Accurate Image-guided (Re)Placement of NIRS Probes. (March 2021)

• Record Type: Journal Article
• Authors: Wu, Shin-Ting; Rubianes Silva, José Angel Iván; Novi, Sergio Luiz; de Souza, Nicolas Gabriel S.R.; Forero, Edwin Johan; Mesquita, Rickson C.
• Is Part Of: Computer methods and programs in biomedicine. Volume 200(2021)
• Publication Date: 2021-03
• Subjects: fNIRS -- Probe placement -- Image-guided placement system -- Reproducibility of measures; 00-01 -- 99-00; Medicine -- Computer programs -- Periodicals; Biology -- Computer programs -- Periodicals; Computers -- Periodicals; Medicine -- Periodicals; Médecine -- Logiciels -- Périodiques; Biologie -- Logiciels -- Périodiques; Biology -- Computer programs; Medicine -- Computer programs; Periodicals; Electronic journals; 610.28
• DOI: 10.1016/j.cmpb.2020.105844
• Languages: English
• ISSNs: 0169-2607

Abstract

Highlights: Development of a neuronavigation system using electromagnetic digitizer. Design of a holder of the digitizer's accessories to improve the accuracy of its alignment with a real head. Presentation of a GPU-based rendering of visual objects at interactive rates. Solution for unambiguous depth perception to improve the accuracy of the placement of probes. Proposal of a procedure to improve the reproducibility of digitized points. Assessment of intra-subject fNIRS reproducibility using the proposed neuronavigation system. Abstract: Background and Objective: Functional near-infrared spectroscopy (fNIRS) has become an attractive choice to neuroscience because of its high temporal resolution, ease of use, non-invasiveness, and affordability. With the advent of wearable fNIRS technology, on-the-spot studies of brain function have become viable. However, the lack of within-subject reproducibility is one of the barriers to the full acceptability of fNIRS. To support the validation of the claim that within-subject reproducibility of fNIRS could benefit from accurate anatomical information, we present in this paper a method to develop an image-based system that improves the placement of the sensors on the scalp at interactive rates. Methods: The proposed solution consists of an electromagnetic digitizer and an interactive visualization system that allows monitoring the movements of the digitizer on a real head with respect to the underlying cerebral cortical structures. GPU-based volume raycasting rendering is applied to unveil these structures from the corresponding magnetic resonance imaging volume. Scalp and cortical surface are estimated from the scanned volume to improve depth perception. An alignment algorithm between the real and scanned heads is devised to visually feedback the position of the stylus of the digitizer. Off-screen rendering of the depthmaps of the visible surfaces makes spatial positioning of a 2D interaction pointer possible. Results: We evaluated the alignment accuracy using four to eight anatomical landmarks and found seven to be a good compromise between precision and efficiency. Next, we evaluated reproducibility in positioning five arbitrarily chosen points on three volunteers by four operators over five sessions. In every session, seven anatomical landmarks were applied in the alignment of the real and the scanned head. For the same volunteer, one-way analysis of variance (ANOVA) revealed no significant differences within the five points digitized by the same operator over five sessions ( α = 0.05). In addition, preliminary study of motor cortex activation by right-hand finger tapping showed the potential of our approach to increase functional fNIRS reproducibility. Conclusions: Results of experiments suggest that the enhancement of the visualization of the location of the probes on the scalp, relative to the underlying cortical structures, improves reproducibility of fNIRS measurements. As further work, we plan to study the fNIRS reproducibility in other cortical regions and in clinical settings using the proposed system.