Abstract:Acute lymphoblastic leukemia (ALL) is the most malignant form of leukemia and the most common cancer in adults and children. Traditionally, leukemia is diagnosed by analyzing blood and bone marrow smears under a microscope, with additional cytochemical tests for confirmation. However, these methods are expensive, time consuming, and highly dependent on expert knowledge. In recent years, deep learning, particularly Convolutional Neural Networks (CNNs), has provided advanced methods for classifying microscopic smear images, aiding in the detection of leukemic cells. These approaches are quick, cost effective, and not subject to human bias. However, most methods lack the ability to quantify uncertainty, which could lead to critical misdiagnoses. In this research, hybrid deep learning models (InceptionV3-GRU, EfficientNetB3-GRU, MobileNetV2-GRU) were implemented to classify ALL. Bayesian optimization was used to fine tune the model's hyperparameters and improve its performance. Additionally, Deep Ensemble uncertainty quantification was applied to address uncertainty during leukemia image classification. The proposed models were trained on the publicly available datasets ALL-IDB1 and ALL-IDB2. Their results were then aggregated at the score level using the sum rule. The parallel architecture used in these models offers a high level of confidence in differentiating between ALL and non-ALL cases. The proposed method achieved a remarkable detection accuracy rate of 100% on the ALL-IDB1 dataset, 98.07% on the ALL-IDB2 dataset, and 98.64% on the combined dataset, demonstrating its potential for accurate and reliable leukemia diagnosis.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the detection and classification of acute lymphoblastic leukemia (ALL). Specifically, the paper aims to develop an efficient and accurate automated system for identifying the presence of ALL and classifying it through microscopic blood sample images. Traditional ALL detection methods rely on professional medical personnel to perform microscopic analysis of bone marrow or blood smears. This method is not only costly and time - consuming, but also highly dependent on the operator's skills and knowledge. In addition, although most existing deep - learning methods perform well in image classification, they lack the quantification of the uncertainty of their results, which may lead to serious consequences. To this end, the paper proposes a hybrid feature - fusion deep - learning framework that combines gated recurrent units (GRU) and uncertainty quantification. The framework utilizes three pre - trained convolutional neural network (CNN) models: MobileNetV2 - GRU, InceptionV3 - GRU, and EfficientNetB3 - GRU, to extract features and perform classification through these models. To optimize the model performance, the study adopts Bayesian optimization techniques to select the best set of hyper - parameters. In addition, in order to increase the confidence in the classification stage and avoid over - confident diagnosis results, the study also introduces the uncertainty quantification method of deep ensemble. Finally, by aggregating the output results of three different hybrid deep - learning models at the score level using the summation rule, high - precision detection of ALL is achieved. The main contributions of the paper include: 1. Proposing a unique hybrid deep - learning model that combines three different hybrid DL models (MobileNetV2 - GRU, InceptionV3 - GRU, and EfficientNetB3 - GRU), which can automatically extract features from the data set and accurately distinguish ALL from non - ALL. 2. Introducing well - known uncertainty quantification methods such as deep ensemble to increase the confidence level in the classification stage and avoid over - confident diagnosis of the disease. 3. Using Bayesian optimization techniques to optimize the hyper - parameters of the deep - learning model to ensure that the model can achieve the best performance. 4. By training these three different hybrid deep - learning models on two publicly available blood sample data sets of leukemia patients (ALL - IDB1 and ALL - IDB2) and aggregating the results at the score level using the summation rule, leukemia detection of the input microscopic blood sample images is finally achieved. Through the above methods, the detection accuracy of the system proposed in the paper reaches 100% on the ALL - IDB1 data set, 98.07% on the ALL - IDB2 data set, and 98.64% on the combined data set. These results indicate that the system has the potential for rapid diagnosis and treatment in the actual clinical environment.

A Hybrid Feature Fusion Deep Learning Framework for Leukemia Cancer Detection in Microscopic Blood Sample Using Gated Recurrent Unit and Uncertainty Quantification

Feature Fusion Based Ensemble of Deep Networks for Acute Leukemia Diagnosis Using Microscopic Smear Images

Utilizing Deep Feature Fusion for Automatic Leukemia Classification: An Internet of Medical Things-Enabled Deep Learning Framework

Multi-Method Diagnosis of Blood Microscopic Sample for Early Detection of Acute Lymphoblastic Leukemia Based on Deep Learning and Hybrid Techniques

Refinement of ensemble strategy for acute lymphoblastic leukemia microscopic images using hybrid CNN-GRU-BiLSTM and MSVM classifier

A deep transfer learning based approaches for the detection and classification of acute lymphocytic leukemia using microscopic images

Acute Lymphoblastic Leukemia Detection from Microscopic Images Using Weighted Ensemble of Convolutional Neural Networks

BO-ALLCNN: Bayesian-Based Optimized CNN for Acute Lymphoblastic Leukemia Detection in Microscopic Blood Smear Images

A study on deep feature extraction to detect and classify Acute Lymphoblastic Leukemia (ALL)

An efficient decision support system for leukemia identification utilizing nature-inspired deep feature optimization

Acute lymphoblastic leukemia detection using ensemble features from multiple deep CNN models

Automated Detection of Acute Lymphoblastic Leukemia Subtypes from Microscopic Blood Smear Images using Deep Neural Networks

Novel Pipeline for Diagnosing Acute Lymphoblastic Leukemia Sensitive to Related Biomarkers

A Framework for Early Detection of Acute Lymphoblastic Leukemia and Its Subtypes From Peripheral Blood Smear Images Using Deep Ensemble Learning Technique

ALLNet: A Hybrid Convolutional Neural Network to Improve Diagnosis of Acute Lymphocytic Leukemia (ALL) in White Blood Cells

A Hybrid Deep Learning Architecture for Leukemic B-lymphoblast Classification

AUTOMATED ACUTE LYMPHOBLASTIC LEUKEMIA CELL CLASSIFICATION USING OPTIMIZED CONVOLUTIONAL NEURAL NETWORK

Deep learning model for differentiating acute myeloid and lymphoblastic leukemia in peripheral blood cell images via myeloblast and lymphoblast classification

A hybrid detection model for acute lymphocytic leukemia using support vector machine and particle swarm optimization (SVM-PSO)

A fast and efficient CNN model for B‐ALL diagnosis and its subtypes classification using peripheral blood smear images