Automated diagnosis of atherosclerosis using multi-layer ensemble models and bio-inspired optimization in intravascular ultrasound imaging

Nisha K Prajapati ^1,2, Amitkumar Patel ³, Hiren Mewada ^4,5,6

⁰Government Engineering College, Gandhinagar, Gujarat, India.

Medical & Biological Engineering & Computing +

|

September 18, 2024

View abstract on PubMed

Summary

This study introduces an automated method for classifying atherosclerotic plaques using deep learning on IVUS images. The novel approach achieves high accuracy, improving diagnosis efficiency for cardiovascular disease.

Area Of Science

Medical Imaging
Artificial Intelligence
Cardiovascular Disease Research

Background

Atherosclerosis, a primary cause of heart disease, involves plaque buildup in arteries.
Intravascular Ultrasound (IVUS) imaging offers detailed views of coronary arteries and plaque morphology.
Current diagnostic methods, manual or software-based, are time-consuming, necessitating automated solutions.

Purpose Of The Study

To propose an automated atherosclerotic plaque classification method using a hybrid Ant Lion Optimizer with Deep Learning (AAPC-HALODL) on IVUS images.
To enhance the accuracy and efficiency of atherosclerosis detection and classification.
To leverage deep learning and computer vision for automatic plaque characterization.

Main Methods

Utilized a Faster Region Convolutional Neural Network (Faster RCNN) for segmenting diseased regions in IVUS images.
Employed the ShuffleNet-v2 model for feature vector generation, with hyperparameters optimized by the Hybrid Ant Lion Optimizer (HALO) technique.
Implemented an average ensemble classification using Stacked Autoencoder (SAE) and Deep Extreme Learning Machine (DELM) models.

Main Results

The AAPC-HALODL method demonstrated superior performance compared to other deep learning models on the MICCAI Challenge 2011 dataset.
Achieved a maximum accuracy of 98.33%, precision of 97.87%, sensitivity of 98.33%, and F-score of 98.10%.
The hybrid optimization and ensemble classification effectively improved plaque classification accuracy.

Conclusions

The proposed AAPC-HALODL technique offers a highly accurate and efficient automated solution for atherosclerotic plaque classification from IVUS images.
This automated approach has the potential to significantly aid healthcare professionals in diagnosing and managing cardiovascular disease.
Deep learning and optimization techniques show great promise for advancing medical image analysis in cardiology.

Keywords:

Atherosclerotic plaque Computer-aided diagnosis Deep learning Healthcare Intravascular ultrasound images Metaheuristics