Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A novel method to compute the contact surface area between an organ and cancer tissue

Alessandra Bulanti, Alessandro Carfì, Paolo Traverso, Carlo Terrone, Fulvio Mastrogiovanni

TL;DR

A novel approach using 3D reconstructions from computed tomography (CT) scans to provide an accurate CSA estimate using a segmentation protocol and an algorithm that processes reconstructed meshes, demonstrating its reliability and consistency.

Abstract

With "contact surface area" (CSA) we refers to the area of contact between a tumor and an organ. This indicator has been identified as a predictive factor for surgical peri-operative parameters, particularly in the context of kidney cancer. However, state-of-the-art algorithms for computing the CSA rely on assumptions about the tumor shape and require manual human annotation. In this study, we introduce an innovative method that relies on 3D reconstructions of tumors and organs to provide an accurate and objective estimate of the CSA. Our approach consists of a segmentation protocol for reconstructing organs and tumors from Computed Tomography (CT) images and an algorithm leveraging the reconstructed meshes to compute the CSA. With the aim to contributing to the literature with replicable results, we provide an open-source implementation of our algorithm, along with an easy-to-use graphical user interface to support its adoption and widespread use. We evaluated the accuracy of our method using both a synthetic dataset and reconstructions of 87 real tumor-organ pairs.

A novel method to compute the contact surface area between an organ and cancer tissue

TL;DR

A novel approach using 3D reconstructions from computed tomography (CT) scans to provide an accurate CSA estimate using a segmentation protocol and an algorithm that processes reconstructed meshes, demonstrating its reliability and consistency.

Abstract

With "contact surface area" (CSA) we refers to the area of contact between a tumor and an organ. This indicator has been identified as a predictive factor for surgical peri-operative parameters, particularly in the context of kidney cancer. However, state-of-the-art algorithms for computing the CSA rely on assumptions about the tumor shape and require manual human annotation. In this study, we introduce an innovative method that relies on 3D reconstructions of tumors and organs to provide an accurate and objective estimate of the CSA. Our approach consists of a segmentation protocol for reconstructing organs and tumors from Computed Tomography (CT) images and an algorithm leveraging the reconstructed meshes to compute the CSA. With the aim to contributing to the literature with replicable results, we provide an open-source implementation of our algorithm, along with an easy-to-use graphical user interface to support its adoption and widespread use. We evaluated the accuracy of our method using both a synthetic dataset and reconstructions of 87 real tumor-organ pairs.
Paper Structure (12 sections, 6 equations, 10 figures)

This paper contains 12 sections, 6 equations, 10 figures.

Table of Contents

  1. Introduction
  2. Segmentation Protocol
  3. Algorithm
  4. Centroids
  5. Centroid-to-Centroid distance
  6. Threshold
  7. CSA definition
  8. CSA refinment
  9. CSA computation
  10. Evaluation
  11. Conclusions
  12. Acknowledgment

Figures (10)

  • Figure 1: 3D reconstruction of a kidney, in yellow, and a tumor, in purple.
  • Figure 2: It can be observed that as the lower bound varies, the accuracy of the reconstruction differs. In particular, moving from left to right, there is a decreasing of the lower bound, resulting in a more defined and precise reconstruction of the kidney.
  • Figure 3: On the left hand side, the selection of the region of interest in the three planes(axial, coronal and sagittal) on the bases of the application of a threshold. On the right hand side, top, the 3D reconstruction after the threshold application. On the right hand side, bottom, the 3D reconstruction after manual cleanup.
  • Figure 4: On the left is the automated 3D interpolation of the tumor, based on the manually drawn silhouette in the three planes. On the right is the 3D reconstruction of the tumor.
  • Figure 5: Here is an example highlighting the importance of outlining the correct perimeter of the tumor. In the left image, the reconstructed tumor exhibits a missing part, leading to a gap between the tumor and the kidney. The right image depicts the tumor after the missing part has been manually added.
  • ...and 5 more figures