Protocols

Device used:

Software used:

• Epiphan Capture Tool (Windows version 3.32.4.2 64 bit): https://www.epiphan.com/support/dvi2usb-3-0-software-documentation/
• Plus Applications (2.8.0.20191105-Win64.exe): https://plustoolkit.github.io/download?_ga=2.48602278.16905868.1677761357-1650690234.1675951087
• 3D slicer (Windows, MacOS, Linux version 5.2.2): https://download.slicer.org/
• Slicer extensions to be installed in 3D slicer: SlicerIGT, SlicerIGSIO, SlicerOpenIGTLink
• NDI toolbox: https://support.ndigital.com/s/
• Fusion 360 (V.2.0.15299): https://www.autodesk.com/products/fusion-360/overview?term=1-YEAR&tab=subscription

Procedures

1) Polaris (Motion capture device) Calibration:

• Download the NDI Toolbox
• In the package there are several softwares such as: NDI Track, NDI Capture, USB Driver etc.
• Use NDI Track to check if the Polaris system is connected to the laptop
• Close all the softwares before doing the temporal calibration because it either can be connected to Plus Server or NDI software

2) Ultrasound Video:

• Start the ultrasound machine
• Plug the DVI out to the ultrasound machine
• Epiphan Software: start Epiphan
• When it is connected to the laptop successfully, the light of the DVI 2USB turning to green and we have the image of the ultrasound

3) Temporal Calibration:
3.1) Set the experiment:

• Fill the water bath and attach the tool to the probe
• Start the Polaris system
• Make sure we have image from both devices
• Put the probe with the markers and the stylus in a position that Polaris can detect both signals

3.2) When you download the PlusApp package from the Plus Toolkit website, it contains of several softwares such as free-hand calibration (fCal), Plus Server Launcher etc.

• fCal software: start the fCal
• Select our config file which we recreate (In the GitHub link): "Polaris_Epiphan.xml"
• The default config file is "NDI Polaris tracker with passive markers" which contains the reference but we don't need a reference. So we create another config file without reference that also can collect data from multiple devices. This config file can collect the data from both ultrasound device and Polaris system
• Add the config file (*Attention that it only works for Polaris and Epiphan, not other motion capture devices or other DVI software)
• Click on "connect"
• Pay attention to both signals: signals from ProbeToTracker and StylusTipToTracker. They both must be "ok"
• Go to the tab "temporal calibration"
• In the "Moving" section, make sure that it is on ProbeToTracker option
• Click on "start"
• Rapidly move the ultrasound probe from top to bottom, and keep it vertical in the water
• When it's finished, click on "show plots" to see the quality of the calibration

4) Spatial (Probe) Calibration:

• Start "plus server" launcher
• Use the same config file (combination of both device): "Polaris_Epiphan.xml"
• Do not quit Plus Server
• Start 3D slicer
• Choose the module "IGT" and "OpenIGTLinkIF"
• Click on +
• Check 'status'. It must be Active and the status indicator shows 'On'. It means that 3D Slicer is correctly connected to Plus Server
• Click on the left arrow of 'IGTLConnector'. Then click on the left arrow of 'IN'
• First Image-Image is for ultrasound image so make sure the eye is open and choose Image-Image on the top left, so we have ultrasound image
• Put the probe with markers and stylus in a way that Polaris can detect, so now we have another part in 'Name section' which is StylusTipToProbe. And when you move the stylus, we can see the needle move in the image top right
• Choose the module "Sequences"
• On 'Sequence browser' section, create new sequence as 'calib_spatiale'
• For sequences: first one should be Image-Image and second StylusTipToProbe
• Click on the red circle to record the real- time video of ultrasound
• Put the probe and stylus in the water
• We should see the tip of stylus clearly on the ultrasound image, at least four times at each corner of the image
• Then rotate both stylus and probe at least 45 degree and do it again
• Once it's finished, stop the recording

