Deep-MI · m-reuter · Jul 13, 2026 · Jul 13, 2026 · Jul 13, 2026
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -34,7 +34,7 @@ jobs:
         run: |
           bundle exec htmlproofer ./_site \
             --disable-external \
-            --ignore-urls "/^http:\/\/127.0.0.1/,/^http:\/\/0.0.0.0/,/^http:\/\/localhost/,/\/fsqc\//,/\/LaPy\//,/\/BrainPrint/,/\/FastSurfer\//" \
+            --ignore-urls "/^http:\/\/127.0.0.1/,/^http:\/\/0.0.0.0/,/^http:\/\/localhost/,/\/fsqc\//,/\/LaPy\//,/\/BrainPrint/,/\/FastSurfer\//,/\/neuroreg\//,/\/hipsta\//,/\/WhipperSnapPy\//" \
             --allow-hash-href
 
       - name: Upload built site

diff --git a/_config.yml b/_config.yml
@@ -4,12 +4,15 @@ description: "This is the official web page for the Deep Medical Imaging Lab at
 url: "https://Deep-MI.org"
 
 paginate: 5
-plugins: [jekyll-paginate]
+plugins: [ jekyll-paginate ]
 paginate_path: "news/page:num"
 
 permalink: pretty
 markdown: kramdown
-kramdown: 
+kramdown:
   parse_block_html: true
 
-exclude: [vendor]
+exclude:
+  - vendor
+  - .agentbridge
+  - .agent-work
diff --git a/_data/members.yml b/_data/members.yml
@@ -33,7 +33,7 @@
   and structural shape analysis for computer-aided diagnosis and prognosis. His methods are
   widely employed as part of the open source FreeSurfer software package to study neurodegeneration and
   assess disease modifying therapies, e.g., by the Alzheimer''s Disease Neuroimaging Initiative,
-  the Rhineland Study, and other large cohort studies.
+  the [Rhineland Study](https://www.rheinland-studie.de/en/research/research-areas/), and other large cohort studies.
 
 
   After receiving a competitive NIH Career Award (2014) for his research on computational methods for medical image
@@ -51,7 +51,7 @@
   github: santiestrada32
   description: 'Santiago graduated from Technical University of Munich (TUM) with a degree in
   Biomedical Computing. His research focuses on the deployment of deep learning methods for
-  quantifying imaging biomarkers in large cohort studies (e.g. [Rhineland Study](https://www.rheinland-studie.de/)).
+  quantifying imaging biomarkers in large cohort studies (e.g. [Rhineland Study](https://www.rheinland-studie.de/en/research/research-areas/)).
   When not at the lab, Santiago enjoys going to the movies and exploring the city.'
   projects: 'Olfactory Bulb Segmentation, Localized Brain Age Prediction, Automated Tools for Retina Image Analysis'
 

diff --git a/_data/projects.yml b/_data/projects.yml
@@ -4,6 +4,7 @@
 #   name:        display name
 #   tag:         short category label shown on the card
 #   featured:    true → teal top border on card
+#   homepage:    true → shown on the front page (all projects appear on /research/)
 #   description: one or two sentences shown on the card
 #   image:       path to card thumbnail (leave blank for a letter placeholder)
 #   url:         link for "Learn more" — use a local path or external URL
@@ -13,6 +14,7 @@
   name: FastSurfer
   tag: Neuroimaging Pipeline
   featured: true
+  homepage: true
   description: >
     Fast, accurate, deep-learning pipeline for whole-brain segmentation in under
     a minute and full surface reconstruction in about an hour.
@@ -21,6 +23,42 @@
   url: /research/fastsurfer/
   github: deep-mi/FastSurfer
 
