---
title: LC25000 Histopathology Dataset
url: https://www.emergentmind.com/topics/lc25000-dataset
type: topic
---

# LC25000 Histopathology Dataset

The LC25000 dataset is a public repository comprising 25,000 de-identified color histopathology images, curated for machine learning and deep learning research in cancer pathology. Released by Borkowski et al. [1912.12142], the dataset consists of five balanced classes—colon adenocarcinoma, benign colon tissue, lung adenocarcinoma, lung squamous cell carcinoma, and benign lung tissue. All images are RGB, H&E-stained patches (768×768 pixels), and have undergone rigorous annotation and HIPAA-compliant de-identification, making LC25000 a prominent benchmark for automated histopathological image analysis.

## 1. Dataset Composition and Class Semantics

LC25000 contains 25,000 JPEG histopathology images, evenly distributed across five classes ($N_{c_i} = 5{,}000$, $i=1,\dots,5$):

| Class Index | Histologic Entity       | Folder Name    | N (images) |
|:-----------:|:---------------------- |:--------------|:----------:|
| c₁          | Colon adenocarcinoma   | colon_aca     | 5,000      |
| c₂          | Benign colon tissue    | colon_n       | 5,000      |
| c₃          | Lung adenocarcinoma    | lung_aca      | 5,000      |
| c₄          | Lung squamous carcinoma| lung_scc      | 5,000      |
| c₅          | Benign lung tissue     | lung_n        | 5,000      |

Images are stored in a hierarchical directory (lung_colon_image_set/) with organ-specific subfolders. Every class has exactly 20% representation, such that $p(c_i) = 0.20$ for all $i$ [1912.12142, 2601.01056, 2509.02851, 2103.12155].

## 2. Image Acquisition, File Format, and Metadata

Original images were scanned from de-identified, HIPAA-compliant pathology glass slides. The acquisition protocol includes:

- Modality: bright-field (light microscopy)
- Staining: hematoxylin and eosin (H&E)
- Initial resolution: scanned at 1024×768 px; final standard crop is 768×768 px
- Color channels: standard RGB ($C=3$)
- File format: JPEG (typical quality ≈ 90%); some studies report lossless PNG/TIFF as alternatives [2601.01056]
- No explicit magnification level specified in the dataset release

All images are stripped of metadata; no patient identifiers remain. Final tensor form: $\mathbf{I} \in \mathbb{R}^{768 \times 768 \times 3}$ [1912.12142].

## 3. Data Preprocessing and Augmentation Protocols

LC25000 was constructed from 750 raw images (250 per lung class, 250 per colon class), expanded through stochastic augmentation as follows:

- Center-crop to 768×768 px
- Random rotation: angle $\theta \sim \mathcal{U}(-25^\circ, +25^\circ)$ with probability 1.0
- Horizontal and vertical flips: $p=0.5$ each
- No stain normalization or color-space augmentation beyond rotations/flips [1912.12142, 2009.03878]

Downstream experiments frequently adapt images for model input, with resizing to 224×224 px [2103.12155, 2205.15290, 2509.02851, 2510.16310], or to 299×299 px for Inception architectures [2601.01056]. Additional metrics include pixel-value scaling to $[0,1]$ and ImageNet-style normalization.

Complex augmentation pipelines superimpose geometric (rotation, flipping, scaling, shear), color/photometric (contrast, brightness, color jitter, Gaussian blur/noise) transforms during training [2103.12155, 2509.02851]. Some studies eschew augmentation entirely, processing images “as is” [2510.16310].

## 4. Annotation, Validation, and Ground Truth Integrity

Images in LC25000 are annotated per-class by board-certified pathologists. The dataset documentation states “validation” by experts, but does not publish inter-observer reliability statistics such as Cohen’s $\kappa$ [1912.12142]. No region-of-interest masks, bounding boxes, or patient-level metadata are included; only categorical (slide-level) class labels are provided uniformly across studies [2601.01056, 2510.16310, 2507.01279].

## 5. Data Partitioning for Machine Learning Experiments

The official LC25000 data release does not prescribe explicit training/validation/test splits. Consequently, studies have adopted a variety of stratified partitioning schemes:

- Hold-out splits: e.g., 60% train, 20% val, 20% test [2601.01056, 2205.15290]
- Three-way splits (lung-only): e.g., 68% train, 17% val, 15% test [2510.16310]
- Per-class splits (e.g., 75%/9%/15% train/val/test on random subsets) [2507.01279]
- K-fold cross-validation or leave-one-scanner-out procedures (recommended for batch-effect or domain generalization studies) [1912.12142]

All partitioning approaches preserve strict class balance.

## 6. Applications and Model Benchmarking

The LC25000 dataset underpins a large body of deep learning research for cancer histopathology:

- CNN and transformer-based architectures (ResNet, InceptionResNet-v2, VGG, Xception, MobileNet, DenseNet169, Vision Transformer) [2103.12155, 2205.15290, 2510.16310, 2601.01056]
- Hybrid architectures combining convolutional networks with transformer modules, graph-attention, capsule networks [2509.02851]
- Feature engineering: integration of HOG descriptors with deep features, block-wise histogramming, and concatenation protocols [2601.01056]

Notable performance benchmarks (see Table):

| Paper (arXiv)   | Model/Strategy        | Classification Accuracy (%) | Test Protocol                      |
|-----------------|----------------------|----------------------------|------------------------------------|
| 2601.01056      | InceptionResNet-v2    | 96.01                      | 5-class, 60/20/20 split            |
| 2601.01056      | Deep features + NN    | 99.84                      | HOG + deep feature fusion          |
| 2509.02851      | HG-TNet (hybrid)      | 96.0                       | 5-class, aggressive augmentation   |
| 2507.01279      | ResNet+ (subset)      | 98.14                      | 3-class, train/val/test split      |
| 2510.16310      | ResNet-50 (lung)      | 98.8                       | 3-class, 68/17/15 split            |
| 2205.15290      | Vision Transformer    | 100.0                      | 3-class, 5-shot finetuning         |
| 2103.12155      | 8 CNNs                | 96–100                     | 5-class, 80/20 + val split         |

Evaluation metrics include precision, recall, F1-score (macro/weighted), AUROC, and confusion matrices. Multi-class ROC AUCs consistently exceed 0.97 per class [2509.02851, 2601.01056, 2205.15290].

## 7. Accessibility, Licensing, and Usage Restrictions

LC25000 is distributed as a ~1.85 GB archive via multiple public repositories:

- Original GitHub: https://github.com/tampapath/lung_colon_image_set/
- Figshare mirror: https://figshare.com/articles/dataset/Lung_and_Colon_Cancer_Histopathological_Image_Dataset_LC25000_/11504065

The dataset is freely available to AI researchers for non-clinical research and educational use, with no explicit open-source license file. Attribution to Borkowski et al. (arXiv:1912.12142) is requested in publications. All images are de-identified and HIPAA-compliant; no patient re-identification is possible under U.S. regulatory standards.

---

Research employing LC25000 typically capitalizes on its uniform class balance, expert annotation, and rich augmentation protocols to benchmark classifier performance, develop domain-adapted pipelines, and advance computational pathology methodologies. The dataset remains a central resource for validating multi-class, organ-specific cancer detection algorithms and for methodological experimentation in digital histopathology.

Source: https://www.emergentmind.com/topics/lc25000-dataset