Artikel
Identification of active sensory cortex areas during neurosurgical procedures using standard RGB camera equipment for optical imaging – a feasibility study
Suche in Medline nach
Autoren
Veröffentlicht: | 2. Juni 2015 |
---|
Gliederung
Text
Objective: Intraoperative Optical Imaging is a neuro-imaging technique for detection and visualization of changes in cortical optical characteristics. One field of application is the identification of the sensory cortex area during neurosurgical procedures. Therefore, up to now, researchers used sophisticated hardware-setups comprising stabilized light sources, highly sensitive CCD cameras und light wavelength filters. Contrary to this approach, this study investigated the feasibility of the method using standard RGB camera equipment that is part of the surgical microscope.
Method: Eight patients undergoing tumor resection near or adjacent to the somatosensory cortex were included in this study after informed given consent. Optical changes of cortical surface during median nerve stimulation of the patients were recorded simultaneously with a highly sensitive monochrome camera (AxioCamMRm) and with a three channel standard RGB camera (Trio 610). Color coded activity maps, representing the corresponding areas of optical changes on the cortical surface, were calculated from each camera data, respectively from separated RGB color channel. Mean power of each activity map in the area of the segmented sensory cortex was calculated in relation to the power standard deviation of each activity map outside the segmented sensory cortex. This parameter represents a spatial signal-to-noise ratio (SNR) which indicates the reliability of the sensory-cortex identification based on each activity map.
Results: Except two cases, the activity maps calculated from AxioCam data offered best performance. SNR of the activity map calculated from AxioCam data was highest over all patients with a value of 6.75 ± 1.95. Activity maps calculated from the Trio 610 camera revealed an individual, channel related SNR (5.11 ± 1.85 red channel data; 5.13 ± 1.55 blue channel data; 4.91 ± 1.38 green channel data). Regarding the Trio 610 data, activity maps calculated from red channel data, offered best performance in four of the eight cases, from green channel in three cases, and from blue channel in one case. Identification of sensory cortex showed significant results in all cases by choosing the best activity map calculated from one of the RGB channels.
Conclusions: Identification of stimulated sensory cortex area during neurosurgical procedure has been established using standard RGB camera equipment. Quality of the activity maps might be further improved by optimizing color data computation algorithms.