Blood Group and Diabetes Detection from Fingerprints Using CNN

Abstract

The increased demand for speed and accuracy in medical diagnosis has motivated the use of artificial intelligence techniques in the healthcare domain. There is an urgent need to take necessary steps for developing non-invasive and reliable systems to assist in diagnosis. The main aim of this work is to identify and implement modern artificial intelligence methods to speed up and increase the accuracy of blood group detection by using fingerprint images. In order to achieve higher accuracy, a deep learning-based approach of Convolutional Neural Networks (CNNs) has been incorporated and tested in this work. Several fingerprint parameters like ridge flow, ridge density, and minutiae points are considered in this work. In this project, a fingerprint-based dataset comprising different samples of blood groups has been used, as the accurate detection of blood groups plays a very important role in different medical emergencies and healthcare applications. The added advantage of the system is its non-invasive and time-saving nature, making it an effective alternative to traditional blood testing procedures. A module has been designed that predicts an individual\'s blood group by analysing their fingerprint image. This module focuses more on predictive analytics than invasive diagnostics, with a greater emphasis on automation and accessibility in the healthcare domain. The designed architecture will contribute to other AI-based biometric diagnosis systems, and it provides a concrete base for further medical image analysis and artificial intelligence applications. In addition to blood group identification, the system also includes a module for detecting the presence of diabetes based on fingerprint image features, making it useful for early health screening.

Introduction

The healthcare sector is increasingly adopting non-invasive, rapid, and automated diagnostic technologies to overcome the limitations of traditional laboratory-based methods. Blood group determination is critical for transfusions, surgery, trauma care, and organ transplantation, yet conventional methods can be time-consuming and resource-intensive. Fingerprints, as unique and universally available biometric traits, have emerged as a promising source for predicting blood groups, potentially enabling fast, accessible, and non-invasive diagnostics.

Deep learning, particularly Convolutional Neural Networks (CNNs), excels at analyzing visual patterns and extracting features directly from images. This study develops a CNN-based approach to predict blood groups from fingerprint images, aiming to create a time-saving and reliable screening tool. The methodology involves dataset collection across all major ABO and Rh blood groups, image preprocessing (grayscale conversion, resizing, normalization, and augmentation), CNN model design with convolutional, pooling, and fully connected layers, and model training and evaluation using metrics such as accuracy, precision, recall, and F1-score.

The model achieved 95% accuracy during training and 82% accuracy on validation data, demonstrating effective learning of fingerprint ridge patterns, minutiae, and density features relevant to blood group classification. Misclassifications were mostly due to poor image quality or limited samples for certain groups. The results indicate that fingerprint-based CNN models can serve as a rapid, non-invasive, and reliable diagnostic support system, with potential applications in emergency care, routine screening, and integration into broader biometric health monitoring frameworks.

The study highlights the broader trend of combining AI, imaging technologies, and biometric analysis to develop efficient, accessible, and non-contact medical diagnostic tools, with potential extensions to disease detection such as diabetes and cardiovascular risk assessment.

Conclusion

This project explored the idea of finding a person’s blood group by using only their fingerprint image. A Convolutional Neural Network was designed and trained on a fingerprint dataset, and it was able to correctly identify blood groups with 95% accuracy during training and 82% accuracy during testing. These results show that fingerprints contain useful patterns such as ridge flow, density, and minutiae points, which can help in blood group prediction. The main strength of this system is that it is non-invasive, quick, and does not require a blood sample, making it helpful in emergencies, rural areas, and places with limited medical facilities. It also reduces human effort and can be automated for large-scale use. However, the accuracy can still be improved. A larger dataset, better quality fingerprint images, and more diverse samples from different age groups and regions may lead to stronger results. Future work can also focus on developing a real-time application that can be used in hospitals, blood banks, and ambulances. Overall, this study shows that fingerprint-based blood group detection using deep learning is a practical and promising step toward faster and more accessible healthcare support, and it provides a concrete base for other AI-based biometric diagnosis systems, including diabetes prediction.

