HematoFusion: A Weighted Residual-Vision Transformer Ensemble for Automated Classification of Haematologic Disorders in Microscopic Blood Images
Abstract
Haematologic malignancies pose a significant global challenge, with 1.34 million new cases reported in 2019 and leukemia claiming 311,594 lives in 2020. Early diagnosis of these blood disorders increases survival chances by enabling prompt treatment, yet their complexity and variable cellular morphology hinder accurate detection. Advances in Medical Imaging and AI, particularly Image Classification, offer solutions by analyzing blood samples for subtle morphological patterns. This study advances the field by introducing a novel data set for the classification of red blood cells and using open-source data for the classification of leukemia and lymphoma (each covering 29,363; 16,811; and 1,436 images, respectively).
We fine-tuned multiple AI models, including EfficientNetB3, ResNet50V2, and a pretrained Vision Transformer (ViT), and combined their strengths into a weighted ensemble framework. Evaluated across various metrics (including accuracy, precision, recall, etc.), the proposed HematoFusion model excelled, achieving
96% accuracy in the morphology of red blood cells, 99% in Leukemia, and 96% in Lymphoma, surpassing most existing models in terms of accuracy while covering a wider range of haematologic disorders. These findings demonstrate the potential of integrated AI frameworks to improve haematologic diagnostics with precision and reliability.
References
Frank Rosenblatt. “The perceptron: A probabilistic
model for information storage and organization in
the brain”. In: Psychological Review 65.6 (1958),
pp. 386–408.
Jacob Cohen. “A coefficient of agreement for nominal
scales”. In: Educational and psychological measurement
1 (1960), pp. 37–46. doi: 10 . 1177 /
Melvin Earl Maron and John Larry Kuhns. “On relevance,
probabilistic indexing and information retrieval”.
In: Journal of the ACM (JACM) 7.3 (1960),
pp. 216–244. doi: 10.1145/321033.321035.
C Van Rijsbergen. “Information retrieval: theory
and practice”. In: Proceedings of the joint
IBM/University of Newcastle upon tyne seminar on
data base systems. Ed. by B Shaw. Vol. 79. University
Computing Laboratory: University of Newcastle
Upon Tyne Computing Laboratory, 1979, pp. 1–14.
David E Rumelhart, Geoffrey E Hinton, and Ronald
J Williams. “Learning representations by back-propagating errors”. In: Nature 323.6088 (1986),
pp. 533–536.
Yann LeCun et al. “Gradient-based learning applied
to document recognition”. In: Proceedings of the
IEEE 86.11 (1998), pp. 2278–2324. doi: 10.1109/
726791.
Thomas G Dietterich. “Ensemble methods in machine
learning”. In: International workshop on multiple
classifier systems. Ed. by Fabio Roli Josef Kittler.
Cagliari, Italy: Springer, 2000, pp. 1–15.
KS Kim et al. “Analyzing blood cell image to distinguish
its abnormalities”. In: Proceedings of the
eighth ACM international conference on multimedia. New York: Association for Computing Machinery,
, pp. 395–397. doi: https://doi.org/
1145/354384.354543.
Jia Deng et al. “ImageNet: A large-scale hierarchical
image database”. In: 2009 IEEE conference on computer
vision and pattern recognition. Miami: Ieee,
, pp. 248–255. doi: 10 . 1109 / CVPR . 2009 .
Nikita Orlov et al. “Automatic Classification of
Lymphoma Images With Transform-Based Global
Features”. In: IEEE transactions on information
technology in biomedicine : a publication of the
IEEE Engineering in Medicine and Biology Society
(2010), pp. 1003–13. doi: 10.1109/TITB.
2050695.
Lior Rokach. “Ensemble-based classifiers”. In: Artificial
intelligence review 33 (2010), pp. 1–39. doi:
1007/s10462-009-9124-7.
Alex Krizhevsky, Ilya Sutskever, and Geoffrey E.
Hinton. “ImageNet Classification with Deep Convolutional
Neural Networks”. In: Advances in Neural
Information Processing Systems. Ed. by F. Pereira et
al. Vol. 25. Lake Tahoe, Nevada: Curran Associates,
Inc., 2012, pp. 1097–1105.
