Caredda, Charly;
Ezhov, Ivan;
Sdika, Michaël;
Lange, Frederic;
Giannoni, Luca;
Tachtsidis, Ilias;
Montcel, Bruno;
(2024)
Pixel-wise and real-time estimation of optical mean path length using deep learning: application for intraoperative functional brain mapping.
In: Elson, Daniel S and Gioux, Sylvain and Pogue, Brian W, (eds.)
Proceedings Volume 13009:Clinical Biophotonics III.
(pp. 130090F-1-130090F-4).
SPIE: Bellingham, WA, USA.
Preview |
Text
130090F.pdf - Published Version Download (4MB) | Preview |
Abstract
Optical imaging is a non-invasive technique that is able to monitor hemodynamic and metabolic brain response following neuronal activation during neurosurgery. However, it still lacks robustness to be used as a clinical standard. In particular, the quantification o f t he b iomarkers o f b rain f unctionality n eeds t o b e i mproved. The quantification r elies o n t he m odified Be er La mbert la w, wh ich ne eds a co rrect es timation of th e op tical mean path length of traveled photons. Monte Carlo simulations are used for estimating the optical path length, but it is time-consuming, especially when modeling a patient’s brain cortex. In this study, we developed a neural network based on the UNET architecture for a pixel-wise and real-time estimation of optical mean path length. The neural network was trained with segmentation of brain cortex as input and mean path length data as target. This deep learning approach allows a real time estimation of the optical mean path length. The results can be beneficial a nd u seful w ithin t he f ramework o f o ur E U-funded H yperProbe p roject, w hich a ims a t transforming neuronavigation during glioma resection using novel hyperspectral imaging technology.
| Type: | Proceedings paper |
|---|---|
| Title: | Pixel-wise and real-time estimation of optical mean path length using deep learning: application for intraoperative functional brain mapping |
| Event: | SPIE Photonics Europe, 2024 |
| Location: | Strasbourg, France |
| Dates: | 7 Apr 2024 - 9 Apr 2024 |
| ISBN-13: | 9781510673366 |
| Open access status: | An open access version is available from UCL Discovery |
| DOI: | 10.1117/12.3016632 |
| Publisher version: | http://dx.doi.org/10.1117/12.3016632 |
| Language: | English |
| Additional information: | This version is the version of record. For information on re-use, please refer to the publisher’s terms and conditions. |
| Keywords: | Monte Carlo simulations, mean path length, Deep-learning, Digital instrument simular, optical imaging |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Med Phys and Biomedical Eng |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10197083 |
Loading...
Loading...Loading...Loading...Archive Staff Only
 |
View Item |
