RPS 2014 - Novel techniques and biomarkers for precision care

Author Block: J. T. d. S. d. Castro, H. d. J. Ferreira, D. C. Novais da Silva, D. Yamada, S. San Juan Dertkigil, F. Reis; Campinas/BR
Purpose: Computed tomographic pulmonary angiography (PE-CTA) must achieve ≥250 HU for reliable diagnosis; doing so with ultra-low iodine volumes across different body-weight groups is challenging. The aim is to evaluate the impact of weight-stratified, PSI-adapted PE-CTA protocols on arterial enhancement and iodine load.
Methods or Background: Single-centre retrospective longitudinal analysis within ANGIO-MONITOR. Consecutive PE-CTA exams were split by protocol change (go-live 01-Jun-2025): pre (fixed volumes; N=339) vs post (personalised 30–50 mL by weight with PSI ceilings; N=129). Primary endpoint: arterial attenuation (HU) in the pulmonary trunk. Secondary endpoints: proportion ≥250 HU, contrast volume (mL; mL/kg), peak PSI, and repeat acquisitions. Statistics: Welch’s tests for means; χ² for proportions; 95% CIs.
Results or Findings: Arterial enhancement increased from 313.5 HU (95% CI 301.3–325.7) to 499.7 HU (95% CI 465.5–533.9); p=5.38×10⁻¹⁹. The proportion ≥250 HU rose from 72.3% (95% CI 67.3–76.8; 245/339) to 95.3% (95% CI 90.2–97.9; 123/129); absolute Δ +23.1 pp, RR 1.32; p=5.26×10⁻⁸. Mean contrast volume fell from 60.0 mL (95% CI 58.3–61.8) to 43.8 mL (95% CI 41.5–46.1); −27.0%, p=1.22×10⁻²⁴. Normalised dose decreased from 0.825 mL/kg (95% CI 0.797–0.853) to 0.577 mL/kg (95% CI 0.547–0.606); −30.1%, p=2.31×10⁻²⁸. Peak PSI remained within operational safety, with a modest rise (98.9 to 107.8; 95% CIs 95.5–102.3 vs 102.1–113.5; p=0.0089). Repeat-acquisition rates were rare and stable (p=0.45).
Conclusion: Weight-stratified, PSI-adapted ultra-low iodine PE-CTA substantially increased enhancement above the 250 HU threshold while reducing iodine by ~27% (mL) and ~30% (mL/kg), without operational penalty. This supports iodine stewardship and safe high-flow practice in routine care, highlighting technologist-led implementation and hospital–industry synergy.
Limitations: The study is a single-center retrospective analysis, which may limit the generalizability of the findings.
Funding for this study: Not applicable.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The work has been submitted to the university's Research Ethics Committee (CEP) and is currently under review for evaluation and final opinion.

Author Block: L. Buttex, L. Gallus, Y. Cottier, C. S. D. Reis, S. S. Ghotra; Lausanne/CH
Purpose: To evaluate the application of artificial intelligence(AI) to reconstruct knee MRI imaging compared to standard protocols and identify if there is add-value for clinical practice.
Methods or Background: An exploratory study was conducted to evaluate AI-based image reconstruction methods in MRI compared to the standard(STD) knee MRI protocol. T1 Fast Spin Echo(FSE), PD FSE Fat saturation(FS) and T2 FSE in sagittal view, with PD FSE FS axial, and PD FSE FS coronal views were acquired. Protocols were optimised to enhance image quality(IQ) and reduce acquisition time. The final protocol was applied in 10 healthy volunteers using a 3T MRI to acquire the images. Visual assessment was performed by three MRI-experienced observers using the ViewDEX software. Visual Grading Analysis(VGA), Kappa statistics for inter-observer agreement and Visual Grading Characteristics(VGC) were performed to analyse data.
Results or Findings: Overall, the VGA scores indicate that the values for AI protocol are generally higher or equivalent to those for STD sequences across nearly all criteria of anatomical reproduction and IQ and 36.9% of time reduction was achieved as well 8min22s versus 13min15s. Kappa statistics indicated a medium to good agreement for AI protocol sequences(0.44-0.71) compared to low to good agreement for STD protocol sequences(0.13-0.71). VGC analysis showed that sequences reconstructed with the AI tool performed statistically better than STD sequences at 3T, with AUCVGC of 0.76 to 0.81(95% CI and p-value≤0.05).
