ALS Credibility Prediction Software--Please Share Treatment Experiences

[markuswoltjer]

Dear ALS Patients and Families,

It's really wonderful to see a place like this so active with support for each other. Thank you for being here.

I'm a computer science student at Oregon State University and for my senior design project, my partners and I are building software that helps researchers at ALS Untangled determine the credibility of treatment offers based on their website design. To train our machine learning algorithm, we are looking to build a data set with treatments and ratings of their legitimacy. Please share your experiences here, and we will use them to improve our software.

It really broke my heart when I heard that scams targeted the communities of the most debilitating chronic diseases, and I had to do all I could to help. Hopefully our software will be of use, or at least provide a start for professionals to perfect.

Thanks,
Markus Woltjer

 

[Nikki J]

You are working with ALS Untangled??? That group is not going to evaluate treatments by website design and neither do we. We evaluate by validity ( or not) of scientific evidence provided if any

You are right we are targeted by unscrupulous individuals. I just don't understand your project. If this is true, will Dr Bedlack confirm it?

 

[David]

Maybe I read wrong, but his software program is going to predict the success rate of treatment based on the design of the website that offers the treatment?

With all respect due, that's ridiculous. I've seen very poorly designed websites that are in fact completely legitimate (.edu sites, etc); and I've seen highly professional sites that are completely scams, along with everything in between. I would hazard an educated guess that the design quality of a website is in fact, completely unrelated to the quality of the treatment offered. It would be like judging a person's level of honesty based on the clothing they wear.

 

[swalker]

I spent over 30 years in the software industry, with much of that as a hands-on software developer specializing in backend data systems and algorithm development.

In my experience, the quality of the algorithms used in the backend (as well as the overall quality of the backend) cannot be predicted by the quality of the user-facing software.

In addition, what looks like an appealing user interface to a novice is often a very impractical user interface for an experienced user. This can lead to incorrect conclusions about the efficacy and usefulness of a system when it's user interface is evaluated by someone who is not an experienced user of the system.

So, attempting to predict quality by looking at web site design seems like you are barking up the wrong tree.

I encourage you to rethink your approach. Doing so could be a very valuable part of your learning experience while doing your senior project.

Steve

 

[affected]

fascinating stuff!

 

[Diane H]

Marcus, I think I understand your intent. You want to design software to evaluate websites that promote treatments by using a database that identifies wording, lack of references to research study results, reliance on testimonial evidence, cost factors, cross-referencing to known scams and unproven treatments as well as to the findings of correctly done research and clinical trials, etc. Your intent is not to critique the treatment yourself, but rather the website by looking for indicators commonly seen in scams and unproven or disproven treatments. Correct? If so, I think your project is well worth while. Far too many people are drawn into trying very questionable and costly treatments by website promotions.

 

[KimT]

Hi Marcus,

If your project is as Diane summarized, I think it would be worthy of senior project material. I was a college professor for 30 years.

I would like to know how you got involved with ALS Untangled. I think once you answer that question, your reception here will be more welcoming.

Regards.

 

[markuswoltjer]

Maybe I read wrong, but his software program is going to predict the success rate of treatment based on the design of the website that offers the treatment?

With all respect due, that's ridiculous. I've seen very poorly designed websites that are in fact completely legitimate (.edu sites, etc); and I've seen highly professional sites that are completely scams, along with everything in between. I would hazard an educated guess that the design quality of a website is in fact, completely unrelated to the quality of the treatment offered. It would be like judging a person's level of honesty based on the clothing they wear.

It wouldn't be a binary decision, and it wouldn't be certain either. This would just preview the websites for ALS Untangled, such as suggest that a website in question is very similar to another. Some of the reasons I believe in this type of method is because defrauders are lazy and unprofessional. They can make websites that look okay, but they don't update regularly, they are built from templates that are other fraud attempts, and their use of frameworks and the like is different.

I'm not asking for an evaluation of my project, just looking for data that we can use to train it. If it isn't usable, at least I learned something. But we have a smart and dedicated team and we should be able to do something that at least helps others direct their efforts better.

