Sclera Segmentation in Images for Bilirubin Level Measurement Using the U-Net Network

Authors

  • Diana A. Mendoza-Mora Universidad Autónoma del Estado de México
  • Adriana H. Vilchis-González Universidad Autónoma del Estado de México
  • Rigoberto Martínez-Méndez Universidad Autónoma del Estado de México
  • Vianney Muñoz-Jiménez Universidad Autónoma del Estado de México
  • Iván Francisco-Valencia Universidad Autónoma del Estado de México

DOI:

https://doi.org/10.47756/aihc.y9i1.164

Keywords:

Segmentation, Sclera, U-Net, RGB imaging, Bilirubin index

Abstract

The sclera is a white membrane rich in collagen and elastin fibres, which gives it an affinity for bilirubin. This yellow substance is produced by the breakdown of the heme group, and when its levels are elevated, it causes jaundice a condition that leads to yellowing of the skin, mucous membranes, and sclera. The intensity of the yellowing in the sclera is directly related to bilirubin levels in the body. This relationship enables the extraction of features to infer these levels using machine learning techniques based on RGB images of the sclera. Sclera segmentation is a transcendent factor in achieving this goal. For this reason, this article presents the results of sclera segmentation using the U-Net network. This is a convolutional network composed of encoding and decoding layers and is used for medical image segmentation.  The model was trained and validated with a set of 181 eye images and their corresponding binary masks. The results obtained during the training phase are Loss (0.006), Precision (0.976), Recall (0.973) and F1-score (0.974), and in the validation phase: Loss (0.145), Precision (0.897), Recall (0.863) and F1-Score (0.880). These results demonstrate the U-Net model's effectiveness in segmenting the sclera, particularly in the training phase where the metrics are highly favorable. However, the slight decrease in performance during the validation phase suggests the need for further refinement. Future work will focus on increasing the dataset size and introducing data augmentation techniques to improve generalization and robustness. Ultimately, accurate sclera segmentation is a critical step toward developing reliable Machine Learning models for non-invasive bilirubin level estimation.

Downloads

Download data is not yet available.

References

N. C. Flores-García, M. Dirac, H. Han, y D. Kershenobich-Stalnikowitz, «La carga de la enfermedad por cirrosis hepática en México», GMM, vol. 159, n.o 6, p. 12605, dic. 2023, doi: 10.24875/GMM.23000370. DOI: https://doi.org/10.24875/GMM.23000370

A. R. Guerra-Ruiz et al., «Bilirrubina: Medición y utilidad clínica en la enfermedad hepática», Advances in Laboratory Medicine / Avances en Medicina de Laboratorio, vol. 2, n.o 3, pp. 362-372, ago. 2021, doi: 10.1515/almed-2021-0016. DOI: https://doi.org/10.1515/almed-2021-0016

L. Ngashangva, V. Bachu, y P. Goswami, «Development of new methods for determination of bilirubin», Journal of Pharmaceutical and Biomedical Analysis, vol. 162, pp. 272-285, ene. 2019, doi: 10.1016/j.jpba.2018.09.034. DOI: https://doi.org/10.1016/j.jpba.2018.09.034

G. Alfieri, R. Mir Villamayor, L. E. Genes De Lovera, E. M. Otazo Arévalos, S. G. Miño Moreno, y J. P. G. Bordón Dure, «Aplicación del bilirrubinómetro no invasivo en recién nacidos», Pediatr (Asunción)., vol. 46, n.o 3, pp. 158-164, nov. 2019, doi: 10.31698/ped.46032019002. DOI: https://doi.org/10.31698/ped.46032019002

C. Enweronu-Laryea et al., «Validating a Sclera-Based Smartphone Application for Screening Jaundiced Newborns in Ghana», Pediatrics, vol. 150, n.o 1, p. e2021053600, jun. 2022, doi: 10.1542/peds.2021-053600. DOI: https://doi.org/10.1542/peds.2021-053600

T. Kihara et al., «Identification and Quantification of Jaundice by Trans-Conjunctiva Optical Imaging Using a Human Brain-like Algorithm: A Cross-Sectional Study», Diagnostics, vol. 13, n.o 10, p. 1767, may 2023, doi: 10.3390/diagnostics13101767. DOI: https://doi.org/10.3390/diagnostics13101767

