Post Your Answer

Can we modify Deep CNN technique for enhancing Brain tumor segmentation?

Algorithms

I am currently working on my research under brain tumour segmentation. I have planned to employ deep convolutional networks for performing segmentation of brain tumour from the image dataset. Few experts have suggested that deep CNN technique has already been used in various research works. Can anyone suggest me an approach which can be integrated with CNN for enhancing the novelty.

Also, please suggest some feasible optimization algorithms for optimizing the performance of Deep CNN models.

All Answers (1 Answers In All)

By Fathima M Answered 3 years ago

The proposed deep CNN can be combined with reinforcement-based learning approaches for achieving effective segmentation accuracy. RL based approaches are effective in providing desired accuracy and performance efficiency. The deep CNN based approach can be combined with RL-based algorithms such as Long-short term memory and generative adversarial networks. This can bring novelty in the proposed approach. Bio-inspired algorithms such as modified BAT algorithm, ANT optimization algorithm and Cuckoo search algorithm can be used for optimizing the performance of deep CNN models. These algorithms are simple for implementation and are effective optimization techniques.