feat(readme): install guide

README.md

@@ -1,17 +1,37 @@
-# SlicerEISeg
+# SlicerEISegMed3D
 
-3D slicer extension for interactive medical image segmentation
+A 3D slicer extension for performing **E**fficient **I**nteractive **Seg**mentation on **Med**ical image in **3D**. Users will guide a deep learning model to perform segmentation by providing positive and negative points.
 
 ## Install
 
-EISegMed3D depends on pypi packages paddlepaddle and paddleseg. You would need to install them before using this plugin.
+If you are just trying out this plugin, there is no prior installation steps requred. You will be prompted to install required packages when the plugin loads.
 
-Open up the python terminal inside 3D Slicer and run the following code.
+To achieve the best performance, it's advised to install the GPU version of paddlepaddle before running this plugin. Steps are as follows:
+
+1. Check your CUDA driver version. Choose the first version that's lower than yours from the followings: 11.6, 11.2, 11.1, 10.2, 10.1
+
+2. Remove the dot in the chosen version, eg: 11.6 -> 116, and replace the [CHOSEN VERSION] part in the following code with it.
 
 ```python
 import sys
 import os
-os.system(f"'{sys.executable}' -m pip install paddlepaddle paddleseg")
+slicer.app.processEvents()
+os.system(f"'{sys.executable}' -m pip install paddlepaddle-gpu==2.3.1.post[CHOSEN VERSION] -f https://www.paddlepaddle.org.cn/whl/linux/mkl/avx/stable.html"))
 ```
 
-You should see the terminal print out 0. Anything else indicates the installation failed.
+For CUDA 11.6, the code you run should be:
+
+```python
+import sys
+import os
+slicer.app.processEvents()
+os.system(f"'{sys.executable}' -m pip install paddlepaddle-gpu==2.3.1.post116 -f https://www.paddlepaddle.org.cn/whl/linux/mkl/avx/stable.html"))
+```
+
+The final line will only show up after installation finshes. If you started 3D Slicer from a terminal, you should be able to see pip install progress there.
+
+The final line should output 0. Anything else indicates the installation has failed.
+
+# Usage
+
+Model parameters: