3D Digitization and Co-Registration to the MNI brain template using OxySoft

How do we know that the most active channels are located over the brain region of interest and not somewhere else? Of course, an experienced researcher just knows where to place the optodes, but is that enough to convince a potential highly-critical reviewer or fellow scientist? In order to have stronger evidence for the actual spatial location, one needs a way to measure the position of the optodes reliably. Since 2016, this is possible directly from within OxySoft using a Polhemus digitizer. After measuring anatomical landmarks, simply pointing the digitizer at the optodes is sufficient and OxySoft reliably knows where the optode is placed with respect to the human head. OxySoft can visualize this in a 3D view (see Figure 1), allowing not only beautiful picture for a publication, but also having a scientifically sound method to present neuroimaging data. In this article, we will explain how this method works and show how easily it works using OxySoft.

Figure 1:  3D visualization of activation of a Brite23 measurement in OxySoft

Figure 1: 3D visualization of activation of a Brite23 measurement in OxySoft

As every individual head and brain is different, researchers have agreed to use so-called standard brain templates, to which the individual anatomy and functional data is transformed to. Thereby the location of a specific brain region is precisely known, if it is already known from the brain template. The Montreal Neurological Institute (MNI) has created one of these templates, the MNI-152. The International Consortium for Brain Mapping (ICBM) has adopted the MNI-152 brain template as their standard template. This the most used brain template in neuroscience research, especially in the fMRI community. This is also the brain template that is implemented in OxySoft. Using this template, you are thus using a brain that is compatible with the vast majority of neuroscientists worldwide. For more information on the brain templates, see [1].

OxySoft supports the digitizers from Polhemus Inc. (see Figure 2). Polhemus is established as the world’s leading digitization and motion tracking company, which devices are not only used extensively in neuroscience research (see e.g. [2] and [3]), but have for example also been used to digitally archive Star Wars artifacts or during filming of the Lord of the Rings movies (http://polhemus.com/scanning-digitizing/case-studies). The Polhemus digitizer measures the position of sensors in a self-created 3D electromagnetic field. OxySoft uses these 3D coordinates to localize the positon of the optodes with respect to the participant’s head. The tool of choice for digitizing is the stylus. While similar in shape as a pen, it is connected to the Polhemus device. Upon clicking on the button on the stylus, OxySoft gets notified to record the momentarily position.

Figure 2: Polhemus digitizer on a Brite23. The stylus is used to record the position of the optodes. Head movements are automatically co-recorded using the dedicated head motion sensor.

OxySoft is the only fNIRS recording software which has this direct interface between fNIRS optode template (also called probe setup) and actual positions of the optodes with respect to the participant’s head. Of course, OxySoft will automatically coregister the digitized points to the MNI-152 template, allowing you without any effort to report back the actual positions of the optodes in MNI-space.

After having established a connection with the Polhemus digitizer, OxySoft will initially ask to digitize fiducial points. The fiducials are anatomical landmarks, which are used to infer the size and shape of the participant’s head. OxySoft will ask for five fiducial points. the nasion (the valley on your nasal bone), the inion (the small bulge on the back of your head), the two pre-auricular points (just in front of your ears) and the vertex, which is the crossing point of the nasion-inion line and the left and right pre-auricular points. Note that using five points allow for a more precise coregistration, and thus more accurate results than used in most articles, and is far superior to the common 4-point registration (see also [2], [3] and [4]). Note that we of course record and correct for movement of the participant simultaneously using a dedicated head motion sensor. The position of these five fiducial points are also known from the MNI template, so that OxySoft performs the coregistration automatically, i.e. compute the five-point rigid-body transformation.

After some sanity checks whether the digitization was successful, the procedure can seamlessly continue with the digitization of the optodes. As OxySoft knows the optode template that you are using, it specifically asks for the position of each optode specifically (see Figure 3). OxySoft then automatically coregisters the optode position and projects it to the scalp surface of the MNI-tempate. You can thus immediately double-check whether the digitization was successful and accurate. In contrast to 3rd party toolboxes, you thus do not need to match the digitized position with the optode, and you overcome the cumbersome search for errors (and sudden enlightment, that you missed an optode and have to start all over again). With this integrated approach, digitization of for example 100-channels takes less than a minute and no additional work from you. You can thus spend more of your valuable time for answering important neuroscientific research questions.

Figure 3: OxySoft optode digitization screen. Since OxySoft has a direct interface to the Polhemus digitzer and knows the optode template, OxySoft directly performs the coregistration of the digitized positions and presents them during the digitization process to you. This way, it is so easy to verify the output of the coregistration and digitization!

Are you interested in obtaining a Polhemus digitizer and the 3D-plugin for OxySoft? Contact us now! We are happy to provide you with our integrated solution, helping you to get the best out of your research.


[1] Brett, Matthew, Ingrid S. Johnsrude, and Adrian M. Owen. 2002. “The Problem of Functional Localization in the Human Brain.” Nature Reviews Neuroscience 3 (3): 243–49. doi:10.1038/nrn756.

[2] Singh, Archana K., Masako Okamoto, Haruka Dan, Valer Jurcak, and Ippeita Dan. 2005. “Spatial Registration of Multichannel Multi-Subject FNIRS Data to MNI Space without MRI.” NeuroImage 27 (4): 842–51. doi:10.1016/j.neuroimage.2005.05.019.

[3] Whalen, Christopher, Edward L. Maclin, Monica Fabiani, and Gabriele Gratton. 2008. “Validation of a Method for Coregistering Scalp Recording Locations with 3D Structural MR Images.” Human Brain Mapping 29 (11): 1288–1301. doi:10.1002/hbm.20465

[4] Wu, Xue, Adam T. Eggebrecht, Silvina L. Ferradal, Joseph P. Culver, and Hamid Dehghani. 2015. “Evaluation of Rigid Registration Methods for Whole Head Imaging in Diffuse Optical Tomography.” Neurophotonics 2 (3): 035002–035002. doi:10.1117/1.NPh.2.3.035002.