Abstract:
Doctors rely on volumetric imaging. Convolutional neural networks (CNN) for volumetric image analysis require detailed training data and GPU memory.
This project presents the volumetric image classification problem as a multi-instance classification problem and proposes a novel method to adaptively select positive instances from positive bags during training.
Extreme value theory is used to model the feature distribution of images without pathologies and identify positive instances of imaged pathologies.
The experimental results on three image classification tasks—classifying retinal OCT images according to fluid build-ups, detecting emphysema in pulmonary 3D-CT images, and detecting cancerous regions in 2D histopathology images—show that the proposed method produces classifiers that perform similarly to fully supervised methods and achieves state-of-the-art performance in all test cases.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
