vix.ing · top · new · best · stats · spec

Hierarchical Manifold Clustering on Diffusion Maps for Connectomics (MIT\n 18.S096 final project)

2016/07/19 by Gergely Ódor, Odor, Gergely
Computer Science · Health Professions · Medicine · #Computer Vision and Pattern Recognition (cs.CV) #Data Management and Algorithms #FOS: Computer and information sciences #Osteoarthritis Treatment and Mechanisms #Rough Sets and Fuzzy Logic #Temporomandibular Joint Disorders #Topological and Geometric Data Analysis

paper · pdf · doi:10.48550/arxiv.1607.06318

openalex publication_date 2016/07/19 · openalex created_date 2022/08/27 · openalex updated_date 2026/07/28

Abstract

In this paper, we introduce a novel algorithm for segmentation of imperfect\nboundary probability maps (BPM) in connectomics. Our algorithm can be a\nconsidered as an extension of spectral clustering. Instead of clustering the\ndiffusion maps with traditional clustering algorithms, we learn the manifold\nand compute an estimate of the minimum normalized cut. We proceed by divide and\nconquer. We also introduce a novel criterion for determining if further splits\nare necessary in a component based on it's topological properties. Our\nalgorithm complements the currently popular agglomeration approaches in\nconnectomics, which overlook the geometrical aspects of this segmentation\nproblem.\n

Citations

Related