+- id: neuroreg
+  name: NeuroReg
+  tag: Image Registration
+  homepage: true
+  description: >
+    GPU-accelerated Python toolkit for neuroimaging registration — same-modality,
+    cross-modal, boundary-based, and multi-timepoint — plus transform and volume
+    utilities. Replaces FreeSurfer's mri_robust_register, mri_coreg, and
+    bbregister. Usable from the command line or as a library.
+  image: /static/pub/reuter_2010_invconsregistration.jpg
+  url: /neuroreg/
+  github: deep-mi/neuroreg
+
+- id: hipsta
+  name: HipSTA
+  tag: Shape Analysis
+  homepage: true
+  description: >
+    Automated geometry-based method for hippocampal shape and thickness analysis
+    from structural MRI, enabling fine-grained morphometric studies of the
+    hippocampus.
+  image: /static/pub/diers_2023.jpg
+  url: /hipsta/
+  github: deep-mi/hipsta
+
+- id: neurocade
+  name: NeuroCade
+  tag: Neuroimaging Platform
+  homepage: true
+  description: >
+    AI-powered neuroimaging platform combining an MRI viewer with integrated
+    pipelines like FastSurfer. Run complex structural imaging workflows through
+    a GUI or conversational AI — no command line required.
+  image: /static/img/research/neurocade/teaser.jpg
+  url: https://neurocade.org
+
 - id: fatsegnet
   name: FatSegNet
   tag: MRI Segmentation
@@ -45,6 +83,7 @@
 - id: p3cnn
   name: p3CNN
   tag: Cortical Surface
+  homepage: true
   description: >
     Cortical surface parcellation using projection-based 3D convolutional neural
     networks operating on spherical mesh representations.
@@ -54,13 +93,25 @@
 - id: lapy
   name: LaPy
   tag: Python Library
+  homepage: true
   description: >
     Python library for geometric mesh processing and finite element method (FEM)
     differential geometry computations on triangular and tetrahedral meshes.
   url: /LaPy/
   image: /static/pub/reuter_2009_lbeigenvaluesfunctions_lapy.png
   github: deep-mi/LaPy
 
+- id: whippersnappy
+  name: WhipperSnapPy
+  tag: Visualization
+  description: >
+    Python/OpenGL tool for rendering brain surface meshes with color overlays and
+    generating publication-quality screenshots from the command line, Jupyter
+    notebooks, or an interactive desktop GUI.
+  image: /static/img/research/whippersnappy/teaser.jpg
+  url: /WhipperSnapPy/
+  github: deep-mi/WhipperSnapPy
+
 - id: fsqc
   name: FastSurfer QC
   tag: Quality Control

diff --git a/index.html b/index.html
@@ -167,11 +167,19 @@ <h1>AI-powered analysis of the human brain.</h1>
 <!-- RESEARCH PROJECTS -->
 <section>
     <div class="lab-wrap">
-        <h2 class="lab-section-title">Research &amp; Projects</h2>
-        <p class="lab-section-lead">We build open, validated computational tools for neuroimaging &mdash;
-            from whole-brain segmentation pipelines to geometric shape analysis libraries.</p>
-        <div class="lab-proj-grid">
-            {% for project in site.data.projects %}
+        <div class="home-team-strip">
+            <div>
+                <h2 class="lab-section-title">Research &amp; Projects</h2>
+                <p class="lab-section-lead" style="margin-bottom:0;">We build open, validated computational tools for
+                    neuroimaging &mdash;
+                    from whole-brain segmentation pipelines to geometric shape analysis libraries.</p>
+            </div>
+            <a class="lab-btn lab-btn-outline" href="/research/" style="white-space:nowrap;flex-shrink:0;">
+                View all research &rarr;
+            </a>
+        </div>
+        <div class="lab-proj-grid" style="margin-top:2rem;">
+            {% for project in site.data.projects %}{% if project.homepage %}
             <a class="lab-proj-card{% if project.featured %} featured{% endif %}" href="{{ project.url }}">
                 {% if project.image %}
                 <img class="lab-proj-img" src="{{ project.image }}" alt="{{ project.name }}">
@@ -185,7 +193,7 @@ <h2 class="lab-section-title">Research &amp; Projects</h2>
                     <span class="lab-proj-link">Learn more &rarr;</span>
                 </div>
             </a>
-            {% endfor %}
+            {% endif %}{% endfor %}
         </div>
     </div>
 </section>

diff --git a/research/fatsegnet.md b/research/fatsegnet.md
@@ -25,7 +25,10 @@ layout: lab
 
 <div class="lab-wrap rp-page">
 
-FatSegNet is a fast and fully automated deep learning pipeline to accurately identify, segment, and quantify visceral and subcutaneous adipose tissue (VAT and SAT) within a consistent, anatomically defined abdominal region on Dixon MRI scans. The pipeline implements a three-stage design with our Competitive Dense Fully Convolutional Network (CDFNet) architecture at its core for localizing the abdominal region and segmenting abdominal adipose tissue.
+FatSegNet is a fast and fully automated deep learning pipeline to accurately identify, segment, and quantify visceral
+and subcutaneous adipose tissue (VAT and SAT) within a consistent, anatomically defined abdominal region on Dixon MRI
+scans. The pipeline implements a three-stage design with our Competitive Dense Fully Convolutional Network (CDFNet)
+architecture at its core for localizing the abdominal region and segmenting abdominal adipose tissue.
 
