An anemia screening tool based on deep learning with conjunctiva images

Xiaoyan HU; Haoyang LI; Xiang LIU; Yujie LI; Lifang TAN; Yongshuai LI; Yuwen CHEN; Bin YI

doi:10.16016/j.2097-0927.202301049

您当前的位置：

首页 >

文章列表页 >

An anemia screening tool based on deep learning with conjunctiva images

Monographic Reports: Governance and artificial intelligence algorithms of big data during perioperative period | 更新时间：2026-01-09

- An anemia screening tool based on deep learning with conjunctiva images
- Journal of Army Medical University Vol. 45, Issue 8, Pages: 746-752(2023)
- 作者机构：
  
  1.400038 重庆, 陆军军医大学(第三军医大学)第一附属医院麻醉科
  2.400038 重庆, 陆军军医大学(第三军医大学)基础医学院学员五大队
  3.400714 重庆, 中国科学院重庆绿色与智能技术研究所
- 作者简介：
- 基金信息：
  
  the National Key Research and Development Program(2018YFC0116700)
- DOI：10.16016/j.2097-0927.202301049
  CLC： R319;R322.91;R556.04
- Received：01 January 2023，
  
  Revised：2023-02-15，
  
  Published：30 April 2023
- 稿件说明：
移动端阅览
Xiaoyan HU, Haoyang LI, Xiang LIU, et al. An anemia screening tool based on deep learning with conjunctiva images[J]. Journal of Army Medical University, 2023, 45(8): 746-752.
DOI：

Xiaoyan HU, Haoyang LI, Xiang LIU, et al. An anemia screening tool based on deep learning with conjunctiva images[J]. Journal of Army Medical University, 2023, 45(8): 746-752. DOI： 10.16016/j.2097-0927.202301049.

摘要

目的

探索深度学习技术在利用睑结膜图像作为输入的贫血自动筛查方面的应用。

方法

前瞻性收集和分析2021年3月18日到4月26日陆军军医大学第一附属医院麻醉科284例择期手术患者的睑结膜图像，依据对应的血红蛋白浓度，将图像标记为正常和贫血两个类别。基于InceptionV3、ResNet50V2、EfficientNetV2B0、DenseNet121等4种深度学习算法构建贫血预测模型。采用受试者特征工作曲线(receiver operating characteristic curve，ROC曲线)、准确性、敏感度、特异性、阳性预测值、阴性预测值评估模型性能。

结果

基于4种深度学习算法，贫血预测风险模型的ROC曲线下面积(area under curve，AUC)分别为：0.709 (95%

：0.643~0.769)、0.661 (95%

：0.594~0.725)、0.670 (95%

：0.603~0.733)、0.695 (95%

：0.628~0.756)。InceptionV3在测试集上的AUC(95%

)、准确度、敏感度、特异性、阳性预测值、阴性预测值分别为：0.709 (95%

：0.643~0.769)、0.695、0.750、0.412、0.707、0.629，其模型性能最优；基于最优算法，开发了一个网络服务应用程序(

http://150.158.58.4

)用于对贫血的在线预测。

结论

本研究基于深度学习算法利用睑结膜图像建立的贫血筛查模型快速高效，InceptionV3模型的综合预测性能更优。

Abstract

Objective

To explore the application of deep learning in automatic classification of anemia with conjunctival images as input.

Methods

The conjunctival images of 284 patients undergoing elective surgery in the Department of Anesthesiology of the First Affiliated Hospital of Army Medical University from March 18 to April 26

2021 were collected and analyzed prospectively. The images divided into 2 types: normal and anemia according to the corresponding hemoglobin concentration. Four deep learning algorithms

including InceptionV3

ResNet50V2

EfficientNetV2B0 and DenseNet121

were used to construct a prediction model for anemia. The performance of the model was evaluated by receiver operating characteristic (ROC) curve with accuracy

sensitivity

specificity

positive predictive value and negative predictive value.

