WD for 3D-WNO

The aim of this project was to develop a Wavelet Decomposition (WD)module as part of the larger effort to create a 3D-Wavelet Neural Operator. This module is crucial for breaking down complex, high-dimensional data into its constituent components, making it easier to analyze and process. I worked with a US-based scientific consortium, the Open Health System Laboratory (OHSL), where my work was specifically centered on the development of a open source Digital Twin for Oncology.

For this project, the WD module was illustrated on 3D Brain MRI images to demonstrate its practical applications in medical imaging. The wavelet transformation enabled the extraction of multi-resolution features, helping in the efficient representation and analysis of complex MRI datasets. This approach has significant implications for improving the accuracy and efficiency of neural network models used in the diagnosis and treatment of brain-related conditions. The details of the project, including the complete code and article, can be found here: 3D-Wavelet Decomposition on Brain MRI Data.