Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Explainable Ensemble Learning With Stain Normalization and Deep Feature Extraction for Acute Lymphoblastic Leukaemia

Faysal Ahmmed¹, Md Sadi Al Huda², Md Asraf Ali³

¹Department of Computer Science and Engineering American International University-Bangladesh (AIUB) Dhaka Bangladesh.

Healthcare Technology Letters

|July 15, 2026

Summary

Related Concept Videos

Classification of Leukocytes

Classification of Leukocytes

Leukocytes are classified into two groups based on the presence or absence of cytoplasmic granules. Granular leukocytes, which contain granules, belong to the myeloid lineage and are divided into three subtypes: neutrophils, eosinophils, and basophils. These cells are roughly spherical and characterized by the granules in their cytoplasm.
Neutrophils are the most abundant type of granular leukocytes, comprising 50-70% of all leukocytes. They feature small, evenly distributed granules and a...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

CBAM-Xception: An Attention-Guided Framework for Skin Cancer Classification.

Journal of imaging informatics in medicine·2026

Same author

BTdiagAI: A Web-Deployed Hybrid Framework for Brain Tumor Classification Using Optimized MRI Preprocessing and Deep Learning Fusion.

Healthcare technology letters·2026

Same author

A robust hydroponic system for horticulture farming using deep learning, IoT, and mobile application.

PloS one·2025

Same author

Walking Speed Classification from Marker-Free Video Images in Two-Dimension Using Optimum Data and a Deep Learning Method.

Bioengineering (Basel, Switzerland)·2022

Same author

Bearing Fault Diagnosis Using Multidomain Fusion-Based Vibration Imaging and Multitask Learning.

Sensors (Basel, Switzerland)·2022

Same author

Vision-Based Autonomous Crack Detection of Concrete Structures Using a Fully Convolutional Encoder-Decoder Network.

Sensors (Basel, Switzerland)·2019

Early diagnosis of acute lymphoblastic leukaemia (ALL) is crucial. A new AI system using peripheral blood smear images achieves 99.95% accuracy for rapid ALL detection and subtyping, improving patient outcomes.

Area of Science:

Oncology
Medical Imaging
Artificial Intelligence

Background:

Acute lymphoblastic leukaemia (ALL) is a significant global health concern, with advanced stages leading to poorer prognoses.
Current diagnostic methods for ALL are often invasive, costly, and time-consuming, potentially delaying critical treatment.
Peripheral blood smear (PBS) image analysis offers a preliminary screening tool, but interpretation challenges can lead to misdiagnosis.

Purpose of the Study:

To develop and validate an automated system for the early and accurate diagnosis of ALL and its subtypes using PBS images.
To overcome the limitations of manual interpretation and improve diagnostic efficiency and precision.

Main Methods:

Utilized a dataset of 20,000 colour-normalized PBS images (Vahadane method).

Keywords:

B‐cell ALL (B‐ALL)VGG16 Vahadane acute lymphoblastic leukaemia (ALL)deep neural network image processing machine learning peripheral blood smear images stacking ensemble learning

Related Experiment Videos

Employed deep neural networks for feature extraction combined with a stacking ensemble learning approach for classification.

Developed a system capable of distinguishing ALL from benign conditions (haematogones) and identifying specific ALL subtypes (Pro-B, Pre-B).

Main Results:

The proposed system achieved exceptional performance metrics: 99.95% accuracy, 99.95% recall, 99.95% precision, and 99.95% F1-score.
Demonstrated the ability to accurately differentiate ALL cases from haematogones and classify ALL subtypes.
Comparative analysis confirmed superior performance over established machine learning algorithms.

Conclusions:

The AI-powered system shows significant potential as a proof-of-concept for early ALL detection and subtyping.
This approach can streamline the diagnostic process, reducing clinician and patient burden.
Further clinical validation is required prior to deployment in healthcare settings.