Results

The area under ROC curve (AUC) was 0.709 (95%

: 0.643~0.769)

0.661 (95%

: 0.594~0.725)

0.670 (95%

: 0.603~0.733)

and 0.695 (95%

: 0.628~0.756)

respectively for the 4 deep learning algorithms. The InceptionV3 model showed superior predictive performance on the test set

with an AUC value of 0.709 (95%

: 0.643~0.769)

an accuracy of 0.695

a sensitivity of 0.750

a specificity of 0.412

a positive predictive value of 0.707 and a negative predictive value of 0.629. Based on the optimal algorithm

a network service application which can be used for online prediction of anemia was developed (

http://150.158.58.4

Conclusion

Our model

which is established based on deep learning algorithm with conjunctiva image as input

has a good performance on fast and automatic prediction for anemia. The InceptionV3model has better comprehensive prediction performance.

关键词

Keywords

references

WHO. Worldwide prevalence of anaemia 1993-2005: WHO Global Database on Anaemia[EB/OL ] . [2023-04-19 ] . https://www.who.int/publications-detail-redirect/9789241596657 https://www.who.int/publications-detail-redirect/9789241596657 .

STEVENS G A, FINUCANE M M, DE-REGIL L M, et al . Global, regional, and national trends in haemoglobin concentration and prevalence of total and severe anaemia in children and pregnant and non-pregnant women for 1995-2011: a systematic analysis of population-representative data[J ] . Lancet Glob Health, 2013, 1(1): e16-e25. DOI: 10.1016/S2214-109X(13)70001-9.

CIRELI E, MERTOGLU A. The impact of anemia on the mortality of COPD patients hospitalized for acute exacerbation resulting in respiratory failure[J]. Monaldi Arch Chest Dis, 2022. 2022, 93(2). DOI:10.4081/monaldi.2022.2254.

JONAITIENE N, RAMANTAUSKAITE G, LAUKAITIENE J. Anaemia in heart failure patients, associated with angiotensin-renin-aldosterone system altering medications[J]. Heart Views, 2021, 22(3): 196-200. DOI: 10.4103/HEARTVIEWS.HEARTVIEWS_211_20.

KE Y H, HWANG K Y, THIN T N, et al . The usefulness of non-invasive co-oximetry haemoglobin measurement for screening pre-operative anaemia[J ] . Anaesthesia, 2021, 76(1): 54-60. DOI: 10.1111/anae.15171.

颜娅, 葛衡江. 无创连续血红蛋白监测在手术中的应用[J]. 第三军医大学学报, 2012, 34(23): 2425-2427. DOI: 10.16016/j.1000-5404.2012.23.009.

YAN Y, GE H J. Value of continuous noninvasive hemoglobin monitoring during surgery[J]. J Third Mil Med Univ, 2012, 34(23): 2425-2427. DOI: 10.16016/j.1000-5404.2012.23.009.

GALVAGNO S M Jr, HU P, YANG S, et al . Accuracy of continuous noninvasive hemoglobin monitoring for the prediction of blood transfusions in trauma patients[J ] . J ClinMonit Comput, 2015, 29(6): 815-821. DOI: 10.1007/s10877-015-9671-1.

SARKAR P K, PAL S, POLLEY N, et al . Development and validation of a noncontact spectroscopic device for hemoglobin estimation at point-of-care[J ] . J Biomed Opt, 2017, 22(5): 55006. DOI: 10.1117/1.JBO.22.5.055006

SPINELLI M G, SOUZA J M, SOUZA S B, et al . Reliability and validity of palmar and conjunctival pallor for anemia detection purposes[J ] . Rev Saúde Pública, 2003, 37(4): 404-408. DOI: 10.1590/s0034-89102003000400003.

KALANTRI A, KARAMBELKAR M, JOSHI R, et al . Accuracy and reliability of pallor for detecting anaemia: a hospital-based diagnostic accuracy study[J ] . PloS One, 2010; 5(1): e8545. DOI: 10.1371/journal.pone.0008545

TANG B, YU X, XU L, et al . Continuous noninvasive hemoglobin monitoring estimates timing for detecting anemia better than clinicians: a randomized controlled trial[J ] . BMC anesthesiology, 2019;19(1): 80. DOI: 10.1186/s12871-019-0755-1

JIN L, YANG J, KUANG K, et al . Deep-learning-assisted detection and segmentation of rib fractures from CT scans: Development and validation of FracNet[J ] . EBio Medicine, 2020, 62: 103106. DOI: 10.1016/j.ebiom.2020.103106.

