Usage

Introduction

Welcome to our toolbox usage guide! This document offers a roadmap on how to utilise the software, focusing on the primary tasks: data preparation, target localisation, and custom model training. Each step is detailed below.

Prepare Your Data

To localise a surgical target, you need to generate connectivity features using tractography from your diffusion MRI data. This section offers a step-by-step guide to help you accomplish that. If you’ve already prepared your connectivity data and are poised to train or utilise a model, you can skip ahead to Localise a Target (Predict Mode).

Localise a Target (Predict Mode)

If you haven’t generated tract-density maps (i.e., connectivity features) required to run the localise model, please follow the steps in Prepare Your Data, otherwise you can proceed with the following steps to either localise a surgical target using our pre-trained model, or a custom model tailored for your own data.

Steps to Localise a Target:

Prerequisites: Before starting the localisation, ensure you have the required setup and data. This step outlines the essential preconditions.
Prediction: Utilise the ‘predict’ mode for localisation. Depending on your needs, you can use the toolbox’s pre-trained model or a custom one you’ve trained.

Train a Custom Model (Train Mode)

Our toolbox comes with pre-trained models. However, there may be instances where these models aren’t a perfect fit for your data or the specific structure you’re aiming to localise. In such cases, our software allows you to train a custom model tailored to your needs.

Steps to Train a Custom Model:

Prerequisites: Before diving into model training, understand the requirements and ensure you have all the necessary data and configurations in place.
Training: Dive deep into the training procedure, setting parameters and training the model on your dataset. This tailored model can then be used for precise target localisation.