A Gray-Level Dependence Matrix (GLDM) is a second-order texture analysis method that quantifies gray-level dependencies in an image. It measures how often a center voxel with a specific intensity value has a defined number of neighboring voxels within a set distance that share a similar intensity, characterizing the homogeneity of local texture patterns.
Glossary
Gray-Level Dependence Matrix (GLDM)
A statistical matrix that measures the frequency of a center voxel's intensity being dependent on its neighboring voxels within a defined distance, quantifying coarse and fine textural patterns in medical images.
Radiomics Feature Extraction
What is Gray-Level Dependence Matrix (GLDM)?
A statistical matrix quantifying the frequency with which a center voxel's intensity depends on its neighboring voxels within a defined distance, capturing coarse texture homogeneity.
Unlike co-occurrence matrices that analyze pixel pairs, GLDM focuses on the dependence count—the number of connected neighbors dependent on the center voxel. This makes it particularly effective for distinguishing between coarse, uniform regions and fine, heterogeneous textures in medical images, providing features like Small Dependence Emphasis and Large Dependence High Gray-Level Emphasis for predictive modeling.
QUANTITATIVE TEXTURE METRICS
Key GLDM-Derived Features
The Gray-Level Dependence Matrix (GLDM) quantifies the frequency a center voxel's intensity is dependent on its neighbors within a defined distance δ. The following features are derived from the resulting probability matrix P(i,j), where i is the gray level and j is the dependence count.
01
Small Dependence Emphasis (SDE)
Measures the preponderance of small dependence counts in the image. A high SDE value indicates a less homogeneous texture where center voxels rarely share intensity with many neighbors.
- Formula: Σᵢ Σⱼ [P(i,j) / j²] / Σᵢ Σⱼ P(i,j)
- Interpretation: High SDE suggests fine, heterogeneous textures.
- Clinical Context: Often elevated in aggressive tumors with chaotic cellular architecture.
02
Large Dependence Emphasis (LDE)
Quantifies the distribution of large dependence counts. A high LDE value indicates a coarse, homogeneous texture where center voxels are dependent on many neighbors at the same intensity.
- Formula: Σᵢ Σⱼ [j² · P(i,j)] / Σᵢ Σⱼ P(i,j)
- Interpretation: High LDE suggests large, uniform structural patterns.
- Example: Benign lesions often exhibit higher LDE than malignant ones due to organized cellular growth.
03
Gray Level Non-Uniformity (GLN)
Assesses the variability of gray-level intensity values across the dependence matrix. A lower value indicates greater homogeneity in intensity distribution.
- Formula: Σᵢ [Σⱼ P(i,j)]² / Σᵢ Σⱼ P(i,j)
- Interpretation: High GLN means many different intensity values are present.
- Technical Note: This feature is sensitive to the number of discrete gray levels chosen during intensity discretization.
04
Dependence Non-Uniformity (DN)
Measures the similarity of dependence counts throughout the image. A low value indicates that all dependence counts are equally distributed.
- Formula: Σⱼ [Σᵢ P(i,j)]² / Σᵢ Σⱼ P(i,j)
- Interpretation: High DN means the image is dominated by a few specific dependence sizes.
- Utility: Captures the structural complexity independent of raw intensity values.
05
Dependence Entropy (DE)
Quantifies the randomness in the distribution of dependence counts. It is derived from the marginal sum of dependence sizes.
- Formula: -Σⱼ [p(j) · log₂(p(j))] where p(j) = Σᵢ P(i,j) / Σᵢ Σⱼ P(i,j)
- Interpretation: High DE indicates chaotic, unpredictable texture patterns.
- Benchmark: IBSI consensus values provide standardized reference calculations for this feature.
06
Low Gray Level Emphasis (LGLE)
Highlights the distribution of low-intensity values within the dependence matrix. A high value indicates a concentration of dark, dependent structures.
- Formula: Σᵢ Σⱼ [P(i,j) / i²] / Σᵢ Σⱼ P(i,j)
- Interpretation: High LGLE suggests dark, necrotic regions dominate the texture.
- Clinical Use: Helps differentiate necrotic tumor cores from viable enhancing margins in contrast-enhanced scans.
COMPARATIVE ANALYSIS
GLDM vs. Other Texture Matrices
A feature-level comparison of the Gray-Level Dependence Matrix against other primary texture analysis matrices used in radiomics.
| Feature | GLDM | GLCM | GLRLM | GLSZM |
|---|---|---|---|---|
Spatial Relationship Measured | Dependence of center voxel on neighbors within a distance δ | Joint probability of pixel pairs at a specific offset and angle | Length of consecutive, collinear pixels with the same intensity | Size of connected, homogeneous regions of identical intensity |
Rotational Invariance | ||||
Captures Intensity Homogeneity | ||||
Captures Local Contrast | ||||
Computational Complexity | O(N * V) | O(N * G²) | O(N * G * L) | O(N * G * Z) |
Primary Clinical Application | Quantifying homogeneous texture in solid tumors | Assessing tissue anisotropy and local intensity variations | Characterizing structural roughness and fine texture | Measuring tumor heterogeneity and granularity |
IBSI Standardized |
Enabling Efficiency, Speed & Accuracy
Intelligent Analysis, Decision & Execution
We build AI systems for teams that need search across company data, workflow automation across tools, or AI features inside products and internal software.
Talk to Us
Search across company data
Give teams answers from docs, tickets, runbooks, and product data with sources and permissions.
Useful when people spend too long searching or get different answers from different systems.
Enterprise searchRAGPermissions
Read more
Automate internal workflows
Use AI to route work, draft outputs, trigger actions, and keep approvals and logs in place.
Useful when repetitive work moves across multiple tools and teams.
AI agentsWorkflow automationGovernance
Read more
Add AI to products and internal tools
Build assistants, guided actions, or decision support into the software your team or customers already use.
Useful when AI needs to be part of the product, not a separate tool.
AI integrationDecision supportModel routing
Read moreGLDM DEEP DIVE
Frequently Asked Questions
Explore the mechanics, clinical relevance, and technical nuances of the Gray-Level Dependence Matrix, a cornerstone of modern radiomic texture analysis.
A Gray-Level Dependence Matrix (GLDM) is a second-order statistical texture analysis matrix that quantifies the frequency at which a center voxel's intensity is dependent on its neighboring voxels within a defined distance. Unlike matrices that measure consecutive runs (GLRLM) or paired occurrences (GLCM), the GLDM specifically evaluates spatial dependence—how often a specific gray-level value appears as a connected region of a certain size. The algorithm iterates through every voxel in the Region of Interest (ROI), counting the number of connected neighbors (within a Chebyshev distance) that share the same discretized intensity. This count forms the 'dependence' value, populating a matrix where rows represent gray levels and columns represent dependence counts. This makes it exceptionally sensitive to homogeneous textural patterns.
About the author
Prasad Kumkar
CEO & MD, Inference Systems
Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.
His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.
LinkedIn
Limited slotsGet a Free AI Consultation
Partnered with leading AI, data, and software stack.
How We Work
Custom AI workflows for your Business
One-fit-all AI don't work for modern businesses. At Inferensys, we aim to understand your business & custom requirements; which we use to define most efficient agentic workflows, the data, and the tools for your business.
01
Review the use case
We understand the task, the users, and where AI can actually help.
Read more02
Pick the right approach
We define what needs search, automation, or product integration.
Read more03
Build the first useful version
We implement the part that proves the value first.
Read more04
Improve from there
We add the checks and visibility needed to keep it useful.
Read moreThe first call is a practical review of your use case and the right next step.
Talk to Us