<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: LorDave1</title>
    <description>The latest articles on DEV Community by LorDave1 (@lordave1).</description>
    <link>https://dev.to/lordave1</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4252506%2F553e1aaa-ffd6-4ec4-89c8-c132762928d9.png</url>
      <title>DEV Community: LorDave1</title>
      <link>https://dev.to/lordave1</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/lordave1"/>
    <language>en</language>
    <item>
      <title>I'm not a scientist, so I started an open-source dementia project for people</title>
      <dc:creator>LorDave1</dc:creator>
      <pubDate>Wed, 18 Nov 2026 19:20:08 +0000</pubDate>
      <link>https://dev.to/lordave1/im-not-a-scientist-so-i-started-an-open-source-dementia-project-for-people-36ik</link>
      <guid>https://dev.to/lordave1/im-not-a-scientist-so-i-started-an-open-source-dementia-project-for-people-36ik</guid>
      <description>&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;The classic signal shows up: slow theta rhythms rise and alpha falls. Age and sex alone only reach about 59%, so the EEG is doing the work. But telling Alzheimer's from FTD is much harder: our best attempt (* a person-level model plus a second, eyes-open recording) reaches 63.3%, only a little above guessing. More people, not more features, is probably the fix. That's the open problem, and where I'd love help.&lt;/p&gt;

&lt;p&gt;Model selection uses nested cross-validation, so the numbers aren't tuned on test people. Honest caveats: a small single-site sample, and it's a research benchmark, &lt;strong&gt;not a diagnostic tool&lt;/strong&gt;.&lt;/p&gt;

&lt;h2&gt;
  
  
  How it works
&lt;/h2&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Experts&lt;/strong&gt; (doctors, scientists) share insights through a short form, no Git needed, or join as scientific advisors.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Builders&lt;/strong&gt; pick those ideas up as issues and implement them on open data.&lt;/li&gt;
&lt;li&gt;Results go back to experts for review (&lt;code&gt;needs-expert-review&lt;/code&gt;) before anything is presented.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Every change is reviewed, CI blocks committed data files and wording that sounds like a diagnosis, and everyone, expert or developer, is credited by name.&lt;/p&gt;

&lt;h2&gt;
  
  
  Where you can help
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Finding more open dementia EEG datasets (more people is the main bottleneck)&lt;/li&gt;
&lt;li&gt;Pretrained EEG foundation models for Alzheimer's vs FTD&lt;/li&gt;
&lt;li&gt;Testing on a second, independent EEG dataset&lt;/li&gt;
&lt;li&gt;Leading a new module: tau spread simulation, drug repurposing with knowledge graphs, or speech markers&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Website: &lt;a href="https://lordave1.github.io/openengram/" rel="noopener noreferrer"&gt;https://lordave1.github.io/openengram/&lt;/a&gt;&lt;br&gt;
Code: &lt;a href="https://github.com/LorDave1/openengram" rel="noopener noreferrer"&gt;https://github.com/LorDave1/openengram&lt;/a&gt;&lt;br&gt;
Cite (Zenodo DOI): &lt;a href="https://doi.org/10.5281/zenodo.24281637" rel="noopener noreferrer"&gt;https://doi.org/10.5281/zenodo.24281637&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;If you work in neurology, neuroscience or ML, I'd be grateful for your critique. That's exactly what the project needs.&lt;/p&gt;

</description>
      <category>opensource</category>
      <category>machinelearning</category>
      <category>healthcare</category>
      <category>python</category>
    </item>
  </channel>
</rss>