K. He et al. Deep Residual Learning for Image
Recognition. Preprint at https : / / arxiv . org /
abs/1512.03385. 2015.
Mahsa Lotfi et al. “The detection of dacrocyte, schistocyte
and elliptocyte cells in iron deficiency anemia”.
In: 2015 2nd International conference on pattern
recognition and image analysis (IPRIA). Rasht,
Iran: IEEE, 2015, pp. 1–5.
O. Russakovsky et al. ImageNet Large Scale Visual
Recognition Challenge. Preprint at https://
arxiv.org/abs/1409.0575. 2015.
J. C. Chapin and M. T. Desancho. “Hematologic
Dysfunction in the ICU”. In: Critical Care. Ed. by
J. M. Oropello, S. M. Pastores, and V. Kvetan. New
York: McGraw-Hill Education, 2016.
Kaiming He et al. Identity Mappings in Deep Residual
Networks. Preprint at http : / / arxiv . org /
abs/1603.05027. 2016.
Kenneth Kaushansky et al. Williams Hematology.
New York: McGraw-Hill Education, 2016.
Gao Huang et al. “Densely Connected Convolutional
Networks”. In: Proceedings of the IEEE conference
on computer vision and pattern recognition.
Honolulu: IEEE, 2017, pp. 4700–4708.
Yiyue Jiang et al. “Label-free detection of aggregated
platelets in blood by machine-learning-aided
optofluidic time-stretch microscopy”. In: Lab on a
Chip 17.14 (2017), pp. 2426–2434. doi: 10.1039/
C7LC00396J.
Mazin Z Othman, Thabit S Mohammed, and Alaa B
Ali. “Neural network classification of white blood
cell using microscopic images”. In: International
Journal of Advanced Computer Science and Applications
5 (2017), pp. 99–103.
Mohammad Fadly Syahputra, Anita Ratna Sari, and
Romi Fadillah Rahmat. “Abnormality classification
on the shape of red blood cells using radial basis
function network”. In: 2017 4th International Conference
on Computer Applications and Information
Processing Technology (CAIPT). Kuta Bali, Indonesia: IEEE, 2017, pp. 1–5. doi: 10 . 1109 / CAIPT .
8320739.
Ashish Vaswani et al. “Attention is all you need”. In:
Advances in Neural Information Processing Systems
(2017). doi: https://doi.org/10.48550/
arXiv.1706.03762.
Hajara Abdulkarim Aliyu et al. “Red blood cell classification:
deep learning architecture versus support
vector machine”. In: 2018 2nd international conference
on biosignal analysis, processing and systems
(ICBAPS). Kuching, Malaysia: IEEE, 2018,
pp. 142–147. doi: 10 . 1109 / ICBAPS . 2018 .
Paul Mooney. Blood Cell Images. 2018. url:
https : / / www . kaggle . com / datasets /
paultimothymooney/blood-cells.
Mariam Nassar et al. “Label-free identification of
white blood cells using machine learning”. In: Cytometry
Part A 95.8 (2019), pp. 836–842. doi: 10.
/cyto.a.23794.
N. C. Shenggan. BCCD Dataset. https : / /
github.com/Shenggan/BCCD_Dataset. 2019.
Mingxing Tan and Quoc V. Le. “EfficientNet: Rethinking
Model Scaling for Convolutional Neural
Networks”. In: Proceedings of the 36th International
Conference on Machine Learning. Vol. 97.
Long Beach, California: PMLR, 2019, pp. 6105–
Laith Alzubaidi et al. “Classification of red blood
cells in sickle cell anemia using deep convolutional
neural network”. In: Intelligent Systems Design
and Applications. Ed. by Ajith Abraham et
al. Vol. 1. Cham: Springer International Publishing,
, pp. 6–8. doi: 10.1007/978-3-030-16657-
_51.
Yasmin M Kassim et al. “Clustering-Based Dual
Deep Learning Architecture for Detecting Red
Blood Cells in Malaria Diagnostic Smears”. In:
IEEE Journal of Biomedical and Health Informatics
5 (2020), pp. 1735–1746. doi: 10.1109/JBHI.
3034863.
Thomas Wolf et al. HuggingFace’s Transformers:
State-of-the-art Natural Language Processing.
arXiv: 1910.03771 [cs.CL]. url: https:
//arxiv.org/abs/1910.03771.