 

[markuswoltjer]

Steve, thanks for your input. I appreciate your experience and am interested in taking it into account since we are still starting development. The project was proposed by a computer science professor at my university, Bill Smart, who met the ALS Untangled people at a conference. It's possible that our data isn't suitable. Regardless, I am invested in it and will do my best, and if that means reporting what flaws our data has that makes this type of scam detection unusable, then I will do that. That's part of science and engineering as I understand it. I would love to hear more about your insights.

 

[markuswoltjer]

Nikki, we are working under Dr. Bill Smart, and haven't corresponded with the researchers at Duke University. However, Dr. Smart said we may if something important comes up, but the tool we are creating doesn't need much input from his team. If you question the validity of our effort, hopefully this will confirm that.

ACPS Validity - Imgur

 

[markuswoltjer]

Hi Marcus,

If your project is as Diane summarized, I think it would be worthy of senior project material. I was a college professor for 30 years.

I would like to know how you got involved with ALS Untangled. I think once you answer that question, your reception here will be more welcoming.

Regards.

I don't know if my direct replies will show here publicly and are just waiting for approval. I'll respond in more detail after my classes.

For now,

http://m.imgur.com/account/MarkusWoltjer/images/2kzFXBv

 

[markuswoltjer]

You are working with ALS Untangled??? That group is not going to evaluate treatments by website design and neither do we. We evaluate by validity ( or not) of scientific evidence provided if any

You are right we are targeted by unscrupulous individuals. I just don't understand your project. If this is true, will Dr Bedlack confirm it?

Good question. I haven't emailed the Duke team. Our point of contact is alocal professor who spoke with them at a conference.

 

[swalker]

Thanks for clarifying your objectives. I will take a stab an providing some input.

I don't really know of any treatments. Most of what I have experienced is monitoring and support.

Here is a list of things that I believe make a positive difference:

1. Maintain weight
2. Get a feeding tube early (which I have not done)
3. Don't fall
4. Don't over exert yourself
5. Start using a wheelchair early so you don't fall and conserve your energy
6. Monitor your breathing through Pulmonary Function Test (PFT) or Spirometry
7. Monitor CO2 levels via Arterial Blood Gas tests
8. Get all the rest you need
9. Have a support system of friends and relatives ready and willing to step in and help so you are not tempted to expend your energy unnecessarily
10. Don't get sick (don't be around sick people)
11. See a doctor at the first sign of a respiratory infection
12. Avoid aspiration
13. Get a flu shot
14. Get a pneumonia shot
15. Work on gentle range of motion for affected joints

Here are things I think are not proven to make a difference (though some are under investigation)

1. Stem cell therapy (see Celtex for an example)
2. Special diets (though a good diet is wise)
3. Phytoconjugates
4. Various protocols

The bottom line is that we have a progressive, terminal disease that cannot be cured. We all hope for medical advancements that will change that. I am encouraged by some of the research I read about, but am doubtful that meaningful results will be available in time to help me.

So, anything that says "cure" is likely a scam (in my opinion).

Steve

 

[markuswoltjer]

Thanks for that Steve. I'm still interested in your technical recommendations for the project, though we are somewhat constrained, I would like to make it as useful as possible.

As for the advice, that is more or less what I expected from my research of ALS. Here I'm mainly looking for reviews of the symptom treatments, which are probably somewhat effective but not near a cure. The more input I get, the better.

 

[lgelb]

Hi, Markus, I hate to see you guys wasting time on this. The people most likely to be scammed are least likely to use an algorithmic filter (or any other evaluation criteria apart from their hope). And it doesn't take software outside the brain to know when something's a scam, if you're looking in that direction.

So my two cents would be to bag this one and find a basic science or translational project related to ALS (they are all underfunded, just about) that could use your expertise.

But you said you weren't interested in evaluations of what you are doing, only help in doing it. Honestly, I think Google is all you need. Basically, pretty much anything that has a branded site is a scam or unintentionally clueless. The generic tx like BiPAP are spread over thousands of sites, not one, but are the only tx of note so far.

And you won't find certainty here as to scams because of selection bias -- those here who try something the rest find fishy are more likely to report it worked -- that pesky cognitive dissonance --

Best,
Laurie