Applying Transfer Learning on 3D Brain Images and an MLOPS Study for Deployment
Millions of life can be saved worldwide if there is a possibility to identify the brain tumor at an early stage and if it can be integrated cost effectively with the medical imaging system. The MRI images which are generated for other treatments like accidents, pain therapy etc also can be automatic...
Saved in:
| Published in: | 2023 9th International Conference on Smart Computing and Communications (ICSCC) pp. 541 - 547 |
|---|---|
| Main Authors: | , |
| Format: | Conference Proceeding |
| Language: | English |
| Published: |
IEEE
17-08-2023
|
| Subjects: | |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Millions of life can be saved worldwide if there is a possibility to identify the brain tumor at an early stage and if it can be integrated cost effectively with the medical imaging system. The MRI images which are generated for other treatments like accidents, pain therapy etc also can be automatically analyzed in detail helping the doctors to identify the initial stages of brain tumor even in unsuspected cases. But it is always challenging to find a perfect algorithm with a high accuracy to detect the brain tumors at an earlier stage. In this study we propose to use the transfer learning combined with ResGANet model for an improved accuracy. The algorithm was evaluated on BRATS-2020 and BRATS-2021 data sets. The final aim of any machine learning project is to develop the product and bring this into solving real world problems very quickly. Normally this is not accomplished at normal hospitals due to the complexity of integration and the lack of deep computer science skills. This study proposes a cost-effective method for deploying this model in google cloud which can be integrated with any hospital system and can be operated by normal technicians with little effort on training. |
|---|---|
| DOI: | 10.1109/ICSCC59169.2023.10335014 |