Blood Cell Image Processing for Rapid Accurate Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing blood cell analysis methods are prone to errors, slow, and time-intensive, hindering timely diagnosis and treatment of blood disorders.
Innovation Solution
A method involving image preprocessing with contrast amplification, de-noising, edge detection, and classification using convolutional neural networks to accurately identify and classify blood cells and abnormalities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual blood cell analysis methods are used, then pathologists can identify cell types and abnormalities, but the process is slow and time-intensive
Solution Approach 1:
The patent replaces manual mechanical analysis by pathologists with an automated image processing system using contrast amplification, de-noising, edge detection, and classification algorithms. This substitution of mechanical/manual processes with automated computational methods achieves both high identification accuracy and rapid diagnosis speed, resolving the contradiction between measurement precision and productivity.
2Reliability
If manual blood cell analysis methods are used, then pathologists can diagnose blood disorders, but the overall diagnosis efforts and time increase
Solution Approach 1:
The automated image processing system replaces time-consuming manual analysis while maintaining diagnostic reliability. The systematic approach using contrast amplification, de-noising, edge detection, and classification ensures accurate identification of blood cell abnormalities, thereby reducing diagnosis time without compromising reliability.
Solution Approach 2:
The patent performs preliminary image processing operations (contrast amplification, de-noising, edge detection) before final classification. These preliminary actions prepare the images for accurate analysis, enabling rapid and reliable diagnosis by pre-processing the visual data in a systematic manner before the classification stage.
3Measurement precision
If image processing is applied to enhance smear image quality, then cell type identification accuracy improves, but processing complexity increases
Solution Approach 1:
The patent segments the image processing task into distinct sequential stages: contrast amplification, de-noising, edge detection, and classification. Each stage addresses a specific aspect of image enhancement and analysis, making the overall complex process more manageable and systematic while improving cell identification accuracy through targeted processing at each stage.
Data Source
AI summary
A method of image processing and classifying target entities with an image is disclosed that may include applying a contrast amplification procedure to a Lightness parameter associated with an input image, to amplify contrast of the input image and obtain an amplified-contrast image. The method may further include de-noising the amplified-contrast image by iteratively performing on the amplified-contrast image a blur correction, an erosion correction, and a dilation correction, to obtain a de-noised image corresponding to the amplified-contrast image. The method may further include determining edges of each of one or more target entities associated with the input image from the de-noised image and identifying the one or more target entities associated with the input image based on the identified edges, to generate a contoured image. The method may further include classifying the one or more target entities into one or more predefined classes, using a classification model.