Cancer Cell Classification From Peripheral Blood Smear Data Using the YOLOv8 Architecture

The accurate classification of cancer cells in the peripheral blood is essential for the diagnosis of leukemia and has traditionally been carried out by analyzing laboratory images. In this context, the use of deep learning techniques facilitates decision-making and accelerates the early diagnosis o...

Full description

Saved in:
Bibliographic Details
Main Authors: Joao C. S. Nunes, Jose E. B. De S. Linhares, Miguel Angel Orellana Postigo, Daniel Guzman del Rio, Angilberto Muniz Ferreira Sobrinho, Israel Gondres Torne
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11014521/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850230646725672960
author Joao C. S. Nunes
Jose E. B. De S. Linhares
Miguel Angel Orellana Postigo
Daniel Guzman del Rio
Angilberto Muniz Ferreira Sobrinho
Israel Gondres Torne
author_facet Joao C. S. Nunes
Jose E. B. De S. Linhares
Miguel Angel Orellana Postigo
Daniel Guzman del Rio
Angilberto Muniz Ferreira Sobrinho
Israel Gondres Torne
author_sort Joao C. S. Nunes
collection DOAJ
description The accurate classification of cancer cells in the peripheral blood is essential for the diagnosis of leukemia and has traditionally been carried out by analyzing laboratory images. In this context, the use of deep learning techniques facilitates decision-making and accelerates the early diagnosis of the disease, allowing preventive measures to be adopted for the patient. This study explores the application of the YOLOv8 deep learning architecture for the classification of cancer cells in blood smear images owing, to its ability to perform this task quickly and accurately. The models were trained on two datasets, ALL and C-NMC Leukemia, and evaluated using top-1 accuracy, top-5 accuracy and loss metrics. The proposed approach achieved a top-1 accuracy of 99.982% and top-5 accuracy of 100% on the validation and test subsets of the ALL dataset respectively, with a final loss of 0.02952. For the C-NMC Leukemia dataset, the model obtained 66.897% and 89.612% top-1 accuracy in the validation and test subsets, respectively, while maintaining 100% top-5 accuracy in both subsets, with a loss of 0.18289. These results demonstrate the effectiveness of YOLOv8 in classifying cancer cells, particularly in the ALL set. However, future strategies such as expanding the data set and fine-tuning the hyperparameters could contribute to better generalization of the model to different data distributions.
format Article
id doaj-art-b83eb258573c41ff820a8b5a27d0bbf4
institution OA Journals
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-b83eb258573c41ff820a8b5a27d0bbf42025-08-20T02:03:47ZengIEEEIEEE Access2169-35362025-01-0113919119192410.1109/ACCESS.2025.357327711014521Cancer Cell Classification From Peripheral Blood Smear Data Using the YOLOv8 ArchitectureJoao C. S. Nunes0https://orcid.org/0009-0006-3826-7765Jose E. B. De S. Linhares1https://orcid.org/0000-0002-7397-4426Miguel Angel Orellana Postigo2https://orcid.org/0000-0003-4827-7677Daniel Guzman del Rio3https://orcid.org/0000-0003-4321-0465Angilberto Muniz Ferreira Sobrinho4https://orcid.org/0000-0002-5558-3931Israel Gondres Torne5https://orcid.org/0000-0002-1267-1878PPGEEL-Postgraduate Program in Electrical Engineering, State University of Amazonas (UEA), Manaus, BrazilFederal Institute of Education, Science, and Technology of Amazonas-IFAM, Manaus, BrazilPPGEEL-Postgraduate Program in Electrical Engineering, State University of Amazonas (UEA), Manaus, BrazilPPGEEL-Postgraduate Program in Electrical Engineering, State University of Amazonas (UEA), Manaus, BrazilPPGEEL-Postgraduate Program in Electrical Engineering, State University of Amazonas (UEA), Manaus, BrazilPPGEEL-Postgraduate Program in Electrical Engineering, State University of Amazonas (UEA), Manaus, BrazilThe accurate classification of cancer cells in the peripheral blood is essential for the diagnosis of leukemia and has traditionally been carried out by analyzing laboratory images. In this context, the use of deep learning techniques facilitates decision-making and accelerates the early diagnosis of the disease, allowing preventive measures to be adopted for the patient. This study explores the application of the YOLOv8 deep learning architecture for the classification of cancer cells in blood smear images owing, to its ability to perform this task quickly and accurately. The models were trained on two datasets, ALL and C-NMC Leukemia, and evaluated using top-1 accuracy, top-5 accuracy and loss metrics. The proposed approach achieved a top-1 accuracy of 99.982% and top-5 accuracy of 100% on the validation and test subsets of the ALL dataset respectively, with a final loss of 0.02952. For the C-NMC Leukemia dataset, the model obtained 66.897% and 89.612% top-1 accuracy in the validation and test subsets, respectively, while maintaining 100% top-5 accuracy in both subsets, with a loss of 0.18289. These results demonstrate the effectiveness of YOLOv8 in classifying cancer cells, particularly in the ALL set. However, future strategies such as expanding the data set and fine-tuning the hyperparameters could contribute to better generalization of the model to different data distributions.https://ieeexplore.ieee.org/document/11014521/Leukemiacomputer visionobject classificationdeep learningYOLOv8
spellingShingle Joao C. S. Nunes
Jose E. B. De S. Linhares
Miguel Angel Orellana Postigo
Daniel Guzman del Rio
Angilberto Muniz Ferreira Sobrinho
Israel Gondres Torne
Cancer Cell Classification From Peripheral Blood Smear Data Using the YOLOv8 Architecture
IEEE Access
Leukemia
computer vision
object classification
deep learning
YOLOv8
title Cancer Cell Classification From Peripheral Blood Smear Data Using the YOLOv8 Architecture
title_full Cancer Cell Classification From Peripheral Blood Smear Data Using the YOLOv8 Architecture
title_fullStr Cancer Cell Classification From Peripheral Blood Smear Data Using the YOLOv8 Architecture
title_full_unstemmed Cancer Cell Classification From Peripheral Blood Smear Data Using the YOLOv8 Architecture
title_short Cancer Cell Classification From Peripheral Blood Smear Data Using the YOLOv8 Architecture
title_sort cancer cell classification from peripheral blood smear data using the yolov8 architecture
topic Leukemia
computer vision
object classification
deep learning
YOLOv8
url https://ieeexplore.ieee.org/document/11014521/
work_keys_str_mv AT joaocsnunes cancercellclassificationfromperipheralbloodsmeardatausingtheyolov8architecture
AT joseebdeslinhares cancercellclassificationfromperipheralbloodsmeardatausingtheyolov8architecture
AT miguelangelorellanapostigo cancercellclassificationfromperipheralbloodsmeardatausingtheyolov8architecture
AT danielguzmandelrio cancercellclassificationfromperipheralbloodsmeardatausingtheyolov8architecture
AT angilbertomunizferreirasobrinho cancercellclassificationfromperipheralbloodsmeardatausingtheyolov8architecture
AT israelgondrestorne cancercellclassificationfromperipheralbloodsmeardatausingtheyolov8architecture