Conclusion: This study demonstrates that the use of AI in knee MRI investigations has a tangible impact on the overall IQ, as well as on acquisition time, which was reduced by 36.9%. Further research is required to emphasise the clinical validation of the results with a larger sample size and cases including pathology.
Limitations: Small sample of healthy volunteers
Funding for this study: Not applicable
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethical approval Number: 2024-01679

Author Block: A. Roletto¹, A. Savio², E. Bertagna², N. S. Chiodini¹, G. R. Bonfitto¹, A. Chiti¹, S. Zanoni²; ¹Milan/IT, ²Brescia/IT
Purpose: While the environmental impact of diagnostic imaging is increasingly acknowledged, comprehensive analyses of nuclear medicine remain scarce. This study presents a Life Cycle Assessment (LCA) of routine [¹⁸F]FDG PET/CT examinations, with the aim of quantifying their Greenhouse Gas (GHG) footprint and identifying opportunities for impact reduction.
Methods or Background: A cradle-to-grave LCA was performed in accordance with ISO 14040 standards within a nuclear medicine department. A standard [¹⁸F]FDG PET/CT protocol was evaluated. Direct measurements were used to determine scanner electricity consumption, whereas heating, ventilation, and air conditioning (HVAC) energy demand was estimated for each relevant area based on data from the hospital trigeneration plant. Radiopharmaceutical production and single-use materials were modeled using SimaPro software.
Results or Findings: GHG footprint of a single PET/CT examination was estimated at 7.6 kg CO₂ eq. HVAC systems were the largest contributor (47% of CO₂ eq, 3.6 kg CO₂ eq.), encompassing the scan room, control room, injection room, and cyclotron facility. Radiopharmaceutical production represented 36% of the footprint, while direct scanner electricity accounted for only 5.4%. Single-use disposables were notable contributors to toxicity-related impacts and waste generation (2.1 kg CO₂ eq., 28%). Sensitivity scenarios suggested that ongoing decarbonization of the electricity grid could lower PET/CT-related carbon emission by up to 50%.
Conclusion: The environmental impact of [¹⁸F]FDG PET/CT is predominantly driven by HVAC energy demand and radiopharmaceutical production. Mitigation strategies, such as enhancing HVAC efficiency, increasing reliance on renewable energy, and replacing single-use items with reusable or lower-impact alternatives, could substantially reduce the carbon emission of nuclear medicine services.
Limitations: This analysis was conducted in a single-center setting with one representative PET/CT protocol, which may limit generalizability of the results.
Funding for this study: n/a
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: K. Borg Grima, J. L. Portelli, J. Borg, A. Rasalam Iris; Msida/MT
Purpose: Differentiated thyroid cancer (DTC) is primarily treated with surgery and radioactive iodine (RAI) ablation, achieving excellent outcomes in most patients. However, 5–10% of metastatic cases develop RAI-refractory disease (RAI-R), representing 60–70% of those with progressive disease. RAI-R patients face poor prognoses, with an average survival of 3–5 years. At present, clinicians lack biomarkers to predict RAI refractoriness, delaying the initiation of targeted therapies. The RADAR project aims to integrate the results of nuclear medicine imaging and molecular profiling into a radiogenomic framework of a unique population sample.
Methods or Background: This prospective study involves patients with DTC presenting at the only oncology centre in Malta. Data relating to clinical, biochemical, and histopathological findings, combined with imaging (I-131 SPECT/CT, 18F-FDG PET/CT) and molecular analyses, such as the microRNA profile, are collected longitudinally. Radiogenomic correlations are performed to identify patterns of RAI uptake and FDG avidity, potentially predictive of resistance. Candidate biomarkers will be validated and incorporated into a predictive risk algorithm.
Results or Findings: Previous research has shown that high FDG uptake (specificity 87.5%; sensitivity 83.3%) and certain genetic variants in the symporters are strongly associated with early loss of RAI avidity. By project completion, the study aims to achieve Technology Readiness Level 5 by delivering a validated biomarker panel and predictive model capable of stratifying high- versus low-risk patients for RAI-R progression.
Conclusion: The RADAR study introduces an innovative radiogenomic precision oncology approach, combining data obtained from nuclear medicine, radioiodine therapy and pathology. Early prediction of RAI-R could enable personalised treatment planning, reduce unintended toxicity and improve overall survival of DTC patients.
