Healthc Inform Res.  2025 Jan;31(1):57-65. 10.4258/hir.2025.31.1.57.

Deep Learning Model-Based Detection of Anemia from Conjunctiva Images

• Sehar N
• Krishnamoorthi N
• Vinoth Kumar C

• KMID: 2564352
• DOI: http://doi.org/10.4258/hir.2025.31.1.57

Abstract

Objectives
Anemia is characterized by a reduction in red blood cells, leading to insufficient levels of hemoglobin, the molecule responsible for carrying oxygen. The current standard method for diagnosing anemia involves analyzing blood samples, a process that is time-consuming and can cause discomfort to participants. This study offers a comprehensive analysis of non-invasive anemia detection using conjunctiva images processed through various machine learning and deep learning models. The focus is on the palpebral conjunctiva, which is highly vascular and unaffected by melanin content.
Methods
Conjunctiva images from both anemic and non-anemic participants were captured using a smartphone. A total of 764 conjunctiva images were augmented to 4,315 images using the deep convolutional generative adversarial network model to prevent overfitting and enhance model robustness. These processed and augmented images were then utilized to train and test multiple models, including statistical regression, machine learning algorithms, and deep learning frameworks.
Results
The stacking ensemble framework, which includes the models VGG16, ResNet-50, and InceptionV3, achieved a high area under the curve score of 0.97. This score demonstrates the framework’s exceptional capability in detecting anemia through a noninvasive approach.
Conclusions
This study introduces a noninvasive method for detecting anemia using conjunctiva images obtained with a smartphone and processed using advanced deep learning techniques.

Keyword

Anemia, Conjunctiva, Hemoglobins, Erythrocytes, Machine Learning

Figure

• 

  Figure 1 Overall workflow. ROI: region of interest.

• 

  Figure 2 Preprocessing of the acquired images.

• 

  Figure 3 Segmentation output.

• 

  Figure 4 A deep convolutional generative adversarial network (DCGAN) architecture: (A) generator and (B) discriminator. ReLU: rectified linear unit.

• 

  Figure 5 Architecture of the voting ensemble model. ReLU: rectified linear unit.

• 

  Figure 6 Stacking ensemble model architecture. ReLU: rectified linear unit.

• 

  Figure 7 Architecture of the GoogLeNet model.

• 

  Figure 8 Confusion matrix and receiver operating characteristic curve (ROC) for the (A) machine learning models and (B) deep learning models. SVM: support vector machine, KNN: k-nearest neighbor, AUC: area under the curve.

Reference

References

1. Bevilacqua V, Dimauro G, Marino F, Brunetti A, Cassano F, Di Maio A, et al. A novel approach to evaluate blood parameters using computer vision techniques. In : Proceedings of In 2016 IEEE International Symposium on Medical Measurements and Applications (MeMeA); 2016 May 15–18; Benevento, Italy. p. 1–6. https://doi.org/10.1109/MeMeA.2016.7533760.

2. Delgado-Rivera G, Roman-Gonzalez A, Alva-Mantari A, Saldivar-Espinoza B, Zimic M, Barrientos-Porras F, et al. Method for the automatic segmentation of the palpebral conjunctiva using image processing. In : Proceedings of 2018 IEEE International Conference on Automation/XXIII Congress of the Chilean Association of automatic control (ICA-ACCA); 2018 Oct 17–19; Concepcion, Chile. p. 1–4. https://doi.org/10.1109/ICA-ACCA.2018.8609744.

3. Srivastava SK. Adoption of electronic health records: a roadmap for India. Healthc Inform Res. 2016; 22(4):261–9. https://doi.org/10.4258/hir.2016.22.4.261.

4. Chaparro CM, Suchdev PS. Anemia epidemiology, pathophysiology, and etiology in low- and middle-income countries. Ann N Y Acad Sci. 2019; 1450(1):15–31. https://doi.org/10.1111/nyas.14092.

