Related Experiment Video
Updated: May 30, 2026
08:32
Using Computer-based Image Analysis to Improve Quantification of Lung Metastasis in the 4T1 Breast Cancer Model
Published on: October 2, 2020
An innovative iterative thresholding algorithm for tumour segmentation and volumetric quantification on SPECT images:
M Pacilio1, C Basile, S Shcherbinin
1Department of Medical Physics, Azienda Ospedaliera S. Camillo Forlanini, Rome 00151, Italy. mpacilio@scamilloforlanini.rm.it
Medical Physics
|August 6, 2011
Summary
A new iterative thresholding method (RIThM) improves tumor segmentation accuracy in single-photon emission computed tomography (SPECT) imaging. This method, validated with Monte Carlo simulations, offers a robust and efficient approach for clinical applications.
Area of Science:
- Medical Imaging
- Nuclear Medicine
- Computational Imaging
Background:
- Accurate tumor segmentation in Positron Emission Tomography (PET) and Single-Photon Emission Computed Tomography (SPECT) is crucial but challenging.
- Existing methods often struggle with reproducibility and precision in delineating functioning organ or tumor volumes.
Purpose of the Study:
- To introduce and implement an innovative iterative thresholding method, Recovering Iterative Thresholding Method (RIThM), for enhanced tumor segmentation in SPECT imaging.
- To investigate the utility of Monte Carlo (MC) simulations for system calibration within the RIThM framework.
- To assess the accuracy and robustness of RIThM in comparison to experimental data and MC simulations.
Main Methods:
- RIThM is an iterative MATLAB-based algorithm that refines volume estimates by incorporating system recovery coefficients (RC) and experimental threshold-volume calibrations.
- The algorithm iteratively corrects the source-to-background ratio (SBR) and adjusts image thresholding until convergence.
- Monte Carlo (MC) codes (SIMIND, SimSet) were used for system calibration, and the method was validated using phantoms, test objects, and clinical patient data (brain metastases, gliomas).
Main Results:
- MC simulations showed good agreement with experimental calibration data, with differences within experimental errors for thresholds and RCs.
- RIThM achieved volume determination accuracies between -9% and 15% for simulated hot spheres (4-270 ml) and 9%-18% for the Zubal head phantom.
- Experimental test images demonstrated accuracy within +/- 10% for volumes between 20-110 ml, and preliminary clinical application showed suitability.
Conclusions:
- The proposed RIThM method, guided by MC simulations, offers a robust, accurate, and fast solution for tumor volume delineation in SPECT imaging.
- MC-guided calibration can potentially reduce experimental acquisition time.
- Further research is ongoing to address challenges with irregularly shaped tumors and non-uniform background activity.