Limitations: RAI-R cases present a limited number of patients; however, data collection will run for the whole duration of the RADAR project.
Funding for this study: Funding for the RADAR project was obtained from the government agency , Xjenza Malta, responsible for promoting and coordinating scientific research in Malta.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: RADAR project team are currently in the process of obtaining ethical permission from the University of Malta Research Ethics Committee (UREC)

Author Block: M. F. Mcentee, A. England, C. Rainey, A. O’Donovan, S. Coakley, J. O'Neill; Cork/IE
Purpose: Adolescent Idiopathic Scoliosis (AIS) is a spinal deformity requiring accurate, high-quality imaging for diagnosis and treatment. While Digital Radiography (DR) remains the gold standard, EOS, a low-dose, biplanar imaging system, have emerged as an alternative. This study compares EOS and DR formage quality for whole-spine X-rays.
Methods or Background: A quantitative design assessed whole-spine imaging using anthropomorphic phantoms in two size configurations: Standard and Large. AP and lateral images were acquired with EOS and DR. Subjective image quality was assessed via a Visual Grading Analysis (VGA) questionnaire completed by 39 radiology professionals. Objective image quality was evaluated using Signal-to-Noise Ratio (SNR) and Contrast-to-Noise Ratio (CNR) metrics on ImageJ software, with statistical analysis via SPSSv27.
Results or Findings: Subjectively, DR scored higher in AP projections (standard: 72%vs.68%, p = 0.015; large:64%vs.60%, p=0.0004), while EOS led in lateral views (standard: 73%vs.69%, p=0.0001; large: 66%vs.60%, p<0.0001).EOS received more “Good” and “Moderate” ratings, but DR had more “Perfect” scores. Both DR and EOS had nearly identical numbers of "Inadequate" ratings. Objectively, DR showed higher SNR and CNR values, but significance was only seen for SNR in the large phantom AP view (DR 46.05 vs. EOS 28.00, p=0.004). Both systems showed reduced image quality with increased phantom size.
Conclusion: EOS demonstrated more consistent imaging across sizes and projections. DR showed marginally better SNR/CNR in some conditions. Subjective results favoured DR in AP views, and EOS in the lateral. Both modalities offer comparable image quality. EOS remains a promising option for AIS imaging; however, several key factors must be considered.
Limitations: Future research should involve real patient imaging, ideally in a multicenter, prospective design. Including a diverse sample of AIS patients would allow assessment of diagnostic performance, interobserver agreement, and real-world utility.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethics was approved by the School of Medicine's Social Ethics committee

Author Block: A. Collaku¹, Z. Mulla¹, E. Dybeli², F. Goga¹; ¹Tirana/AL, ²Elbasan/AL
Purpose: Coronary Computed Tomography Angiography (CTA) has emerged as a powerful non-invasive tool for evaluation of coronary artery disease (CAD). this study aimed to corelate CTA findings with Invasive Coronary Angiography (ICA), consider the gold standard, over a 10 year experience.
Methods or Background: We retrospectively reviewed cases in which patients underwent both CTA and ICA within a comparable clinical timeframe. Degree of stenosis in major coronary arteries was categorized, and CTA results were compared with ICA findings. Diagnostic accuracy, including sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), was calculated for detection of significant stenosis (≥50%)
Results or Findings: CTA demonstrated high sensitivity and NPV in ruling out significant CAD, showing strong concordance with ICA in the majority of cases. The technique was particularly reliable in single- and double-vessel disease, whereas specificity decreased in the presence of extensive calcification or motion artifacts. Despite these limitations, CTA consistently reduced the need for unnecessary invasive procedures.
Conclusion: CTA correlates strongly with ICA in the detection of significant CAD and can serve as an effective gatekeeper to invasive testing. In routine clinical practice, CTA offers a safe, accurate, and patient-friendly alternative that complements—but does not fully replace—the diagnostic value of ICA.
Limitations: No
Funding for this study: No
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: S. Recordon, M. Scheidegger, M. Thiébaud, I. Buchillier-Decka, K. Sprengers, V. Grilj; Lausanne/CH
Purpose: FLASH radiotherapy (FLASH-RT) delivers ultra-high-dose rates (UHDR >40 Gy/s) with the potential to preserve healthy tissue while maintaining tumor reduction efficacy. Accurate dosimetry is essential for clinical implementation. This study evaluates the performance of the Hyperscint™ RP-200 plastic scintillation detector for routine dosimetry in FLASH-RT, in comparison with detectors considered as references.
