MidSurfer: A Parameter-Free Approach for Mid-Surface Extraction from Segmented Volumetric Data

TL;DR

MidSurfer tackles the absence of a parameter-free method for mid-surface extraction from segmented volumetric data, focusing on thinly bounded biological compartments. It presents a three-module pipeline: Ridge Field Transformation to generate a symmetric ridge height field, Mid-Polyline Extraction to trace per-slice midlines, and Polyline Zipper triangulation to assemble a coherent mid-surface mesh. The approach yields smooth, uniformly triangulated surfaces with preserved holes and higher mesh regularity, outperforming prior methods like screened PSR in several quality metrics, and is released as a ParaView plugin for broad accessibility. The work demonstrates robustness across topologies, enables quantitative surface morphometrics, and supports realistic membrane modeling, facilitating automated, one-click analyses in structural biology workflows.

Abstract

In the field of volumetric data processing and analysis, extracting mid-surfaces from thinly bounded compartments is crucial for tasks such as surface area estimation and accurate modeling of biological structures, yet it has lacked a standardized approach. To bridge this gap, we introduce MidSurfer--a novel parameter-free method for extracting mid-surfaces from segmented volumetric data. Our method produces smooth, uniformly triangulated meshes that accurately capture the structural features of interest. The process begins with the Ridge Field Transformation step that transforms the segmented input data, followed by the Mid-Polyline Extraction Algorithm that works on individual volume slices. Based on the connectivity of components, this step can result in either single or multiple polyline segments that represent the structural features. These segments form a coherent series across the volume, creating a backbone of regularly distributed points on each slice that represents the mid-surface. Subsequently, we employ a Polyline Zipper Algorithm for triangulation that connects these polyline segments across neighboring slices, yielding a detailed triangulated mid-surface mesh. Our findings demonstrate that this method surpasses previous techniques in versatility, simplicity of use, and accuracy. Our approach is now publicly available as a plugin for ParaView, a widely-used multi-platform tool for data analysis and visualization, and can be found at https://github.com/kaust-vislab/MidSurfer .

Figures (13)

• Figure 1: An overview of the overall method: From left to right: Volumetric segmented data represented in slice-wise view, the extracted mid-polylines from the data, the triangular surface generated from mid-polylines, and a use case of the mid-surface with the modeling results visualized over the extracted mid-surface mesh.
• Figure 2: Comparative illustration of (A) isosurfaces, (B) medial surfaces, and (C) mid-surfaces to emphasize the unique advantages of mid-surfaces in capturing the median geometry. The segmentation is shown in light green color and the respective surfaces with a thick black line.
• Figure 3: Schematic overview of the Mid-surface Extraction Algorithm, showing steps from Ridge Field Transformation (\ref{['sec:ridge_field_transformation']}) that transforms the binary segmentation data; through the Mid-Polyline Extraction Algorithm (\ref{['sec:centerLines']} that extracts the mid-polylines from slices; to the final mesh generation using the Polyline Zipper Algorithm (\ref{['sec:meshing']}).
• Figure 4: Visualization of the connected components identification and mid-polyline tracing on a binary mask slice. (A) The original binary mask. (B) The connected components within the binary mask, each assigned a unique ID. (C) Tracing the mid-polyline within these identified connected components. (D) Tracing the mid-polyline on the dilated connected components, avoiding ending the line prematurely around narrow boundaries.
• Figure 5: Visualization of Ridge Field Transformation and Mid-Polyline Extraction Algorithm on a slice: (A) Initial binary segmentation, showcasing the structure of interest. (B) Signed distance field derived from the initial segmentation. (C) Ridge line height field (a smoothed version of the signed distance field to ensure a consistent non-zero derivative across all pixels). (D) Curvature tensor field visualization, derived from the ridge line height field, using superquadric tensor glyphs kindlmann2004superquadric. (E) Visualization of normalized eigenvectors associated with the smallest eigenvalue, represented as a line field. (F) Golden section search optimization: a white polyline depicts a streamline within an eigenvector field drifting from the ridge, while a green line shows the golden section search perpendicular to the eigenvector, ensuring the streamline's alignment with the ridge.