 {::nomarkdown}
 <figure class="rp-fig">
@@ -36,7 +39,9 @@ FatSegNet is a fast and fully automated deep learning pipeline to accurately ide
 
 ## CDFNet Architecture
 
-Our CDFNet adopts a Dense-Unet architecture extended toward competitive learning via maxout activations. The Maxout operation computes the maximum at each spatial location across feature maps instead of concatenating them. Benefits of competitive learning through Maxout activations:
+Our CDFNet adopts a Dense-Unet architecture extended toward competitive learning via maxout activations. The Maxout
+operation computes the maximum at each spatial location across feature maps instead of concatenating them. Benefits of
+competitive learning through Maxout activations:
 
 - It promotes dedicated local sub-networks and improves feature selectivity through local competition.
 - Each sub-network can learn a sub-problem more efficiently.
@@ -51,14 +56,22 @@ Our CDFNet adopts a Dense-Unet architecture extended toward competitive learning
 {:/nomarkdown}
 
 For more information on competitive learning:
-- Estrada et al. "Competition vs. concatenation in skip connections of fully convolutional networks." MLMI 2018. [arxiv](https://arxiv.org/pdf/1807.07803.pdf)
-- Goodfellow et al. "Maxout networks." arXiv:1302.4389 (2013). [arxiv](https://proceedings.mlr.press/v28/goodfellow13.pdf)
-- Srivastava et al. "Understanding locally competitive networks." arXiv:1410.1165 (2014). [arxiv](https://arxiv.org/abs/1410.1165)
-- Liao &amp; Carneiro. "Competitive multi-scale convolution." arXiv:1511.05635 (2015). [arxiv](https://arxiv.org/abs/1511.05635)
+
+- Estrada et al. "Competition vs. concatenation in skip connections of fully convolutional networks." MLMI
+  2018. [arxiv](https://arxiv.org/pdf/1807.07803.pdf)
+- Goodfellow et al. "Maxout networks." arXiv:1302.4389 (
+  2013). [arxiv](https://proceedings.mlr.press/v28/goodfellow13.pdf)
+- Srivastava et al. "Understanding locally competitive networks." arXiv:1410.1165 (
+  2014). [arxiv](https://arxiv.org/abs/1410.1165)
+- Liao &amp; Carneiro. "Competitive multi-scale convolution." arXiv:1511.05635 (
+  2015). [arxiv](https://arxiv.org/abs/1511.05635)
 
 ## Proof of Concept
 
-FatSegNet was tested and validated in the [Rhineland Study](https://www.rheinland-studie.de/), a large prospective cohort study based in Bonn, Germany. We evaluate the whole pipeline with respect to robustness and reliability against two independent test sets: a manually edited set and a test-retest set.
+FatSegNet was tested and validated in
+the [Rhineland Study](https://www.rheinland-studie.de/en/research/research-areas/), a large prospective cohort study
+based in Bonn, Germany. We evaluate the whole pipeline with respect to robustness and reliability against two
+independent test sets: a manually edited set and a test-retest set.
 
 {::nomarkdown}
 <table class="rp-table">
@@ -100,10 +113,15 @@ FatSegNet was tested and validated in the [Rhineland Study](https://www.rheinlan
 
