Using camera and 3D scanner to measure upper arm lymphoedema volume

碩士 === 國立高雄科技大學 === 資訊管理系 === 107 === The great majority of patients with upper extremity lymphedema are the women with breast cancer. Breast cancer related lymphedema typically affects the ipsilateral upper extremity. Breast cancer surgery is also the most common cause of secondary lymphedema in th...

Full description

Bibliographic Details
Main Authors: CHEN-YUNG CHUN, 陳詠淳
Other Authors: WEN- CHEN HUANG
Format: Others
Language:zh-TW
Published: 2019
Online Access:http://ndltd.ncl.edu.tw/handle/9rte57
Description
Summary:碩士 === 國立高雄科技大學 === 資訊管理系 === 107 === The great majority of patients with upper extremity lymphedema are the women with breast cancer. Breast cancer related lymphedema typically affects the ipsilateral upper extremity. Breast cancer surgery is also the most common cause of secondary lymphedema in the upper extremity. Early-stage diagnosis and proper treatment are necessary for these patients; otherwise, it may cause irreversible damage. Such early-stage lymphedema is considered reversible with treatment because the skin and tissues haven't been permanently damaged. Although a clinical measure device of lymphedema is already seen. A reliable, valid, real-time monitoring and easy-operating measurement of lymphedema is still expected. In this study, we hope to develop a lightweight device with above features that patients with upper extremity lymphedema can use and can monitor the volume of the lymphedema on their own whenever they want to. These digitalized data can provide a better follow-up for medical use. The study can be divided into two part. First, we choose 3D scanner as a tool and calculate the scanned model volume. “iSense” is the scanner selected in consideration of the light-weight feature. While the devices can operate more efficiently on triangles that are groups into meshes, this is how we calculated a tetrahedral volume mesh. Each tetrahedron has 4 triangular faces, we sum the tetrahedrons to get the volume; and for the repeating issue, we solve it with “signed.” In the second part of the study, we replaced the 3D scanner with a mobile phone camera. We use the camera to capture images of the models, then developed a volume predicting method from calculating the image pixels. We used the python program platform supported with open source “Opencv.” It provides many commonly used functions for image recognition. After taking the pictures, we converted them into grayscale images. And, the filter function allows us to filter a range of data based on the criteria we had defined while the edge calculation helps us to find the target's location. Based on all calculus algorithm, the volume of the target is available.