Browse Prior Art Database

Two Stage Wheal Segmentation from RGB Skin Prick Test Images

IP.com Disclosure Number: IPCOM000244509D
Publication Date: 2015-Dec-17

Publishing Venue

The IP.com Prior Art Database

Abstract

A new method for automated estimation of wheal size from skin prick test images. The method consists of two phases: a) wheal segmentation from captured images and b) estimating the diameter of the detected wheal in physical units. In the first stage of the segmentation, we perform a color transformation from RGB to YCbCr and discard the luminance channel (Y) to reduce the impacts of illumination variation and different skin tones on segmentation. We then performed unsupervised color segmentation (using for example, K-means, EM, fuzzy c-means etc) on the blue-difference (Cb) and red-difference (Cr) color channels, which provides a rough segmentation of the wheal and the surrounding erythema. The second stage then leverages an interactive segmentation algorithm (e.g., graph-cut, grow-cut, random walk etc.) for the fine segmentation of the wheal from the captured image. The seed points required in the second stage are automatically determined based on the rough segmentation calculated in the first stage. Following wheal detection, we estimate the physical size of the wheal by performing a perspective transformation using a plane-to-plane homography.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 14% of the total text.

Page 01 of 11

Two Stage Wheal Segmentation from RGB Skin Prick Test Images

Skin prick test is a commonly used method for diagnosis of allergic diseases (e.g., pollen allergy, food allergy, etc.) in allergy clinics. The results of this test are erythema and wheal provoked on the skin where the test is applied. The sensitivity of the patient against a specific allergen is determined by the physical size of the wheal. The diameter of the wheal is typically measured by a nurse or medical personnel using a ruler. This manual method for measuring wheal size is a cumbersome process and prone to errors and inter-observer variations. The SPT results of a patient can be interpreted differently due to measurement errors and inter-observer variations of different medical personnel, which can easily cause inconsistencies in the diagnosis of the patient.

In this disclosure, we describe a new method for automated estimation of wheal size from skin prick test images. The method consists of two phases: a) wheal segmentation from captured images and b) estimating the diameter of the detected wheal in physical units. The novelty of this ID lies in the first phase, where we propose a two stage method for the segmentation of wheal from captured RGB images. In the first stage of the segmentation, we perform a color transformation from RGB to YCbCr and discard the luminance channel (Y) to reduce the impacts of illumination variation and different skin tones on segmentation. We then performed unsupervised color segmentation (using for example, K-means, EM, fuzzy c-means etc) on the blue-difference (Cb) and red-difference (Cr) color channels, which provides a rough segmentation of the wheal and the surrounding erythema. The second stage then leverages an interactive segmentation algorithm (e.g., graph-cut, grow-cut, random walk etc.) for the fine segmentation of the wheal from the captured image. The seed points required in the second stage are automatically determined based on the rough segmentation calculated in the first stage. Following wheal detection, we estimate the physical size of the wheal by performing a perspective transformation using a plane-to-plane homography.


1. Background

Skin prick test is a commonly used method for diagnosis of allergic diseases (e.g., pollen allergy, food allergy, etc.) in allergy clinics. The results of this test are erythema and wheal provoked on the skin where the test is applied. The sensitivity of the patient against a specific allergen is determined by the physical size of the wheal. The diameter of the wheal is typically measured by a nurse or medical personnel using a ruler. This manual method for measuring wheal size is a cumbersome process and prone to errors and inter-observer variations. The SPT results of a patient can be interpreted differently due to measurement errors and inter-observer variations of different medical personnel, which can easily cause inconsistencies in the diagnosis of the patient.

Though manual...