Probabilistic Fluorescence-Based Synapse Detection

Anish K. Simhal,Cecilia Aguerrebere,Forrest Collman,Joshua T. Vogelstein,Kristina D. Micheva,Richard J. Weinberg,Stephen J. Smith,Guillermo Sapiro
DOI: https://doi.org/10.1371/journal.pcbi.1005493
2016-11-17
Abstract:Brain function results from communication between neurons connected by complex synaptic networks. Synapses are themselves highly complex and diverse signaling machines, containing protein products of hundreds of different genes, some in hundreds of copies, arranged in precise lattice at each individual synapse. Synapses are fundamental not only to synaptic network function but also to network development, adaptation, and memory. In addition, abnormalities of synapse numbers or molecular components are implicated in most mental and neurological disorders. Despite their obvious importance, mammalian synapse populations have so far resisted detailed quantitative study. In human brains and most animal nervous systems, synapses are very small and very densely packed: there are approximately 1 billion synapses per cubic millimeter of human cortex. This volumetric density poses very substantial challenges to proteometric analysis at the critical level of the individual synapse. The present work describes new probabilistic image analysis methods for single-synapse analysis of synapse populations in both animal and human brains.
Computer Vision and Pattern Recognition,Neurons and Cognition
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: how to detect and analyze individual synapses efficiently and accurately through fluorescence microscope (FM) image data, especially the molecular characteristics of large - scale synaptic populations. Specifically, the researchers hope to develop a new probabilistic fluorescence microscope synaptic detection method to overcome the limitations in existing technologies. ### Background and Problems 1. **Limitations of Existing Technologies** - **Electron Microscope (EM)**: Although EM is the gold standard for identifying synapses, with nanometer - level resolution and can clearly show synaptic structural features (such as synaptic vesicles, synaptic clefts, and postsynaptic density regions), it has limited molecular discrimination ability, and is complex to operate, slow, and costly. - **Fluorescence Microscope (FM)**: FM has excellent molecular resolution ability, high imaging speed, and low cost, but it has challenges in distinguishing synaptic signals from non - synaptic background signals, especially in high - density synaptic regions. 2. **Research Objectives** - Develop a probabilistic synaptic detection algorithm based on multi - channel fluorescence microscope (muxFM) data. - This algorithm can generate a probability map of each voxel belonging to a synapse and can reproduce the known cortical synaptic distribution characteristics. - Verify the effectiveness of the algorithm by comparing it with human - annotated conjugate array tomography (cAT) EM data. - Provide a method that does not rely on EM cross - validation and is suitable for the analysis of standard confocal and super - resolution FM images. ### Solutions The researchers proposed an unsupervised probabilistic synaptic detection method, which mainly includes the following steps: 1. **Foreground Probability Calculation** - Calculate the probability \( p_B(x, y, z) \) of each voxel belonging to background noise, and then derive the probability \( p_F(x, y, z)=1 - p_B(x, y, z) \) of it belonging to the foreground (i.e., potential synapses). 2. **Two - Dimensional Blob Probability Calculation** - Process the foreground pixels and calculate the probability \( p_P(x, y, z) \) of them forming two - dimensional blobs, which is achieved by the product of pixels in the neighborhood. 3. **Three - Dimensional Blob Probability Calculation** - Consider blobs across multiple slices, introduce a factor \( f(x, y, z) \) to reduce the probability of blobs that do not span multiple slices, and finally get the three - dimensional blob probability \( p_{3DP}(x, y, z)=p_P(x, y, z)\cdot f(x, y, z) \). 4. **Pre - and Post - Synaptic Blob Adjacency Evaluation** - Detect the adjacency of presynaptic and postsynaptic markers, find high - probability presynaptic signals by the grid search method, and combine them with postsynaptic signals to generate the final synaptic probability map \( p_{synap}(x, y, z)=p_{pres}(x, y, z)\cdot p_{post}(x, y, z) \). ### Experimental Verification This method was verified on the conjugate array tomography (cAT) dataset. The results show that this algorithm can successfully detect synapses from muxFM data, with performance close to the human - annotation level, and can reproduce the known cortical synaptic distribution characteristics. ### Summary This research provides a novel probabilistic synaptic detection method, which can analyze large - scale synaptic populations efficiently and accurately without EM cross - validation, providing a powerful tool for in - depth understanding of brain synaptic networks.