That does sound very similar to other devices that already exist- especially in the medical community.

Very good feedback! (No pun intended)

[iBusyBrain]

Machine learning is very interesting science. I have worked with three- dimensional data arrays, but have yet to learn in depth about the machine learning field of study--wouldn't take much though. I have also entertained the idea of going into robotics--but I think my time would be largely building retail or military application robots. So I think machine learning would have a broader spectrum of application, to include health care research.

[Wylly]

I know there are ADHD associated genes but heard 1) that there are many and 2) there are none that come near to being a good predictor. This would seem to indicate that there are environmental factors and/or ADHD is due to an interaction between genes. I am sure some have looked at interactions. You could probably get at this with a CHAID analysis. You would, however, have to specify it in a way that the cells are not too small. My guess is that you will only be able to explain a small portion of ADHD but it could help to identify sub-types and be used to guide treatment eventually.

I am somewhat of data geek and have some free time in the next couple. I would be happy to look at any data you can get your hand on.

[StoneJeweler]

Good for you!