• Choose the module 'Fiducial Registration wizard'
• In the 'From fiducials' section, choose 'create new markups fiducial as...' and rename it to 'ImageFiducials'
• In the 'To fiducials' section, choose 'create new markups fiducial as...' and rename it to 'ProbeFiducials'
• In the 'Place fiducials using transforms' section, make sure that first one is 'none' and second one is 'StylusTipToProbe'
• In the 'registration results' section, choose 'create new transform as...' and rename it to 'ImageToProbe'
• Select 'similarity' and 'manual update'
• Play the spatial calibration sequence'
• When you see the stylus tip on the image, press 'pause'
• Click on the icon with the arrow upward in the 'From fiducials' section and click on the stylus tip in the image
• Then click on 'Place To' for every fiducial that you add
• Once you finish marking fiducials, click on "Update" to get RSME error (should be in 1 to 3)
• Save all the elements of the scene

4) Volume Reconstruction:

• Start Plus Server
• Use this config file: "Epiphan_Polaris_Acq.xml"
• Start 3D slicer
• Upload the data from spatial calibration
• Choose the module "IGT" and "OpenIGTLinkIF"
• Click on +
• Check 'status'. It must be Active and the status indicator shows 'On'. It means that 3D Slicer is correctly connected to Plus Server
• Click on the left arrow of 'IGTLConnector'. Then click on the left arrow of 'IN'
• Make sure that they eyes of Image-Image and ProbeToTracker are open
• In the 'Data' section, you can check 'Subject hierarchy' 'Transform Hierarchy' and 'All nodes'
• In the 'Transform hierarchy' you need to set the hierarchy like this:

• We need to put Image-Image under ImageToProbe in order to image follow the movements of the probe, and put both of them under ProbeToTracker so that the motion capture device can track both (*Tip: If it didn't work like this, try to create a transform manually with the exact matrix of ImageToProbe transform and then put Image-Image under this Manual Transform)
• Go to 'Volume Reslice Driver' module
• Choose 'Image-Image' for Driver section and 'Transverse' for Mode section (Both in red area)
• Now we also see Image in 3D window and the image will follow the movement of probe
• Go to 'Volume Reconstruction' module
• Input method: Live volume reconstruction (if you're not doing live reconstruction, choose Recorded sequence)
• Input volume node: choose 'Image-Image'
• Output volume node: choose create new volume
• ROI node: choose create new Annotation ROI
• Interpolation mode: choose 'Linear'
• Compounding mode: choose 'Mean'
• The ultrasound image includes all the writings, marks etc from the screen hence we need to crop the image to make sure we only reconstruct the image
• Put the mouse in point A and find the coordinates of it in the bottom left of the screen (1)
• Then put the mouse in point B and based on the coordinate of these two points, calculate the size of the image (2)
• Click on start
• Do the scan
• We are scanning a ruler so the probe is vertical inside the water, you start from top of the ruler and scan it till you reach the bottom of it, then stop recording
• Go back to module 'Data' and 'Subject hierarchy'
• Click on the eye of the 'volume' to see it in 3D
• Go to 'Volume Rendering' module
• Choose volume in 'Volume' section and make sure the eye is open
• The preset 'MR-Default' is good for visualisation
• You can change the shift value to improve the rendering
• To only see the volume reconstructed, you can close the eye in red section
• Save your data

4) Segmentation:

• Strat '3D Slicer'
• Upload your data from volume reconstruction
• Go to the module 'Segment editor'
• In segmentation section: create a new segmentation
• In master volume: choose the reconstructed volume
• Click on 'Add' to add a segment
• Choose 'paint' icon in the Effects section. You can change the diameter of the brush
• The segmentation should be done in one of the plans: 'Axial' 'Sagittal' or 'Coronal'
• Draw the volume outline with the mouse (click and maintain). Then move to next volume and to it again
• Click on 'Show 3D' to see the segmentation in 3D view
• Once you're finished, choose the 'Fill between slices' effect
• Click on 'Initialise' and 'Show 3D' to get a preview
• Click on 'Apply'
• Choose the module 'Segment statistics'
• For 'Segmentation' choose the segmentation
• For 'Scalar volume' choose the volume
• For 'Output table' create a new output table
• In 'Advanced' section, only choose 'Scalar volume statistics' and click on 'options' in front of it
• Click on 'Apply' to get the volume statistics