BSBal Sanghera

Bal Sanghera

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Bal Sanghera has a total of 240 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    [18F]F-Prostate-specific Membrane Antigen-1007 and [68Ga] Ga-Prostate-specific Membrane Antigen-11 Positron Emission Tomography in Prostate Cancer: Do Standardized Uptake Value Correlations Dictate Choice?

    by Bal Sanghera, Gerry Lowe, Sophie Sanghera, Wai-Lup Wong

    Objective: To characterize correlations (ρ) between imaging and patient-related parameters in [ 18 F] F-prostate-specific membrane antigen (PSMA)-1007 and [ 68 Ga] Ga-PSMA-11 positron emission tomography-computed tomography (PET-CT) and to determine whether observed differences influence clinical tracer interchangeability. Materials and Methods: Two retrospective PET-CT cohorts from a single-center database were analyzed, comprising patients imaged for suspected prostate cancer with either tracer. Spearman rank intra-tracer correlation coefficients were calculated between weight, body surface area (BSA), lean body mass (LBM), lesion maximum standardized uptake value (SUVmax) (T), liver mean SUV (SUVmean) (B), tumor-to-background ratio (T/B), and metabolic tumor volume. Inter-tracer comparisons were performed using Fisher-transformed paired correlation coefficients, with Bonferroni correction to control type I error. Results: Mann–Whitney testing demonstrated no significant differences in weight ( P = 0.50), BSA ( P = 0.50), or LBM ( P = 0.47) distributions between cohorts. Intra-tracer heatmaps showed broadly similar correlation patterns and significance distributions across tracers. Inter-tracer analysis identified significant differences in correlations between liver SUVmean with weight (ρ = 0.43, P = 0.0001), BSA (ρ = 0.29, P = 0.0001), and LBM (ρ = 0.29, P = 0.0001). These differences were nullified when standard weight-based liver SUVmean normalization for [ 68 Ga] Ga-PSMA-11 was replaced with LBM normalization (weight: ρ = −0.01, P = 0.96; BSA: ρ = −0.04, P = 0.75; LBM: ρ = −0.04, P = 0.75) or BSA normalization (weight: ρ = −0.09, P = 0.49; BSA: ρ = −0.12, P = 0.36; LBM: ρ = −0.12, P = 0.36). Conclusion: Correlation structures for the two tracers were broadly comparable across most imaging and demographic variables. Observed differences in liver SUVmean correlations with body habitus parameters were mitigated by BSA or LBM normalization. These findings support clinical tracer interchangeability. However, full characterization is recommended when alternative SUV normalizations are applied or when liver SUVmean is used as background reference tissue in tumor-to-background trial selection criteria.

  • Open Access

    Quantitative characterization of 18F-PSMA-1007 and [68Ga]Ga-PSMA-11 PET-CT Imaging in Suspected Prostate Cancer: A Single-centre Experience

    by Bal Sanghera, Gerry Lowe, Sophie Sanghera, Wai Lup Wong

    Quantitative characterization of 18F-PSMA-1007 and [68Ga]Ga-PSMA-11 PET-CT Imaging in Suspected Prostate Cancer: A Single-centre Experience - Molecular Imaging and Radionuclide Therapy

  • Open Access

    Tomographic Reconstruction with Real-time a priori Acquisition

    by Muhammad Wishal Khan, Hooman Oroojeni M. J., Bal Sanghera, Tim Blackwell, Mohammad Majid Al-Rifaie


  • Open Access

    Positron emission tomography PET/CT harmonisation study of different clinical PET/CT scanners using commercially available software

    by Gerry Lowe, Bruce Spottiswoode, Jerome Declerck, Keith Sullivan, Mohammad-Saeed Sharif, Wai-Lup Wong, Bal Sanghera

    Harmonisation is the process whereby SUV values from different scanners can be made comparable. This positron emission tomography (PET) - CT pilot study aimed to evaluate the effectiveness of harmonisation of a modern scanner with image reconstruction incorporating resolution recovery (RR) with another vendor older scanner operated in two-dimensional (2D) mode, and for both against a European standard (EARL). The vendor-proprietary software EQ•PET was used, which achieves harmonisation with a gaussian smoothing. A sub study investigated effect of RR on harmonisation. Phantom studies on each scanner were performed to optimise smoothing the parameters required to achieve successful harmonisation. 80 patients were retrospectively selected; half were imaged on each scanner. As proof of principle, a cohort of 10 patients was selected from the modern scanner subjects to study the effects of RR on harmonisation. Before harmonisation the modern scanner without RR adhered to EARL specification. Using the phantom data, filters were derived for optimal harmonisation between scanners and with and without RR as applicable, to the EARL standard. The 80-patient cohort did not reveal any statistically significant differences. In the 10-patient cohort SUVmax for RR >no RR irrespective of harmonisation but differences lacked statistical significance (one-way ANOVA F(3.36) = 0.37, p = 0.78). Bland-Altman analysis showed that harmonisation reduced the SUVmax ratio between RR and no RR to 1.07 (95% CI 0.96–1.18) with no outliers. EQ•PET successfully enabled harmonisation between modern and older scanners and against the EARL standard. Harmonisation reduces SUVmax and dependence on the use of RR in the modern scanner. EQ•PET is feasible to harmonise different PET-CT scanners and reduces the effect of resolution recovery on SUVmax.