5. Bauskar S, Jain P, Gyanchandani M. A noninvasive computerized technique to detect anemia using images of eye conjunctiva. Pattern Recognit Image Anal. 2019; 29:438–46. https://doi.org/10.1134/S1054661819030027.

6. Fuadah YN, Sa’idah S, Wijayanto I, Patmasari R, Magdalena R. Non invasive anemia detection in pregnant women based on digital image processing and K-nearest neighbor. In : Proceedings of 2020, 3rd International Conference on Biomedical Engineering (IBIOMED); 2020 Oct 6–8; Yogyakarta, Indonesia. p. 60–4. https://doi.org/10.1109/IBIOMED50285.2020.9487605.

7. Wu Q, Chen Y, Meng J. DCGAN-based data augmentation for tomato leaf disease identification. IEEE Access. 2020; 8:98716–28. https://doi.org/10.1109/ACCESS.2020.2997001.

8. Appiahene P, Arthur EJ, Korankye S, Afrifa S, Asare JW, Donkoh ET. Detection of anemia using conjunctiva images: a smartphone application approach. Med Nov Technol Devices. 2023; 18:100237. https://doi.org/10.1016/j.medntd.2023.100237.

9. Purwanti E, Amelia H, Bustomi MA, Yatijan MA, Putri RN. Anemia Detection using convolutional neural network based on palpebral conjunctiva images. In : Proceedings of 2023, 14th International Conference on Information & Communication Technology and System (ICTS); 2023 Oct 4–5; Surabaya, Indonesia. p. 117–22. https://doi.org/10.1109/ICTS58770.2023.10330869.

10. Dimauro G, Baldari L, Caivano D, Colucci G, Girardi F. Automatic segmentation of relevant sections of the conjunctiva for non-invasive anemia detection. In : Proceedings of 2018, 3rd International Conference on Smart and Sustainable Technologies (SpliTech); 2018 Jun 26–29; Split, Croatia. p. 1–5.

11. Kasiviswanathan S, Vijayan TB, John S. Ridge regression algorithm based non-invasive anaemia screening using conjunctiva images. J Ambient Intell Humaniz Comput. 2020. Oct. 22. [Epub]. https://doi.org/10.1007/s12652-020-02618-3.

12. Sevani N, Fredicia , Persulessy GB. Detection anemia based on conjunctiva pallor level using k-means algorithm. IOP Conf Ser Mater Sci Eng. 2018; 420:012101. https://doi.org/10.1088/1757-899X/420/1/012101.

13. Mannino RG, Myers DR, Tyburski EA, Caruso C, Boudreaux J, Leong T, et al. Smartphone app for noninvasive detection of anemia using only patient-sourced photos. Nat Commun. 2018; 9(1):4924. https://doi.org/10.1038/s41467-018-07262-2.

14. Kumar RD, Guruprasad S, Kansara K, Rao KR, Mohan M, Reddy MR, et al. A novel noninvasive hemoglobin sensing device for anemia screening. IEEE Sens J. 2021; 21(13):15318–29. https://doi.org/10.1109/JSEN.2021.3070971.

15. Dimauro G, Griseta ME, Camporeale MG, Clemente F, Guarini A, Maglietta R. An intelligent non-invasive system for automated diagnosis of anemia exploiting a novel dataset. Artif Intell Med. 2023; 136:102477. https://doi.org/10.1016/j.artmed.2022.102477.

16. Dimauro G, Guarini A, Caivano D, Girardi F, Pasciolla C, Iacobazzi A. Detecting clinical signs of anaemia from digital images of the palpebral conjunctiva. IEEE Access. 2019; 7:113488–98. https://doi.org/10.1109/ACCESS.2019.2932274.

17. Epomedicine. Pallor [Internet]. Epomedicine;2013. [cited at 2025 Jan 5]. Available from: https://epomedicine.com/clinical-medicine/clinical-examination-pallor/.

18. Pflipsen M, Massaquoi M, Wolf S. Evaluation of the painful eye. Am Fam Physician. 2016; 93(12):991–8.