 |
|
 |
|
| Scientific Publications |  |
| News and Events |  |
| Conferences | |
| Workshops | |
| Request Info | |
| Newsletter | |
| Corporate | |
| Contact | |
 |
|
|
|
| Research Products | |
| Clinical Products | |
| Research | |
| BEL | |
|
|
|
|
| Clinical EEG Overview | |
|
|
|
|
| Support | |
|
|
|
|
| Home | |
|
|
|
|
 |
GPS Overview |
 |
Patient Positioning |
 |
Photogrammetry Process |
 |
Model Window and Tools |
GPS Overview
|
GPS is based on the concepts of positioning, acquisition, and marking and modeling. Position your patient in the center of the GPS dome and view the images in Net Station to ensure good visibility of sensors. After acquiring the images, mark and model the sensor positions on the images, and export the data for use with a source-localization software program. |
| |
Patient Positioning
|
Positioning the patient involves centering him or her vertically and horizontally within the GPS dome. Pictured above is a top-down view (nose-up) of the dome, providing a sense of the different lines of sight for the different cameras. In general, pay special attention to the patient's position in regard to cameras in the back (4, 5, and 9). |
| |
Photogrammetry Process
|
GPS requires that a sensor be visible in at least two cameras for its position to be determined. The GPS software takes the 2D user marks that you make on the acquired images, determines the sensors' correct ID numbers (according to the built-in GSN sensor layout map), and uses triangulation to create a 3D model of the GSN. The software projects those 3D points back through the "cameras" (applying the camera calibration values to them), and displays them as points in a 3D Point Cloud model in the camera views. The differences between the 2D user marks and the 3D Point Cloud marks are indicative of the accuracy of your reconstructed sensor model. |
| |
Model Window and Tools
|
With so many sensors and cameras involved, the result is a large number of lines of sight. Because of this, the GPS software tends to be conservative in what it considers a "solved" sensor position. Therefore, even if you mark sensors in the Model Window (a diagram of it is shown above) very carefully, you will probably have to use the Model Window tools to solve some "unsolved" positions manually. The Sensor Registration Images webpage shows many of these tools displayed in different camera views. |