  • Prognostic value of 18FDG PET/CT volumetric parameters in the survival prediction of patients with pancreatic cancer

    by Eyas Mohamed, Alexander Needham, Eftychia Psarelli, Melvyn Carroll, Sobhan Vinjamuri, Bal Sanghera, Wai Lup Wong, Christopher Halloran, Paula Ghaneh


  • Radionuclide calibrator intercomparison study of clinical PET centres in England to a single traceable 68Ge syringe source

    by Bal Sanghera, Andrew Fenwick, Gerry Lowe, Keith Sullivan, Wai-Lup Wong

    Objectives The aim of this study was to characterize national variation in radionuclide calibrator activity response to a single National Institute of Standards and Technology (NIST) traceable reference 68Ge source used as a surrogate for 18F at clinical PET centres in England using National Physical Laboratory approved techniques. Methods Readings from 20 instruments at 13 centres using local 18F and 68Ge factor settings were recorded with the source located in vial and syringe positions. Ten repeat measurements were conducted to investigate repeatability using % coefficient of variability (COV). Comparison ratios to investigate accuracy were made between calibrator responses and decay-corrected NISTref reference activity for syringe and vial position measurements. Results The maximum %COV was 0.79%, while 90, 95 and 80% of calibrators conformed to 5% accuracy for 18F syringe, 68Ge syringe and 68Ge vial position readings, respectively. We revealed a trend towards reduced bias in measurements using Veenstra devices for 18F and using Capintec devices for 68Ge factor settings. Conclusions This study demonstrated good repeatability in local device measurements. In total, 70% of English calibrators tested and 88% of all measurements performed achieved 5% accuracy. While statistically significant bias was exhibited between different vendor equipment dependent upon radioisotope selected, our study recommends regular traceability checks for optimum instrument performance conducted within National Metrology Institutes guidelines.

  • Open Access

    An Accurate Ensemble Classifier for Medical Volume Analysis: Phantom and Clinical PET Study

    by Mhd Saeed Sharif, Maysam Abbod, Ali Al-Bayatti, Abbes Amira, Ahmed S. Alfakeeh, Bal Sanghera

    The predominant application of positron emission tomography (PET) in the field of oncology and radiotherapy and the significance of medical imaging research have led to an urgent need for effective approaches to PET volume analysis and the development of accurate and robust volume analysis techniques to support oncologists in their clinical practice, including diagnosis, arrangement of appropriate radiotherapy treatment, and evaluation of patients’ response to therapy. This paper proposes an efficient optimized ensemble classifier to tackle the problem of analysis of squamous cell carcinoma in patient PET volumes. This optimized classifier is based on an artificial neural network (ANN), fuzzy C-means (FCM), an adaptive neuro-fuzzy inference system (ANFIS), K-means, and a self-organizing map (SOM). Four ensemble classifier machines are proposed in this study. The first three are built using a voting approach, an averaging technique, and weighted averaging, respectively. The fourth, novel ensemble classifier machine is based on the combination of a modified particle swarm optimization (PSO) approach and weighted averaging. Experimental National Electrical Manufacturers Association and International Electrotechnical Commission (NEMA IEC) body phantom and clinical PET studies of participants with laryngeal squamous cell carcinoma are used for the evaluation of the proposed approach. Superior results were achieved using the new optimized ensemble classifier when compared with the results from the investigated classifiers and the non-optimized ensemble classifiers. The proposed approach identified the region of interest class (tumor) with an average accuracy of 98.11% in clinical datasets of patients with laryngeal tumors. This system supports the expertise of clinicians in PET tumor analysis.


  • Open Access

    Treatment of classical Hodgkin lymphoma in young adults aged 18–30 years with a modified paediatric Hodgkin lymphoma protocol. Results of a multicentre phase II clinical trial (CRUK/08/012)

    by William Townsend, Sarah Leong, Peter Hoskin, Patricia Diez, Pip Patrick, David Linch, Wai‐Lup Wong, Irfan Kayani, Bal Sanghera, Andre Lopes, et al.

    Summary: This phase II trial was designed to determine the safety and efficacy of a modified paediatric risk‐stratified protocol in young adults (18–30 years) with classical Hodgkin Lymphoma. The primary end‐point was neurotoxicity rate. The incidence of grade 3 neurotoxicity was 11% (80% CI, 5–19%); a true rate of neuropathy of >15% cannot be excluded. Neuropathy and associated deterioration in quality of life was largely reversible. The overall response rate was 100% with 40% complete remission (CR) rate. Twelve months disease‐free survival (DFS) was 91%. We demonstrate that a risk‐stratified paediatric combined modality treatment approach can be delivered to young adults without significant irreversible neuropathy.


  • Estimating Uncertainty in Deep Learning for Reporting Confidence to Clinicians when Segmenting Nuclei Image Data

    by Biraja Ghoshal, Allan Tucker, Bal Sanghera, Wai Lup Wong

    Deep Learning, which involves powerful black box predictors, has achieved a state-of-the-art performance in medical image analysis such as segmentation and classification for diagnosis. However, in spite of these successes, these methods focus exclusively on improving the accuracy of point predictions without assessing the quality of their outputs. Knowing how much confidence there is in a prediction is essential for gaining clinicians' trust in the technology. Monte-Carlo dropout in neural networks is equivalent to a specific variational approximation in Bayesian neural networks and is simple to implement without any changes in the network architecture. It is considered state-of-the-art for estimating uncertainty. However, in classification, it does not model the predictive probabilities. This means that we are not capturing the true underlying uncertainty in the prediction. In this paper, we propose an uncertainty estimation framework for classification by decomposing predictive probabilities into two main types of uncertainty in Bayesian modelling: aleatoric and epistemic uncertainty (representing uncertainty in the quality of the data and in the model parameters, respectively). We demonstrate that the proposed uncertainty quantification framework using the Bayesian Residual U-Net (BRUNet) provides additional insight for clinicians when analysing images with help from deep learners. In addition, we demonstrate how the resulting uncertainty depends on the dropout rates using images from nuclei in divergent medical images.


  • Open Access

    Effect of PET Image Reconstruction Techniques on Unexpected Aorta Uptake

    by Hassan Hirji, Keith Sullivan, Imran Lasker, Mhd S. Sharif, Andre Nunes, Chris Shepherd, Wai-Lup Wong, Bal Sanghera

    Objectives: To determine if unexpected aorta uptake seen in some patients is influenced by popular modern reconstruction algorithms using semi-quantitative and qualitative analysis. Methods: Twenty-five consecutive patients without suspected vascular disease were selected for F-18-FDG positron emission tomography/computed tomography (PET/CT) scanning and images of the aorta were created using iterative reconstruction (IT), IT + time of flight (TOF), IT + TOF + point spread function correction (referred collectively as UHD) with and without metal artefact reduction (MAR) algorithms. An experienced radiologist created aorta and blood pool (BP) regions of interests then copied these to all reconstructions for accurate positioning before recording target aorta standardized-uptake-values (SUVmax) and background BP SUVmean. Furthermore, target-to-background ratio (TBRmax) was defined by aorta SUVmax -to-BP SUVmean ratio for more analysis. Results: For aorta SUV max with IT, IT + TOF, UHD, UHD + MAR reconstructions the mean +/- standard deviation recorded were 2.15 +/- 0.43, 2.25 +/- 0.51, 2.25 +/- 0.45 and 2.09 +/- 0.4, respectively. Values for BP SUV mean were 1.61 +/- 0.31, 1.58 +/- 0.28, 1.58 +/- 0.28 and 1.47 +/- 0.25, respectively. Likewise, for TBR max these were 1.35 +/- 0.19, 1.43 +/- 0.21, 1.43 +/- 0.19, 1.43 +/- 0.18, respectively. ANOVA analysis revealed no significant differences for aorta SUVmax (F(0.86) p= 0.46), BP SUVmean (F(1.22) p= 0.31) or TBRmax (F(0.99) p= 0.4). However, the qualitative visual analysis revealed significant differences between IT + TOF with UHD (p= 0.02) or UHD + MAR (p= 0.02). Conclusion: Reconstruction algorithm effect on aorta SUV max or BP SUV mean or TBR max was not statistically significant. However, qualitative visual analysis showed significant differences between IT + TOF as compared with UHD or UHD + MAR reconstructions. Harmonization of techniques with a larger patient cohort is recommended in future clinical trials.