 ## News
 
-- FatSegNet was selected as June 2020 [MRM Highlights](https://blog.ismrm.org/category/highlights/) by the ISMRM community — [Q&amp;A interview](https://blog.ismrm.org/2020/06/19/qa-with-santiago-estrada-and-martin-reuter/) · [reproducible research interview](https://blog.ismrm.org/2020/07/03/reproducible-research-insights-with-santiago-estrada-and-martin-reuter/).
-- The tool has been deployed in 2,738 participants from the [Rhineland Study](https://www.rheinland-studie.de/), processing 16,520 liters of abdominal adipose tissue.
-- We are collaborating with [GenerationR](https://generationr.nl/) to improve domain adaptation to a wider population (age &lt; 30).
-- One of our paper images appeared on the cover of the April 2020 edition of [Magnetic Resonance in Medicine](https://onlinelibrary.wiley.com/toc/15222594/2020/83/4).
+- FatSegNet was selected as June 2020 [MRM Highlights](https://blog.ismrm.org/category/highlights/) by the ISMRM
+  community — [Q&amp;A interview](https://blog.ismrm.org/2020/06/19/qa-with-santiago-estrada-and-martin-reuter/) · [reproducible research interview](https://blog.ismrm.org/2020/07/03/reproducible-research-insights-with-santiago-estrada-and-martin-reuter/).
+- The tool has been deployed in 2,738 participants from
+  the [Rhineland Study](https://www.rheinland-studie.de/en/research/research-areas/), processing 16,520 liters of
+  abdominal adipose tissue.
+- We are collaborating with [GenerationR](https://generationr.nl/) to improve domain adaptation to a wider population (
+  age &lt; 30).
+- One of our paper images appeared on the cover of the April 2020 edition
+  of [Magnetic Resonance in Medicine](https://onlinelibrary.wiley.com/toc/15222594/2020/83/4).
 
 {::nomarkdown}
 <figure class="rp-fig">

diff --git a/research/index.html b/research/index.html
@@ -12,20 +12,33 @@
         gap: 1.25rem;
         margin-top: 2rem;
     }
+
     .res-dir-card {
         background: #fff;
         border: 1px solid var(--lab-border);
         border-left: 4px solid var(--lab-accent);
         border-radius: var(--lab-radius);
         padding: 1.25rem 1.4rem;
     }
+
     .res-dir-card h3 {
-        font-size: 1rem; font-weight: 700; margin: 0 0 0.35rem;
+        font-size: 1rem;
+        font-weight: 700;
+        margin: 0 0 0.35rem;
     }
+
     .res-dir-card p {
-        margin: 0; font-size: 0.91rem; color: var(--lab-muted); line-height: 1.6;
+        margin: 0;
+        font-size: 0.91rem;
+        color: var(--lab-muted);
+        line-height: 1.6;
+    }
+
+    @media (max-width: 680px) {
+        .res-directions {
+            grid-template-columns: 1fr;
+        }
     }
-    @media (max-width: 680px) { .res-directions { grid-template-columns: 1fr; } }
 </style>
 
 <!-- PAGE HEADER -->
@@ -56,7 +69,8 @@ <h1>Research &amp; Projects</h1>
                     <p class="lab-proj-desc">{{ project.description }}</p>
                     {% if project.github %}
                     <span class="lab-proj-link">
-                        <img src="/static/img/logos/github_logo.svg" alt="" style="height:0.85em;vertical-align:middle;margin-right:0.25rem;opacity:0.6">
+                        <img src="/static/img/logos/github_logo.svg" alt=""
+                             style="height:0.85em;vertical-align:middle;margin-right:0.25rem;opacity:0.6">
                         {{ project.github }}
                     </span>
                     {% else %}
@@ -104,7 +118,8 @@ <h3>Computational Geometry &amp; Topology</h3>
                 <h3>Neurodegeneration &amp; Brain Development</h3>
                 <p>Understanding brain development and neurodegeneration (aging, Alzheimer's,
                     Huntington's disease) via multi-modal imaging of large population cohorts
-                    such as the Rhineland Study.</p>
+                    such as the <a href="https://www.rheinland-studie.de/en/research/research-areas/">Rhineland
+                        Study</a>.</p>
             </div>
         </div>
     </div>
@@ -120,7 +135,8 @@ <h2 class="lab-section-title">Publications</h2>
         </p>
         <div style="display:flex;gap:1rem;justify-content:center;flex-wrap:wrap;">
             <a class="lab-btn lab-btn-accent" href="/publications/">View all publications &rarr;</a>
-            <a class="lab-btn lab-btn-outline" href="https://scholar.google.com/citations?user=CbdJk1wAAAAJ&hl=en" target="_blank" rel="noopener">Google Scholar &rarr;</a>
+            <a class="lab-btn lab-btn-outline" href="https://scholar.google.com/citations?user=CbdJk1wAAAAJ&hl=en"
+               target="_blank" rel="noopener">Google Scholar &rarr;</a>
         </div>
     </div>
 </section>
diff --git a/static/img/research/neurocade/teaser.jpg b/static/img/research/neurocade/teaser.jpg
diff --git a/static/img/research/whippersnappy/teaser.jpg b/static/img/research/whippersnappy/teaser.jpg