References

[1] N. Sinha and A. G. Ramakrishnan, “Automation of differential blood count,” in Proc. Conf. Convergent Technol. Asia–Pacific Region (TENCON), Bengaluru, India, Vol. 2, 2003, pp. 547–551. [2] A. T. Anifowoshe, O. A. Owolodun, K. M. Akinseye, O. A. Iyiola, and B. F. Oyeyemi, “Gene frequencies of ABO and Rh blood groups in Nigeria: A review,” Egyptian J. Med. Hum. Genet., Vol. 18, No. 3, Jul. 2017, pp. 205–210, doi: 10.1016/j.ejmhg.2016.10.004. [3] S. Ba?, G. Carello, E. Lanzarone, and S. Yalçindag?, “An appointment scheduling framework to balance the production of blood units from donation,” Eur. J. Oper. Res., Vol. 265, No. 3, Mar. 2018, pp. 1124–1143, doi: 10.1016/j.ejor.2017.08.054. [4] L. Gustafsson, D. Nordling, M. Andersson, and B. Sheldon, “Infectious diseases, reproductive effort and the cost of reproduction in birds,” Phil. Trans. Roy. Soc. London. Ser. B, Biol. Sci., Vol. 346, No. 1317, 1994, pp. 323–331. [5] J. Wu, J. Qiu, W. Jiang, J. Qiu, L. Zhang, R. Zhao, and C. Yu, “Development and validation of a nomogram predicting the probability of type A aortic dissection at a diameter below 55 mm: A retrospective cohort study,” Int. J. Surgery, Vol. 60, Dec. 2018, pp. 266–272. [6] T. Kallenbach, “Anaesthesia for a patient with beta thalassaemia major,” Southern African Journal of Anaesthesia and Analgesia, Vol. 21, No. 5, 2015, pp. 21–24. [7] T. T. A. Dinh, J. Wang, G. Chen, R. Liu, B. C. Ooi, and K.-L. Tan, “BLOCKBENCH: A framework for analyzing private blockchains,” Proceedings of the ACM International Conference on Management of Data, May 2017, pp. 1085–1100. [8] N. M. Karabacak, P. S. Spuhler, F. Fachin, E. J. Lim, V. Pai, E. Ozkumur, J. M. Martel, N. Kojic, K. Smith, P. I. Chen, J. Yang, H. Hwang, B. Morgan, J. Trautwein, T. A. Barber, S. L. Stott, S. Maheswaran, R. Kapur, D. A. Haber, and M. Toner, “Microfluidic, marker-free isolation of circulating tumor cells from blood samples,” Nature Protocols, Vol. 9, No. 3, Feb. 2014, pp. 694–710, doi: 10.1038/nprot.2014.044. [9] S. Cadd, B. Li, P. Beveridge, W. T. O’Hare, A. Campbell, and M. Islam, “The non-contact detection and identification of blood-stained fingerprints using visible wavelength reflectance hyperspectral imaging: Part 1,” Science & Justice, Vol. 56, No. 3, May 2016, pp. 181–190, doi: 10.1016/j.scijus.2016.01.004. [10] Y. Seo, G. Kim, J. Lim, A. Lee, B. Kim, J. Jang, C. Mo, and M. S. Kim, “Non-destructive detection pilot study of vegetable organic residues using VNIR hyperspectral imaging and deep learning techniques,” Sensors, Vol. 21, No. 9, Apr. 2021, p. 2899, doi: 10.3390/s21092899. [11] M. Zhu, D. Huang, X. Hu, W. Tong, B. Han, J. Tian, and H. Luo, “Application of hyperspectral technology in detection of agricultural products and food: A review,” Food Science & Nutrition, Vol. 8, No. 10, Oct. 2020, pp. 5206–5214, doi: 10.1002/fsn3.1852. [12] S. Wang, et al., “Analysis of anti-M antibody status and blood transfusion strategy in Hunan, China,” Annals of Translational Medicine, Vol. 10, 2022, p. 1166. [13] T. Ackfeld, T. Schmutz, Y. Guechi, and C. Le Terrier, “Blood transfusion reactions: A comprehensive review of the literature including a Swiss perspective,” Journal of Clinical Medicine, Vol. 11, 2022, p. 2859. [14] M. de Haas, F. F. Turik, J. M. Koelewijn, and C. E. van der Schoot, “Haemolytic disease of the fetus and newborn,” Vox Sanguinis, Vol. 109, 2015, pp. 99–113. [15] T. Falk, et al., “U-Net: Deep learning for cell counting, detection, and morphometry,” Nature Methods, Vol. 16, No. 1, 2019, pp. 67–70. [16] L. W. Chen, Y. Zhou, R. Zhou, and M. H. Hong, “Microsphere—Toward the future of optical microscopes,” iScience, Vol. 23, No. 6, 2020, Article no. 101211. [17] S. M. Manohar, P. C. Shah, and A. Nair, “Flow cytometry: Principles, applications and recent advances,” Bioanalysis, Vol. 13, No. 3, 2021, pp. 181–198. [18] W. P. Geng, et al., “An ultra-compact acoustofluidic device based on the narrow-path travelling surface acoustic wave (np-TSAW) for label-free isolation of living circulating tumor cells,” Analytica Chimica Acta, Vol. 1255, 2023, Art. no. 341138. [19] D. N. Ku, D. P. Giddens, C. K. Zarins, and S. Glagov, “Pulsatile flow and atherosclerosis in the human carotid bifurcation: Positive correlation between plaque location and low and oscillating shear stress,” Atherosclerosis, Vol. 5, 1985, pp. 293–302, doi: 10.1161/01.atv.5.3.293. [20] A. M. Malek, S. L. Alper, and S. Izumo, “Hemodynamic shear stress and its role in atherosclerosis,” JAMA, Vol. 282, 1999, pp. 2035–2042, doi: 10.1001/jama.282.21.2035. [21] N. Bansal, “Prediabetes diagnosis and treatment: A review,” World Journal of Diabetes, Vol. 6, No. 2, 2015, pp. 296–303; American Diabetes Association, “Glycemic targets: Standards of medical care in diabetes—2020,” Diabetes Care, Vol. 43, No. Supplement 1, 2020, pp. S66–S76. [22] A. Kandwal, L. W. Liu, M. J. Deen, R. Jasrotia, B. K. Kanaujia, and Z. Nie, “Electromagnetic wave sensors for noninvasive blood glucose monitoring: Review and recent developments,” IEEE Transactions on Instrumentation and Measurement, Vol. 72, 2023, pp. 1–15. [23] C. Panagiotakis and A. A. Argyros, “Cell segmentation via region-based ellipse fitting,” Proceedings of the 25th IEEE International Conference on Image Processing (ICIP), Oct. 2018, pp. 2426–2430. [24] H.-J. Jeong, T.-Y. Kim, H.-G. Hwang, H.-J. Choi, H.-S. Park, and H.-K. Choi, “Comparison of thresholding methods for breast tumor cell segmentation,” Proceedings of the 7th International Workshop on Enterprise Networking and Computing in Healthcare Industry, Vol. 27, 2005, pp. 392–395.

Copyright

Copyright © 2025 S Krishna Anand, Anjan B, Chandana D C , Deepika K , Varshini R. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.