Detection of axonal synapses in 3D two-photon images.

摘要: Studies of structural plasticity in the brain often require detection and analysis axonal synapses (boutons). To date, bouton has been largely manual or semi-automated, relying on a step that traces axons before boutons. If tracing axon fails, accuracy is compromised. In this paper, we propose new algorithm does not to detect boutons 3D two-photon images taken from mouse cortex. find most appropriate techniques for task, compared several well-known algorithms interest point feature descriptor generation. The final proposed following main steps: (1) Laplacian Gaussian (LoG) based enhancement module accentuate appearance boutons; (2) Speeded Up Robust Features (SURF) detector candidate locations extraction; (3) non-maximum suppression eliminate candidates were detected more than once same local region; (4) generation descriptors Gabor filters; (5) Support Vector Machine (SVM) classifier, trained features labelled data, was used distinguish between non-bouton candidates. We found our method achieved Recall 95%, Precision 76%, F1 score 84% within dataset make available accessing detection. On average, significantly better current state-of-the-art method, while different. conclusion, article demonstrate approach, which independent tracing, can high level accuracy, improves performance existing approaches. data code (with an easy use GUI) are open source repositories.