MOJAHED D, HA R S, CHANG P, et al . Fully Automated Postlumpectomy Breast Margin Assessment Utilizing Convolutional Neural Network Based Optical Coherence Tomography Image Classification Method[J ] . Acad Radiol, 2020, 27(5): e81-e86. DOI: 10.1016/j.acra.2019.06.018.

CHEN Y M, MIAOU S G, BIAN H. Examining palpebral conjunctiva for anemia assessment with image processing methods[J]. Computer Methods Programs Biomed, 2016, 137: 125-135. DOI: 10.1016/j.cmpb.2016.08.025

BEVILACQUA V, DIMAURO G, MARINO F, et al . A novel approach to evaluate blood parameters using computer vision techniques[C ] //2016 IEEE International symposium on medical measurements and applications(MeMeA). Benevento, Italy: IEEE, 2016: 1-6. DOI: 10.1109/MeMeA.2016.7533760 http://dx.doi.org/10.1109/MeMeA.2016.7533760 ..

TAMIR A, JAHAN C S, SAIF M S, et al . Detection of anemia from image of the anterior conjunctiva of the eye by image processing and thresholding[C ] //IEEE region 10 humanitarian technology conference (R10-HTC). Dhaka, Bangladesh: IEEE, 2017: 697-701. DOI: 10.1109/R10-HTC.2017.8289053 http://dx.doi.org/10.1109/R10-HTC.2017.8289053 ..

R MUTHALAGU, BAI V T, JOHN S. A Smart (phone) solution: An effective tool for screening anaemia-correlation with conjunctiva pallor and haemoglobin levels[J]. TAGA journal, 2018, 14: 2611-2621.

COLLINGS S, THOMPSON O, HIRST E, et al . Non-invasive detection of anaemia using digital photographs of the conjunctiva[J ] . PloS One, 2016, 11(4): e0153286. DOI: 10.1371/journal.pone.0153286

JAIN P, BAUSKAR S, GYANCHANDANI M. Neural network based non-invasive method to detect anemia from images of eye conjunctiva[J]. Int J Imaging Syst Technol, 2019, 30(1): 112-125. DOI: 10.1002/ima.22359.

SELIM, SUNER, MD M, et al . Non-invasive determination of hemoglobin by digital photography of palpebral conjunctiva[J ] . J Emerg Med, 2007, 33(2): 105-111. DOI: 10.1016/j.jemermed.2007.02.011.

DIMAURO G, CAIVANO D, GIRARDI F. A new method and a non-invasive device to estimate Anemia based on digital images of the conjunctiva[J]. IEEE Access, 2018, 6: 46968-46975. DOI: 10.1109/ACCESS.2018.2867110.

SZEGEDY C, VANHOUCKE V, IOFFE S, et al . Rethinking the inception architecture for computer vision[C ] //2016 IEEE conference on computer vision and pattern recognition (CVPR). Las Vegas, USA: IEEE, 2016: 2818-2826. DOI: 10.1109/CVPR.2016.308 http://dx.doi.org/10.1109/CVPR.2016.308 ..

HE K M, ZHANG X Y, REN S Q, et al . Identity mappings in deep residual networks[M ] //Computer Vision-ECCV 2016. Cham: Springer International Publishing, 2016: 630-645. DOI: 10.1007/978-3-319-46493-0_38 http://dx.doi.org/10.1007/978-3-319-46493-0_38 ..

TAN M X, LE Q V. EfficientNetV2: Smaller models and faster training[EB/OL ] . ArXiv, abs/2104.00298. [2023-04-19 ] . https://arxiv.org/abs/2104.00298 https://arxiv.org/abs/2104.00298 .

HUANG G, LIU Z, VAN DER MAATEN L, et al . Densely connected convolutional networks[C ] //2017 IEEE conference on computer vision and pattern recognition (CVPR). Honolulu, USA: IEEE, 2017: 2261-2269. DOI: 10.1109/CVPR.2017.243 http://dx.doi.org/10.1109/CVPR.2017.243 ..

ROSENCHER N, KERKKAMP H E, MACHERAS G, et al . Orthopedic surgery transfusion hemoglobin european overview (OSTHEO) study: blood management in elective knee and hip arthroplasty in Europe[J ] . Transfusion, 2003, 43(4): 459-469. DOI: 10.1046/j.1537-2995.2003.00348.x.

