Related Experiment Video
Updated: Jul 16, 2025
07:42
Magnetic Levitation Coupled with Portable Imaging and Analysis for Disease Diagnostics
Published on: February 19, 2017
8.8K
DSCNet: Deep Skip Connections-Based Dense Network for ALL Diagnosis Using Peripheral Blood Smear Images
Manjit Kaur1, Ahmad Ali AlZubi2, Arpit Jain3
1School of Computer Science and Artificial Intelligence, SR University, Warangal 506371, India.
Diagnostics (Basel, Switzerland)
|September 9, 2023
Summary
A new deep learning model, DSCNet, improves the accuracy of diagnosing acute lymphoblastic leukemia (ALL) from blood smear images. This AI approach offers faster and more reliable detection for better patient outcomes.
Area of Science:
- Hematology
- Medical Imaging
- Artificial Intelligence
Background:
- Acute lymphoblastic leukemia (ALL) is a serious blood cancer requiring prompt diagnosis.
- Manual ALL diagnosis is slow and can impede timely treatment.
- Existing deep learning (DL) models for ALL diagnosis face challenges like complexity and sensitivity to image quality.
Purpose of the Study:
- To develop a novel deep learning network, DSCNet, for accurate ALL diagnosis using peripheral blood smear images.
- To overcome limitations of current DL methods in ALL detection.
Main Methods:
- Proposed a Deep Skip Connections-Based Dense Network (DSCNet) architecture.
- Integrated skip connections, custom image filtering, KL divergence loss, and dropout regularization.
- Trained and evaluated DSCNet on an augmented dataset of ALL peripheral blood smear images.
Main Results:
- DSCNet demonstrated superior performance compared to existing methods.
- Achieved significant improvements in accuracy (1.25%), sensitivity (1.32%), specificity (1.12%), F-score (1.24%), and AUC (1.23%).
Conclusions:
- DSCNet shows significant potential as an effective tool for early and accurate ALL diagnosis.
- The model's performance suggests clinical applicability for improving patient outcomes and advancing leukemia detection research.
