Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Early Diagnosis of Acute Lymphoblastic Leukemia Using YOLOv8 and YOLOv11 Deep Learning Models

Alaa Awad, Salah A. Aly

TL;DR

The paper addresses early detection of Acute Lymphoblastic Leukemia (ALL) from multi-cell blood smear images by applying transfer learning with YOLOv8 and YOLOv11 on a fused two-class Normal vs Cancer dataset derived from ALL image datasets. It employs HSV-based OpenCV segmentation, mosaic and other augmentations, and explores multiple optimizers with 50 training epochs to achieve high accuracy. Key contributions include first use of YOLOv11 for ALL detection, demonstration of strong cross-dataset generalization, and competitive accuracy up to 98.8% under certain configurations, surpassing several prior CNN-based approaches. The work demonstrates potential for robust, real-world ALL screening across diverse samples including hematogones, supporting faster and more reliable clinical decision-making.

Abstract

Leukemia, a severe form of blood cancer, claims thousands of lives each year. This study focuses on the detection of Acute Lymphoblastic Leukemia (ALL) using advanced image processing and deep learning techniques. By leveraging recent advancements in artificial intelligence, the research evaluates the reliability of these methods in practical, real-world scenarios. Specifically, it examines the performance of state-of-the-art YOLO models, including YOLOv8 and YOLOv11, to distinguish between malignant and benign white blood cells and accurately identify different stages of ALL, including early stages. Moreover, the models demonstrate the ability to detect hematogones, which are frequently misclassified as ALL. With accuracy rates reaching 98.8%, this study highlights the potential of these algorithms to provide robust and precise leukemia detection across diverse datasets and conditions.

Early Diagnosis of Acute Lymphoblastic Leukemia Using YOLOv8 and YOLOv11 Deep Learning Models

TL;DR

The paper addresses early detection of Acute Lymphoblastic Leukemia (ALL) from multi-cell blood smear images by applying transfer learning with YOLOv8 and YOLOv11 on a fused two-class Normal vs Cancer dataset derived from ALL image datasets. It employs HSV-based OpenCV segmentation, mosaic and other augmentations, and explores multiple optimizers with 50 training epochs to achieve high accuracy. Key contributions include first use of YOLOv11 for ALL detection, demonstration of strong cross-dataset generalization, and competitive accuracy up to 98.8% under certain configurations, surpassing several prior CNN-based approaches. The work demonstrates potential for robust, real-world ALL screening across diverse samples including hematogones, supporting faster and more reliable clinical decision-making.

Abstract

Leukemia, a severe form of blood cancer, claims thousands of lives each year. This study focuses on the detection of Acute Lymphoblastic Leukemia (ALL) using advanced image processing and deep learning techniques. By leveraging recent advancements in artificial intelligence, the research evaluates the reliability of these methods in practical, real-world scenarios. Specifically, it examines the performance of state-of-the-art YOLO models, including YOLOv8 and YOLOv11, to distinguish between malignant and benign white blood cells and accurately identify different stages of ALL, including early stages. Moreover, the models demonstrate the ability to detect hematogones, which are frequently misclassified as ALL. With accuracy rates reaching 98.8%, this study highlights the potential of these algorithms to provide robust and precise leukemia detection across diverse datasets and conditions.

Paper Structure

This paper contains 8 sections, 5 equations, 7 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Datasets and Data Collections
  3. Models and Methodologies
  4. Performance Metrics
  5. Experimental Results
  6. Related Work
  7. Comparison Study
  8. Conclusion

Figures (7)

  • Figure 1: The implementation process including the data preparation and models training and evaluation
  • Figure 2: Performance metrics for YOLOv11s using SGD optimizer. (a) Train loss, (b) Accuracy
  • Figure 3: Performance metrics for YOLOv11s using AdamW optimizer. (a) Train loss, (b) Accuracy
  • Figure 4: Normalized confusion matrix for YOLOv11s.
  • Figure 5: Performance metrics for YOLOv8s using SGD optimizer. (a) Train loss, (b) Accuracy
  • ...and 2 more figures