The Prostate_Cancer_CISH_HE_Epithelium_Segmentation dataset

<h2>The Prostate_cISH_Epithelium_Segmentation Dataset</h2>

<p>
<strong>Corresponding author:</strong> Henrik Sahlin Pettersen ([e-mail])<br>
Consultant Pathologist / Associate Professor, St. Olav's Hospital / NTNU, Trondheim, Norway
</p>

<hr>

<h3>Short Webpage Description</h3>
<p>
This dataset provides high-resolution histopathological images and corresponding expert-annotated segmentation masks, specifically designed for developing AI models for prostate epithelium segmentation. The images feature Chromogenic In Situ Hybridization (cISH) staining for various miRNAs, alongside controls and standard Hematoxylin & Eosin (HE) stains. Data originates from 70 patients: 30 with prostate cancer (<strong>PCa</strong>) and 40 with benign prostatic hyperplasia (<strong>BPH</strong>).
</p>

<h3>Sample Collection</h3>
<ul>
<li><strong>Prostate Cancer (PCa):</strong> 30 patients. Samples include normal glandular epithelium (core type <code>'a'</code>), Gleason 3 pattern (core type <code>'b'</code>), and Gleason 4 pattern (core type <code>'c'</code>), ideally in triplicate for each marker.</li>
<li><strong>Benign Prostatic Hyperplasia (BPH):</strong> 40 patients. Samples consist of triplicate normal glandular epithelium for each marker.</li>
</ul>

<h3>Markers and Controls</h3>
<p>Data is provided for the following stains, each organized into its own top-level folder:</p>
<ul>
<li><strong>miRNAs:</strong> miR‑550A, miR‑1246, miR‑3614, miR‑4326, miR‑4632, miR‑4742, miR‑4754, miR‑7850</li>
<li><strong>Controls:</strong> U6 (Positive), Scr (Negative)</li>
<li><strong>Standard Stain:</strong> Hematoxylin & Eosin (HE)</li>
</ul>

<h3>Data Format & Organization</h3>
<p>
Each high-resolution image (<code>.jpg</code>) has a corresponding pixel-level segmentation mask (<code>.png</code>) delineating the prostate epithelium, suitable for training deep learning models. Segmentation masks are single-channel images where pixel value <strong>0 indicates background</strong> and pixel value <strong>255 indicates epithelium</strong>.
</p>
<p>
The data is organized first by marker, then by tissue type/origin. Within each marker's top-level folder (e.g., <code>HE/</code>, <code>550A/</code>), the structure is:
</p>
<pre><code>
[Marker]/
├── Normal/
│ ├── Normal_TURP_BPH/ (Images/masks from BPH patients)
│ └── Normal_Prostatectomy/ (Normal core 'a' images/masks from PCa patients)
└── Cancer/
└── Cancer_Prostatectomy/ (Gleason 3/4 core 'b'/'c' images/masks from PCa patients)
</code></pre>
<p>
All image and mask files are located directly within the innermost folders (<code>Normal_TURP_BPH</code>, <code>Normal_Prostatectomy</code>, <code>Cancer_Prostatectomy</code>). Filenames encode marker, patient type, anonymized ID, sample number, core type, and optional experimental details.
</p>

<h3>Terms of Use</h3>
<p>
Distributed under a CC0 license for open research and development.
</p>