Md. M. M. Miah et al., «Non-Invasive Bilirubin Level Quantification and Jaundice Detection by Sclera Image Processing», en 2019 IEEE Global Humanitarian Technology Conference (GHTC), oct. 2019, pp. 1-7. doi: 10.1109/GHTC46095.2019.9033059. DOI: https://doi.org/10.1109/GHTC46095.2019.9033059

A. Mariakakis, M. A. Banks, L. Phillipi, L. Yu, J. Taylor, y S. N. Patel, «BiliScreen: Smartphone-Based Scleral Jaundice Monitoring for Liver and Pancreatic Disorders», Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., vol. 1, n.o 2, pp. 1-26, jun. 2017, doi: 10.1145/3090085. DOI: https://doi.org/10.1145/3090085

F. Outlaw, M. Nixon, O. Odeyemi, L. W. MacDonald, J. Meek, y T. S. Leung, «Smartphone screening for neonatal jaundice via ambient-subtracted sclera chromaticity», PLoS ONE, vol. 15, n.o 3, p. e0216970, mar. 2020, doi: 10.1371/journal.pone.0216970. DOI: https://doi.org/10.1371/journal.pone.0216970

Md. R. Sammir, K. Md. Towhidul Alam, P. Saha, Md. M. Rahaman, y Q. Delwar Hossain, «Design and Implementation of a Non-invasive Jaundice Detection and Total Serum Bilirubin Measurement System», en 2018 10th International Conference on Electrical and Computer Engineering (ICECE), Dhaka, Bangladesh: IEEE, dic. 2018, pp. 137-140. doi: 10.1109/ICECE.2018.8636801. DOI: https://doi.org/10.1109/ICECE.2018.8636801

A. Das et al., «Sclera Segmentation Benchmarking Competition in Cross-resolution Environment», en 2019 International Conference on Biometrics (ICB), Crete, Greece: IEEE, jun. 2019, pp. 1-7. doi: 10.1109/ICB45273.2019.8987414. DOI: https://doi.org/10.1109/ICB45273.2019.8987414

A. Das et al., «Sclera Segmentation and Joint Recognition Benchmarking Competition: SSRBC 2023», en 2023 IEEE International Joint Conference on Biometrics (IJCB), Ljubljana, Slovenia: IEEE, sep. 2023, pp. 1-10. doi: 10.1109/IJCB57857.2023.10448601. DOI: https://doi.org/10.1109/IJCB57857.2023.10448601

R. A. Naqvi y W.-K. Loh, «Sclera-Net: Accurate Sclera Segmentation in Various Sensor Images Based on Residual Encoder and Decoder Network», IEEE Access, vol. 7, pp. 98208-98227, 2019, doi: 10.1109/ACCESS.2019.2930593. DOI: https://doi.org/10.1109/ACCESS.2019.2930593

C. Wang, Y. He, Y. Liu, Z. He, R. He, y Z. Sun, «ScleraSegNet: an Improved U-Net Model with Attention for Accurate Sclera Segmentation», en 2019 International Conference on Biometrics (ICB), Crete, Greece: IEEE, jun. 2019, pp. 1-8. doi: 10.1109/ICB45273.2019.8987270. DOI: https://doi.org/10.1109/ICB45273.2019.8987270

Arzate Cruz Karla Yaneli, «Desarrollo de un dispositivo electrónico para medir la coloración de la esclera y su relación con el nivel de bilirrubina en sangre», Universidad Autónoma del Estado de México, Toluca, Estado de México, 2024.

M. Ríos-Hernández, J. M. Jacinto-Villegas, O. Portillo-Rodríguez, y A. H. Vilchis-González, «User-Centered Design and Evaluation of an Upper Limb Rehabilitation System with a Virtual Environment», Applied Sciences, vol. 11, n.o 20, p. 9500, oct. 2021, doi: 10.3390/app11209500. DOI: https://doi.org/10.3390/app11209500

N. Ibtehaz y M. S. Rahman, «MultiResUNet : Rethinking the U-Net Architecture for Multimodal Biomedical Image Segmentation», Neural Networks, vol. 121, pp. 74-87, ene. 2020, doi: 10.1016/j.neunet.2019.08.025. DOI: https://doi.org/10.1016/j.neunet.2019.08.025

Downloads

Published

2024-11-30

How to Cite

[1]
Mendoza-Mora, D.A. et al. 2024. Sclera Segmentation in Images for Bilirubin Level Measurement Using the U-Net Network. Avances en Interacción Humano-Computadora. 9, 1 (Nov. 2024), 179–184. DOI:https://doi.org/10.47756/aihc.y9i1.164.

Issue

Section

Work in Progress