Methods or Background: The experiments were performed with the Oriatron eRT6 linear accelerator (CHUV, Lausanne) delivering 6 MeV pulsed electron beams in conventional (CONV) and FLASH modes. The RP-200 was tested for linearity, pulse dose response (PDR), repetition rate (PRF), and depth dose profiles (DDP). The measurements were compared with those of a AdvancedMarkus ionisation chamber and Gafchromic™ EBT3 films. The data were normalised by beam monitor and analysed by linear regression and dose ratios.
Results or Findings: The RP-200 demonstrated excellent linearity in CONV mode (R²=0.9996) with doses consistent with the reference detectors. In FLASH mode, a systematic underestimation occurred with deviations of 9 to 30%, increasing with PRF and DPP. PDD profiles differed from the references, particularly at a source-surface distance of 600 mm. These results confirm the limitations of the RP-200 in UHDR conditions, potentially related to Cherenkov light contamination.
Conclusion: Although reliable in conventional dosimetry, the RP-200 requires correction factors for routine dosimetry in FLASH-RT. Newer models such as the RP-100 may offer better UHDR performance. Further research is needed to optimise scintillator dosimetry in clinical FLASH.
Limitations: The main study limitation is the definition with certainty of the exact causes leading to differences in doses measured by the scintillator compared to the doses indicated by the films.
Funding for this study: Not applicable
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: I. Blom¹, M. Olsthoorn²; ¹De Lier/NL, ²Rotterdam/NL
Purpose: Computed Tomography (CT) is frequently used in orthopedic imaging to assess implant failure and complications such as periprosthetic fractures. However, the presence of metal implants often leads to metal artifacts (MA), which can significantly impair diagnostic image quality. Due to the wide variety of implant types and materials, the extent of MA varies considerably. As a result, a combination of artifact reduction techniques is often required to optimize visualization of both the implant and surrounding tissues. To evaluate the effectiveness of metal artifact reduction, multiple kiloelectron volt (keV) reconstructions were performed, with and without Orthopedic Metal Artifact Reduction (OMAR).
Methods or Background: Various orthopedic implants (hip, knee, shoulder, and spine) were scanned using a spectral CT scanner (IQon, Philips, The Netherlands) at 140 kV to reduce metal artifacts. Image reconstructions were performed both with and without OMAR, across a range of virtual monochromatic energy levels up to 200 keV. Hounsfield Units (HU) were measured using regions of interest (ROIs), with average (AVG) and standard deviation (SD) values recorded. Two musculoskeletal radiologists independently assessed diagnostic image quality using a 5-point Likert scale.
Results or Findings: Application of OMAR reduced SD in bone by approximately 10 HU, while increasing metal HU by 40. Bone AVG HU decreased by roughly 150 with OMAR across most joint revisions. Combining OMAR with high keV levels resulted in a 200 HU (AVG) increase in metal component.
Conclusion: Images reconstructed at 130–150 keV with OMAR yielded the highest diagnostic quality scores. However, 130 keV images without OMAR provided the most balanced HU values for both bone and metal, suggesting optimal contrast for diagnostic assessment.
Limitations: Case Study
Funding for this study: Not Applicable
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: J. Santos, F. Caseiro Alves; Coimbra/PT
Purpose: For lung diseases the literature recommends Computed Tomography (CT) for diagnostic, follow-up, planning and intervention. This study aims to redefine image quality indicators in chest CT, for pulmonary pathology, analysing CT exposure and image quality values.
Methods or Background: Ethical approval was obtained. In phase 1, the most relevant clinical CT indications for pulmonary pathology were identified. Dose values (n=90), described in the CT Dose Index (CTDIvol-mGy) and Dose Length Product (mGy.cm), were analysed. In phase 2, new imaging criteria were defined by 10 radiologists, using Delphi method. Image quality was evaluated objectively [10 Regions of Interest (ROI's)] and subjectively [(2 radiologists - Visual Grading Characteristic (VGC)]. In phase 3, dose values were analysed (n=100). Examinations performed on the same patients were analysed in terms of exposures and image quality.
