Quantitative SPECT-CT and textural analysis for neuro-endocrine tumours

Gemmell, Alastair J. (2026) Quantitative SPECT-CT and textural analysis for neuro-endocrine tumours. PhD thesis, University of Glasgow.

Abstract

Background Quantification is an important emerging technique in nuclear medicine imaging. It may provide additional clinical information to add to visual interpretation, whilst reducing inter-operator variability. Technological developments and commercial solutions have recently increased access to quantitative Single Photon Emission Tomography (SPECT). There are few studies optimising acquisition and reconstruction parameters and only a limited number of clinical applications have been investigated. The growth in imaging of Neuroendocrine Tumours (NETs), both for diagnosis and following molecular radiotherapy, highlights Somatostatin Receptor Imaging (SRI) as an ideal candidate for applying quantification.

Aims The thesis sought to assess the accuracy and longitudinal stability of a commercial quantification package for SPECT, Siemens Broad Quantification ("BroadQuant") for two radionuclides, 99mTc and 177Lu. A further aim was to apply quantification to SPECT imaging of NETs, for the integration of quantification into clinical practice.

Methods A series of experiments were designed to measure the accuracy and reproducibility of each stage of the BroadQuant calibration process. Uncertainties were calculated for each step and combined to provide an overall uncertainty on the calibration factor. A range of Quality Control tests were evaluated to determine how best to monitor the longitudinal stability of the calibration. A phantom that mimicked inhomogeneous lesions was used to identify robust textural analysis features for further study in clinical applications. The use of SPECT quantification for imaging of NETs was investigated for a number of clinical applications. The impact of Somatostatin Analogue medication on quantitative SPECT measurement of 99mTc-EDDA/HYNIC-TOC was studied, as well as whether a Tumour Sink Effect (TSE) affected uptake. The use of quantitative metrics and textural analysis features as an alternative to the visual assessment of Krenning Score in 99mTc-EDDA/HYNICTOC SPECT was examined. Post-therapy SPECT imaging following Peptide Receptor Radionuclide Therapy (PRRT) was analysed, to determine the predictive value of quantitative metrics. Statistical harmonisation methods were applied to quantitative data from two gamma cameras with the aim of expansion of quantitative SPECT to multiple sites for future studies.

Results The calibration process was comprehensively examined, providing a number of recommendations for the user, and estimating an uncertainty in the calibration factor of 1.37% for 99mTc and 2.81% for 177Lu respectively. Longitudinal measurements of calibration accuracy determined that stability could be successfully measured with Shewhart charts monitoring Standardised Uptake Value (SUV), with a suitable stable period for 99mTc calibration assessed as being up to 9 months. For patients on Somatostatin Analogue medication, a reduction in physiological uptake of 99mTc-EDDA/HYNIC-TOC was shown, that was not reflected in pathological uptake. No evidence was found for the presence of a TSE for 99mTc-EDDA/HYNIC-TOC. A quantitative Krenning Score was found to be more robust than visual interpretation of Krenning Score, with a comparable level of accuracy when compared to expert assessment. Following feature reduction, a range of textural analysis features were investigated in the context of post-PRRT SPECT imaging. Sphericity was determined to be predictive both of overall survival and for patient-reported symptom response. Finally, statistical harmonisation using Location-Scale adjustment was found to improve the harmonisation between gamma cameras for multiple quantitative metrics, demonstrating a potential tool for combining data from multiple cameras for clinical studies.

Conclusion This thesis addresses the key aspects of the SPECT quantification process, from the calibration of the gamma camera through to clinical applications. It demonstrates that SPECT quantification can be an accurate and robust tool. The use of quantification and textural analysis has been shown to have positive clinical value when applied to the imaging of neuroendocrine tumours.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Additional Information: Supported by funding from the NHS GG&C staff bursary and the NES HSSE programme.
• Subjects: R Medicine > R Medicine (General); R Medicine > RC Internal medicine > RC0254 Neoplasms. Tumors. Oncology (including Cancer)
• Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Medicine, Dentistry & Nursing
• Supervisor's Name: Small, Dr. Sandy, Brown, Dr. Colin and Ray, Professor Surajit
• Date of Award: 2026
• Depositing User: Theses Team
• Unique ID: glathesis:2026-86259
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 13 Oct 2026 15:22
• Last Modified: 13 Oct 2026 15:32
• Thesis DOI: 10.5525/gla.thesis.86259
• URI: https://theses.gla.ac.uk/id/eprint/86259

Related URLs

• Article DOI
• Article DOI
• Article DOI