This project is a part of a Master Thesis in Interaction Technology and Design. The application is the frontend for viewing and translating NIfTI files (medical images) from one domain to another (CT to PET), done on a backend server (HTTP). The application works as a standalone without the server if one dont want to translate files.

Server is located in https://github.com/FraggeGaming/ThesisInferenceServer/tree/main

This is a Kotlin Multiplatform project targeting Desktop (Kotlin 2.1.0) (Compose 1.8.1) (Gradle 8.9) (Java 17).

• /composeApp is for code that will be shared across your Compose Multiplatform applications. It contains several subfolders:
  • desktopMain is for code that’s for the desktop and where the whole project is structured for. With two main screens in the subfolder Screens
  • UploadData is the first entrypoint of the applicaiton, where the user uploads the prefered NIfTI and chooses a translation model
  • MainView is for viewing the images. All input and generated images can be exported in the main view to a ZIP to view the translated files in a prefered external tool

Certain folders contain executable python script to parse the NIfTI into

• external has three executable scripts, one for Windows, one for macOS, and one for Linux and use the Nibabel library to parse NIfTI files into NumPy arrays.
  • When the model has been running, three folders, and two files will appear, input_gz, output_gz, output_npy, and the files config.properties and SavedMappings.txt
  • input_gz Has the user uploaded NIfTI files.
  • output_gz has the AI translated NIfTI files.
  • output_npy has the parsed NumPy from the input and output NIfTI files.
  • config.properties is where the serverIP is located and can be changed.
  • SavedMappings.txt has the information required to store previously translated or uploaded files for easy tracking. When files gets deleted in the UI, they also get deleted in the input/output directories, aswell as in the SavedMappings.txt.

To run the project in intellij:

• Go to Settings - Build, Execution, Deployment - Build Tools - Gradle: and set Gradle JVM to Java 17
• Make sure that the project also has an compatible SDK. Go to Settings - Project Structure - Project: Set SDK and Language Level to >= 17.
• When starting the application for the first time, the executable (external) scripts might be flagged by the OS. In order to use the application, allow the scripts to run if flagged.

• To run the image-to-image translation the app must be running aswell as the server.

• Go to the (external) folder - Locate (config.properties) - If there is no such file, start the application one time for it to appear. There exists an server IP which one can change to the disired path.

Learn more about Kotlin Multiplatform…

License

This project is licensed under the GNU General Public License v3.0 (GPLv3).
You are free to use, modify, and distribute this software under the same license terms.
See the LICENSE file for full license text and conditions.

© 2025 [Fardis Nazemroaya Sedeh]