KARKOUTI K, WIJEYSUNDERA D N, BEATTIE W S, et al . Risk associated with preoperative anemia in cardiac surgery: a multicenter cohort study[ J ] . Circulation, 2008, 117(4): 478-484. DOI: 10.1161/CIRCULATIONAHA.107.718353.

NAKAMORI E, SHIGEMATSU K, HIGASHI M, et al . Postoperative noninvasive hemoglobin monitoring is useful to prevent unnoticed postoperative anemia and inappropriate blood transfusion in patients undergoing total hip or knee arthroplasty: a randomized controlled trial[J ] . Geriatr Orthop Surg Rehabil, 2021, 12: 21514593211060575. DOI: 10.1177/21514593211060575.

ANGGRAENI M D, FATONI A. Non-invasive self-care anemia detection during pregnancy using a smartphone camera[C ] //IOP conference series: materials science and engineering. Purwokerto, Indonesia: IOP Publishing, 2017, 172: 012030. DOI: 10.1088/1757-899x/172/1/012030 http://dx.doi.org/10.1088/1757-899x/172/1/012030 .

CHEN Y M, MIAOU S G. A kalman filtering and nonlinear penalty regression approach for noninvasive anemia detection with palpebral conjunctiva images[J]. J Healthc Eng, 2017, 2017: 9580385. DOI: 10.1155/2017/9580385.

DIMAURO G, GUARINI A, CAIVANO D, et al . Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva[J ] . IEEE Access, 2019, 7: 113488-113498. DOI: 10.1109/ACCESS.2019.2932274.

BIN NOOR N, ANWAR M S, DEY M. Comparative study between decision tree, SVM and KNN to predict anaemic condition[C ] //2019 IEEE International conference on biomedical engineering, computer and information technology for health (BECITHCON). Dhaka, Bangladesh: IEEE, 2020: 24-28. DOI: 10.1109/BECITHCON48839.2019.9063188 http://dx.doi.org/10.1109/BECITHCON48839.2019.9063188 .

SALDIVAR-ESPINOZA B, NÚÑEZ-FERNÁNDEZ D, PORRAS-BARRIENTOS F, et al . Portable system for the prediction of anemia based on the ocular conjunctiva using artificial intelligence[EB/OL ] . 2019: arXiv: 1910.12399. https://arxiv.org/abs/1910.12399 https://arxiv.org/abs/1910.12399 ..

KASIVISWANATHAN S, BAI VIJAYAN T, SIMONE L, et al . Semantic segmentation of conjunctiva region for non-invasive anemia detection applications[J ] . Electronics, 2020, 9(8): 1309. DOI: 10.3390/electronics9081309.

李年华, 高亮, 李彬, 等. 海拔5 000 m以上男性健康习服青年血红蛋白和血氧饱和度参考值研究[J]. 国际检验医学杂志, 2018, 39(2): 204-206. DOI: 10.3969/j.issn.1673-4130.2018.02.025.

LI N H, GAO L, LI B, et al . Study on reference value of hemoglobin and blood oxygen saturation of male healthy acclimatized youth above 5000 m above sea level[J ] . Int J Lab Med, 2018, 39(2): 204-206. DOI: 10.3969/j.issn.1673-4130.2018.02.025.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Integrative model combining deep learning, clinical and radiomic features enhances EGFR mutation prediction in non-small cell lung cancer

Predictive performance of CT images-based 3D ResNet18 model for identifying lung tuberculosis drug resistance

Advances in deep learning and MRI co-assisted genotyping of brain gliomas

Ultrasound thyroid nodule segmentation algorithm based on wavelet transform and CNN-Transformer

Automatic segmentation of female urine control anatomical elements and related structures in MRI images based on deep learning

Related Author

YANG Taotao

CHEN Wei

CHEN Cancan

YAN Wanying

WANG Dawei

XIONG Kunlin

SUN Zhiyuan

WANG Xianqi

Related Institution

Department of Diagnostic Radiology， Eastern Theatre General Hospital

Department of Radiology， Daping Hospital， Army Medical University

Department of Radiology，First Affiliated Hospital， Army Medical University （Third Military Medical University）

Infervision Medical Technology Co.， Ltd

Department of Medical Imaging， Chongqing Public Health Medical Center

⁰