{"id":101,"date":"2016-09-28T14:10:19","date_gmt":"2016-09-28T12:10:19","guid":{"rendered":"http:\/\/thomaskosch.com\/?p=101"},"modified":"2016-09-28T14:37:20","modified_gmt":"2016-09-28T12:37:20","slug":"visualizing-real-time-brain-localization-in-3d","status":"publish","type":"post","link":"https:\/\/thomaskosch.com\/index.php\/2016\/09\/28\/visualizing-real-time-brain-localization-in-3d\/","title":{"rendered":"Visualizing Real-Time Brain Localization in 3D"},"content":{"rendered":"
\"10-20
Figure 1: 10-20 electrode placement system [4].<\/figcaption><\/figure>\n

Recently, measuring brain activity (known as electroencephalography, short EEG) to enhance or analyze\u00a0the experience\u00a0of user interfaces has attracted great attention. Besides of medical applications, several researchers have investigated effort\u00a0to make use of it in the area of human-computer interaction, resulting in several projects and publications ([1, 2, 3], just to name a few).\u00a0To measure brain\u00a0activity at a specific\u00a0spot on a head, an electrode is\u00a0attached to that place. Measurement reveal a voltage, showing\u00a0the difference between the measured voltage on the head and a reference electrode. The reference electrode is placed on an arbitrary place on the body (usually the ear lobe) and is necessary to calculate the voltage drop between measurements received from the electrode and the reference electrode. Electrode spots are also denoted with unique names and usually follow a system (see Figure 1).<\/p>\n

\"3D
Figure 2: 3D printed Ultracortex headset.<\/figcaption><\/figure>\n

This gives us solely measured brain activity at certain spots, but does not reveal the localization of the generated electrical source. Several algorithms exist which solve this problem. One famous algorithm is the sLORETA algorithm [5], which\u00a0accepts the measured voltages\u00a0from electrodes and calculates source localization dependent on the chosen model. Knowing\u00a0the source of measured voltages can be used to track diseases such as Alzheimer, ADHS, or origins of depressions.<\/p>\n

During a research project [6], the sLORETA algorithm\u00a0was implemented using an interactive 3D voxel-based visualization to make brain source localization visible. We used a heatmap-like visualization to make it understandable as possible.\u00a0After running simulations of electrical activity and comparing these to our algorithm, we\u00a0are able to verify the correctness of our implementation. The computation\u00a0delivers a magnitude for each voxel, being responsible to visualize the electrical strength using a\u00a0color code.<\/p>\n

We used an 16 channel OpenBCI [7]\u00a0together with an Ultracortex headset (see Figure 2) to run a pilot study in the wild while delivering stimuli in virtual reality. Implementation and visualization was integrated into the neuromore Studio platform [8], which provides support for recording EEG data and signal processing for post hoc analysis purposes. Results\u00a0showed a working visualization to see where brain activity is generated.\u00a0Red-colored areas show a high probability, that this area is responsible for measured voltages at different electrodes. However, blue areas denote a low probability. Changes in brain activation can be observed in real-time, making it easier to find responsible brain areas for certain input stimuli:<\/p>\n