Yassine Barhoumi and Ghulam Rasool. Scopeformer:
n-CNN-ViT hybrid model for intracranial
hemorrhage classification. Preprint at https : / /
arxiv.org/abs/2107.04575. 2021. doi: https:
//doi.org/10.48550/arXiv.2107.04575.
Alexey Dosovitskiy et al. An Image is Worth 16x16
Words: Transformers for Image Recognition at
Scale. Preprint at https : / / arxiv . org / abs /
11929. 2021.
Mawaddah Harahap et al. “Implementation of Convolutional
Neural Network in the classification
of red blood cells have affected of malaria”. In:
Sinkron: jurnal dan penelitian teknik informatika 5.2
(2021), pp. 199–207. doi: 10 . 33395 / sinkron .
v5i2.10713.
JGraph. diagrams.net, draw.io. Oct. 2021. url:
https://www.diagrams.net/.
Zhencun Jiang et al. “Method for diagnosis of
acute lymphoblastic leukemia based on ViT-CNN
ensemble model”. In: Computational Intelligence
and Neuroscience 2021.1 (2021), p. 7529893. doi:
https://doi.org/10.1155/2021/7529893.
Korranat Naruenatthanaset et al. Red Blood Cell
Segmentation with Overlapping Cell Separation and
Classification on Imbalanced Dataset. Preprint at
https://arxiv.org/abs/2012.01321. 2021.
Maithra Raghu et al. “Do vision transformers see
like convolutional neural networks?” In: Advances
in neural information processing systems 34 (2021),
pp. 12116–12128. doi: https : / / doi . org / 10 .
/arXiv.2108.08810.
Xufeng Yao et al. “Classification of white blood
cells using weighted optimized deformable convolutional
neural networks”. In: Artificial Cells,
Nanomedicine, and Biotechnology 49.1 (2021),
pp. 147–155. doi: 10 . 1080 / 21691401 . 2021 .
Kai Jiang et al. “The encoding method of position
embeddings in vision transformer”. In: Journal of
Visual Communication and Image Representation
(2022), p. 103664. doi: https : / / doi . org /
1016/j.jvcir.2022.103664.
Zahra Mousavi Kouzehkanan et al. “A large dataset
of white blood cells containing cell locations and
types, along with segmented nuclei and cytoplasm”.
In: Scientific Reports 12.1 (2022), p. 1123. doi: 10.
/s41598-021-04426-x.
Sarang Sharma et al. “[Retracted] Deep Learning
Model for the Automatic Classification of White
Blood Cells”. In: Computational Intelligence and
Neuroscience 2022.1 (2022). Retracted due to issues
with the publication process. For retraction details,
see the retraction notice at https : / / pmc .
ncbi.nlm.nih.gov/articles/PMC10732974/.,
p. 7384131.
Dyah Aruming Tyas et al. “Erythrocyte (red blood
cell) dataset in thalassemia case”. In: Data in Brief
(2022), p. 107886. doi: https://doi.org/10.
/j.dib.2022.107886.
Mohammed Hamdi et al. “Hybrid Models Based on
Fusion Features of a CNN and Handcrafted Features
for Accurate Histopathological Image Analysis
for Diagnosing Malignant Lymphomas”. In: Diagnostics
13 (2023), p. 2258. doi: 10 . 3390 /
diagnostics13132258.
Rojina Kashefi et al. Explainability of Vision Transformers:
A Comprehensive Review and New Perspectives.
Preprint at https://arxiv.org/abs/
06786. 2023.
Enquan Yang et al. “DRNet: Dual-stage refinement
network with boundary inference for RGB-D semantic
segmentation of indoor scenes”. In: Engineering
Applications of Artificial Intelligence 125
(2023), p. 106729. issn: 0952-1976. doi: https://
doi.org/10.1016/j.engappai.2023.106729.
Mouna Saadallah. Red Blood Cell Morphology
Dataset for Image Classification. Zenodo, Feb.
doi: 10 . 5281 / 14936017. url: https : / /
zenodo.org/records/14936017.
DOI: https://doi.org/10.31449/inf.v49i16.9397
This work is licensed under a Creative Commons Attribution 3.0 License.