Results or Findings: Microbacterial pathologies (MB), Lung nodules (NO) and Pulmonary fibrosis (FB) were the defined pathologies. The DLP values on phase 1, were 457mGy.cm for MB, 430mGy.cm for NO and 372mGy.cm for FB, without differences in quality criteria. The exposure values on phase 2, were 305mGy.cm for MB, 300mGy.cm for NO, and 372mGy.cm for FB. Analysis of the dose values of the same patients, revealed a variation from 399 to 324mGy.cm for MB, and from 378 to 324mGy.cm for NO. On phase 3 objective image quality identified lower noise values and higher area under the VGC curve (MB and NO).
Conclusion: The definition of new image quality criteria, with the application of robust statistical methods, allows a detailed analysis per imaging quality criteria. A reduction in DLP of 33% for MB, 30% for NO was obtains with an image quality improvement.
Limitations: The chest pathologies in study are focus on lung parenquima.
Funding for this study: No funding.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: CHUC ethical committe approval.

Author Block: Z. Mulla¹, A. Collaku¹, E. Dybeli², F. Goga¹; ¹Tirana/AL, ²Elbasan/AL
Purpose: To evaluate the impact of ultra-fast 640-slice CT (Aquilion ONE) on the reduction of sedation requirements in pediatric patients undergoing emergency and routine imaging.
Methods or Background: A retrospective review was performed on pediatric CT examinations (ages 1–12 years) conducted with the Aquilion ONE 640-slice system over a 12-month period. Parameters recorded included scan time, diagnostic image quality, presence of motion artifacts, and the rate of sedation. Protocols were optimized using individualized low-dose settings and iterative reconstruction techniques.
Results or Findings: The mean scan time was below 1 second for most brain, chest, and abdominal studies. Diagnostic quality was preserved in 95% of cases, with minimal need for repeat scanning. Only 7% of patients required sedation, compared with published rates of 30–40% with conventional CT. Motion artifacts were significantly reduced due to volumetric coverage in a single rotation.
Conclusion: The Aquilion ONE 640-slice CT significantly decreases the need for sedation in pediatric imaging by providing ultra-fast acquisitions with high diagnostic quality at optimized radiation doses. This approach improves patient safety, comfort, and clinical workflow, representing an important advancement in pediatric CT practice.
Limitations: No
Funding for this study: No
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: K. Tissir, P. Spagnolo; Milan/IT
Purpose: To quantitatively evaluate Cardiac imaging with Photon-Counting Computed Tomography (PCCT) for differentiating healthy myocardium, scar tissue, and edema using material decomposition maps and objective imaging metrics.
Methods or Background: Anonymised images from twenty-five patients with confirmed myocardial disease (acute or chronic infarction, myocarditis-related edema) undergoing PCCT with ECG-synchronised iodinated contrast were analysed. Reconstructions included conventional images, monoenergetic images (40–70 keV), and material decomposition maps (iodine, calcium, soft tissue). Myocardial regions of interest were segmented using Cardiac Mgnetic Resonance (CMR) imaging as reference. For each tissue type (healthy, scar and edema), mean attenuation associated to HU or iodine concentration, standard deviation, Signal-to-Noise Ratio (SNR), and Contrast-to-noise Ratio (CNR) were calculated. Receiver Operating Characteristic (ROC) analysis assessed discrimination performance between healthy and pathological tissue.
Results or Findings: Mean attenuation or iodine concentration differed significantly among tissues: healthy myocardium 55 ± 5 HU, scar 80 ± 8 HU, edema 65 ± 6 HU. CNR between scar and healthy tissue was 5.8 ± 1.2, SNR of healthy myocardium 32 ± 4. ROC analysis yielded area under the curve (AUC) of 0.87 for scar versus healthy myocardium and 0.81 for edema versus healthy myocardium. Voxel-wise concordance with CMR segmentation was 86%, indicating high agreement.
Conclusion: PCCT enables quantitative differentiation of myocardial tissues through material decomposition maps, providing reproducible SNR and CNR measurements. This pilot study demonstrates the feasibility of using PCCT for objective assessment of scar and edema, potentially complementing CMR in clinical practice.
Limitations: The study is limited by the small sample size and single-center design. Material decomposition maps require further validation against larger cohorts, and current results may not generalize to all PCCT scanners or acquisition protocols.
Funding for this study: n/a
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information: