Towards a Pseudo-Labeling Workflow for Celltype-Classification from Explanted Brain Slice Recordings
arXiv:2607.06569v1 Announce Type: cross Abstract: This paper proposes an unsupervised workflow to pseudo-label extracellular spikes from human brain slice MEA recordings into two putative cell types: pyramidal cells and interneurons. Here, the raw data from the data acquisition system is used and processed. The pipeline for pre-processing includes bandpass filtering, threshold--based spike detection, frame alignment and normalization. In the ML workflow, dimensionality reduction (PCA, t-SNE, UMAP), clustering (GMM, k-means). To achieve an online system, template matching and OSort under varyin
Advances in AI, particularly machine learning techniques like pseudo-labeling, are enabling more sophisticated analysis of complex biological data, making workflows like this more feasible and robust.
This development allows for improved understanding and classification of neural cell types from human brain activity, crucial for foundational neuroscience research, drug discovery, and neurological disorder treatments.
The ability to automatically and robustly pseudo-label cell types in explanted human brain slices accelerates research by reducing manual annotation efforts and improving data consistency in neural circuit analysis.
- · Neuroscience researchers
- · Pharmaceutical companies (neurological)
- · AI/ML healthcare solution providers
- · Biomedical instrumentation companies
- · Manual data annotation services (long-term)
- · Less efficient traditional cell-typing methods
More accurate and faster identification of specific neuronal cell types in human brain research.
Accelerated understanding of neural network function and dysfunction in diseases like epilepsy or Alzheimer's.
Potential for AI-driven personalized medicine approaches to neurological disorders based on precise cell-type targeting.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG