Centre for In Vivo Imaging

Staff Profiles

Gemma Roberts

Alzheimer's Society Clinical Fellow


I am a nuclear medicine physicist with clinical experience within the NHS.  I am currently working on a PhD within IoN funded by the Alzheimer's Society's Clinician and Healthcare Professional Training Fellowship.  The project aims to improve and develop the ways that quantitative nuclear medicine imaging can be used to assist in the early diagnosis of dementia with Lewy Bodies.

My main supervisor is Prof. Alan Thomas and my supervisors in Medical Physics within Newcastle upon Tyne NHS Trust are Dr. Jim Lloyd and Dr. Andrew Sims.

Further information can be found on the Alzheimer's Society website:

Alzheimer's Society web profile

Alternate email: gemma.roberts4@nhs.net

Google Scholar: Click here.


Improving the early diagnosis of Dementia with Lewy Bodies using quantitative assessment of nuclear medicine scans 

Aims:Early accurate diagnosis of Dementia with Lewy Bodies (DLB) is required for both patient management and clinical trials. This project aims to improve two of the most promising imaging tests for prodromal diagnosis: 123I-FP-CIT and cardiac 123I-MIBG. Both have very good accuracy for established DLB but neither has been studied in the prodromal phase (pDLB). Pathology studies have suggested that cardiac MIBG has the potential to be a better biomarker for pDLB than FP-CIT. However, cardiac MIBG scans are analysed using planar images and subject to variation due to technical factors, which may make the usual methods inadequate for early diagnosis.

Methods:We will evaluate the use of attenuation and scatter corrected SPECT/CT images to improve the accuracy of cardiac MIBG quantification. This will be investigated using physical phantoms and Monte Carlo simulations of phantoms created from patient data. Similar methods will be evaluated for FP-CIT. Absolute quantificationmethods will be compared to semi-quantitative uptake ratios. The optimised methods will be applied to scans of subjects taking part in a longitudinal research study at our centre and compared to standard analysis methods.

Expected outcomes:We will demonstrate the potential variation in cardiac MIBG and FPCIT results obtained via standard methods and the degree to which the use of corrected SPECT/CT data and absolute quantification can minimise this variation. We will estimate mean uptake values and variances for pDLB and non-pDLB subjects with standard and optimised methods. This will inform future power calculations for clinical studies comparing the two scans.

Further information on the project aimed at members of the public can be found on the Alzheimer's Society website: