An Efficient, Lightweight, Tiny 2D-CNN Ensemble Model to Detect Cardiomegaly in Heart CT Images

Area of Science:

• Medical Imaging Analysis
• Artificial Intelligence in Healthcare
• Cardiology

Background:

• Cardiomegaly presents a significant global health challenge, particularly in developing nations.
• Late diagnoses due to limited access to radiography and radiologists in resource-limited settings contribute to increased mortality.
• Existing deep convolutional neural networks (CNNs) may have compatibility issues with object-oriented human factor technology.

Purpose of the Study:

• To develop a lightweight, tiny 2D-CNN ensemble model for computer-aided diagnosis (CAD) of cardiomegaly.
• To facilitate early detection of cardiomegaly, aiming to reduce mortality rates.
• To address the limitations of traditional deep CNNs in medical image analysis.

Main Methods:

• Utilized a tiny 2D-CNN ensemble learning model.
• Trained the model using CT scan images from the Mendeley data center.
• Focused on developing a compatible deep learning model for medical image diagnosis.

Main Results:

• The proposed tiny 2D-CNN ensemble model achieved a high accuracy of 96.32% in detecting cardiomegaly.
• Demonstrated the model's efficiency and accuracy in identifying the condition.
• The model shows promise for practical application in clinical settings.

Conclusions:

• The developed lightweight, tiny 2D-CNN ensemble model is an effective tool for accurate cardiomegaly detection.
• This AI-driven approach can aid in early diagnosis, especially in regions with limited healthcare resources.
• The system offers a viable solution to improve patient outcomes and reduce mortality associated with cardiomegaly.