Acute Lymphoblastic Leukemia Diagnosis Employing YOLOv11, YOLOv8, ResNet50, and Inception-ResNet-v2 Deep Learning Models

TL;DR

The paper addresses the challenge of accurately detecting Acute Lymphoblastic Leukemia from multi-cell blood images by evaluating four deep learning models—YOLOv11, YOLOv8, ResNet50, and Inception-ResNet-v2—on merged datasets that simulate real-world conditions. It proposes a two-stage workflow with HSV-based segmentation and extensive data augmentation, followed by transfer learning and selective fine-tuning of pretrained networks. The results show top accuracies of up to 99.7% (Inception-ResNet-v2) and strong performance across all models, including 97.3–98.2% on validation/test splits, with Inception-ResNet-v2 delivering the best overall metrics (accuracy and specificity). The findings suggest that combining multi-cell ALL datasets and modern deep architectures can yield robust, deployable diagnostics for ALL detection in practical clinical settings, with future work focused on broader data collection and edge-device deployment.

Abstract

Thousands of individuals succumb annually to leukemia alone. As artificial intelligence-driven technologies continue to evolve and advance, the question of their applicability and reliability remains unresolved. This study aims to utilize image processing and deep learning methodologies to achieve state-of-the-art results for the detection of Acute Lymphoblastic Leukemia (ALL) using data that best represents real-world scenarios. ALL is one of several types of blood cancer, and it is an aggressive form of leukemia. In this investigation, we examine the most recent advancements in ALL detection, as well as the latest iteration of the YOLO series and its performance. We address the question of whether white blood cells are malignant or benign. Additionally, the proposed models can identify different ALL stages, including early stages. Furthermore, these models can detect hematogones despite their frequent misclassification as ALL. By utilizing advanced deep learning models, namely, YOLOv8, YOLOv11, ResNet50 and Inception-ResNet-v2, the study achieves accuracy rates as high as 99.7%, demonstrating the effectiveness of these algorithms across multiple datasets and various real-world situations.

Figures (28)

• Figure 1: Workflow Diagram
• Figure 2: The implementation process
• Figure 3: Data samples before and after image segmentation. (a) Before segmentation and (b) After segmentation.
• Figure 4: YOLOv8, YOLOv11, ResNet50 and Inception-ResNet-v2 Models Training and Evaluation
• Figure 5: YOLOv11n accuracy with SGD