A Study on Impacts of Ambient Lights and Skin Tone for Contactless Blood Oxygen Measurement using Visible Spectrum Facial Videosaper

R. Arulselvi, V. Jeyakumar, N. G. Vidya, V. Yuvaraj
{"title":"A Study on Impacts of Ambient Lights and Skin Tone for Contactless Blood Oxygen Measurement using Visible Spectrum Facial Videosaper","authors":"R. Arulselvi, V. Jeyakumar, N. G. Vidya, V. Yuvaraj","doi":"10.1109/DELCON57910.2023.10127508","DOIUrl":null,"url":null,"abstract":"Blood oxygen saturation also termed SpO2, is a measure of oxygen-carrying capacity of the hemoglobin in the blood as a percentage. Normal arterial blood oxygen saturation level among humans varies from 97 to 100 percent (+/- 5%). SpO2 less than 90% is low, so-called Hypoxemia. Today, COVID patients need frequent and precise SpO2 assessments. Current pulse oximeters demand a contact-based measurement setup employing clips or sensors, which raises the possibility of skin irritation and infection. Additionally, depending on the patient’s age and finger size, pulse oximeters don’t always fit comfortably on everyone’s fingertips, which might result in misleading findings. Other restrictions of the current devices include errors brought on by interference, venous pulsations, skin problems, insufficient perfusion, light interference, and the presence of nail polish. This work offers a successful technique for non-contact SpO2 measurement from facial recordings. The videos show 31 participants doing three activities in various lighting situations. Each activity must be recorded with a camera mounted to a tripod. The obtained films go through pre-processing procedures include converting video to frames, discarding the background, finding faces, choosing regions of interest (RoI), and isolating the RGB channels. The SpO2 level is computed using the red and blue channels, which correspond to the red and infrared light used in contact pulse oximeters. In order to determine the pulsatile absorbance rate, the AC and DC components of the Red (R) and Blue (B) channels are separated. SpO2 will be predicted from the R-R ratio and its related parameter using Linear regression. Root mean squared error (RMSE) is the error between estimated and actual SpO2’s that is calculated for various skin tone participants. From the experiment, it is inferred that medium lighting with a lux range of 200-500 is a minimum requirement to estimate SpO2 which offers low RMSE for all skin types.","PeriodicalId":193577,"journal":{"name":"2023 2nd Edition of IEEE Delhi Section Flagship Conference (DELCON)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2023-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2023 2nd Edition of IEEE Delhi Section Flagship Conference (DELCON)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/DELCON57910.2023.10127508","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

Blood oxygen saturation also termed SpO2, is a measure of oxygen-carrying capacity of the hemoglobin in the blood as a percentage. Normal arterial blood oxygen saturation level among humans varies from 97 to 100 percent (+/- 5%). SpO2 less than 90% is low, so-called Hypoxemia. Today, COVID patients need frequent and precise SpO2 assessments. Current pulse oximeters demand a contact-based measurement setup employing clips or sensors, which raises the possibility of skin irritation and infection. Additionally, depending on the patient’s age and finger size, pulse oximeters don’t always fit comfortably on everyone’s fingertips, which might result in misleading findings. Other restrictions of the current devices include errors brought on by interference, venous pulsations, skin problems, insufficient perfusion, light interference, and the presence of nail polish. This work offers a successful technique for non-contact SpO2 measurement from facial recordings. The videos show 31 participants doing three activities in various lighting situations. Each activity must be recorded with a camera mounted to a tripod. The obtained films go through pre-processing procedures include converting video to frames, discarding the background, finding faces, choosing regions of interest (RoI), and isolating the RGB channels. The SpO2 level is computed using the red and blue channels, which correspond to the red and infrared light used in contact pulse oximeters. In order to determine the pulsatile absorbance rate, the AC and DC components of the Red (R) and Blue (B) channels are separated. SpO2 will be predicted from the R-R ratio and its related parameter using Linear regression. Root mean squared error (RMSE) is the error between estimated and actual SpO2’s that is calculated for various skin tone participants. From the experiment, it is inferred that medium lighting with a lux range of 200-500 is a minimum requirement to estimate SpO2 which offers low RMSE for all skin types.
环境光和肤色对可见光谱面部测氧仪非接触式血氧测量的影响研究
血氧饱和度也称为SpO2,是衡量血液中血红蛋白携氧能力的百分比。人的正常动脉血氧饱和度在97%到100%(+/- 5%)之间变化。SpO2低于90%即为低氧血症。今天,COVID患者需要频繁和精确的SpO2评估。目前的脉搏血氧仪需要使用夹子或传感器的接触式测量装置,这增加了皮肤刺激和感染的可能性。此外,根据患者的年龄和手指大小,脉搏血氧仪并不总是适合每个人的指尖,这可能会导致误导性的发现。当前设备的其他限制包括干扰、静脉搏动、皮肤问题、灌注不足、光干扰和指甲油的存在带来的误差。这项工作为面部记录的非接触式SpO2测量提供了成功的技术。视频显示,31名参与者在不同的照明条件下做三件事。每次活动都必须用安装在三脚架上的摄像机记录下来。获得的影片经过预处理程序,包括将视频转换为帧、丢弃背景、寻找人脸、选择感兴趣区域(RoI)和隔离RGB通道。使用红色和蓝色通道计算SpO2水平,这对应于接触式脉搏血氧计中使用的红光和红外光。为了确定脉冲吸光度,将红(R)和蓝(B)通道的交流和直流分量分开。利用线性回归方法从R-R比及其相关参数预测SpO2。均方根误差(RMSE)是为不同肤色参与者计算的估计和实际SpO2之间的误差。从实验中可以推断,lux范围为200-500的中等照明是估计SpO2的最低要求,它为所有皮肤类型提供了较低的RMSE。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信