Background & Context
Computed Tomography (CT) is one of the most widely used clinical imaging methods in Europe, with over 60 million patient examinations annually. Spectral CT (sCT) is an important new development which provides medical images as a function of x-ray energy — it delivers spectral information of the material properties. A typical clinical implementation uses specialised CT scanners that perform scans at two different energy levels.
Based on this spectral information, imaging biomarkers can be developed — quantitative features in medical images that help characterise disease and assess treatment response — thereby supporting personalised medicine.
However, for these benefits to be realised across Europe, several critical gaps must be addressed:
- Lack of traceable phantoms: Hospitals currently lack reliable, standardised reference objects (“phantoms”) that allow them to check whether sCT scanners measure imaging biomarkers accurately. Existing phantoms are not characterised traceable to national measurement standards; no methodology exists for their traceable characterisation.
- Inter-scanner variability: Different manufacturers use different technologies to create sCT images. The same patient scanned on two separate systems can produce different numerical results, making multi-centre studies, result comparisons across hospitals, or clinical thresholds difficult to establish.
- Dosimetry gaps: Existing dosimeter calibration standards (e.g. IEC 61267) are designed for older CT technologies and do not reflect the radiation fields used in sCT. New reference radiation fields and spectrometry methods must be developed. CT presents specific challenges including relatively high radiation dose rates and imaging geometries that generate substantial photon scatter.
- Personalised dose assessment: Patient radiation dose management must comply with EU Council Directive 2013/59/Euratom. Routinely used approaches for calculating patient doses are not scanner- and patient-specific, and proposed personalised approaches are currently too slow and labour-intensive for routine use.
Project Objectives
The project’s overarching objective is to build a reliable, standardised measurement infrastructure that will allow sCT imaging to be used confidently and safely in personalised medicine.
Objective 1 — Traceable phantom characterisation Develop a methodology for traceable characterisation of phantom inserts for phantoms used in spectral CT. This includes:
- Development of a method to measure energy-binned linear attenuation coefficients of reference phantom inserts at specific x-ray photon energies with a target uncertainty of 1%, established by at least 2 participating organisations
- Characterisation of inserts for at least 1 reference standard phantom
- Development of a method to assess uncertainties of energy-binned linear attenuation coefficients using virtual phantoms
Objective 2 — Multi-scanner variability study Assess the variability of imaging biomarkers between multiple sCT scanners in a multisystem (multicentre and vendor) study using the phantom inserts from Objective 1. Produce a Good Practice Guide on the assessment and correction of biomarker measurement variability and harmonised protocols for multicentre studies.
Objective 3 — Spectral characterisation of sCT radiation fields Develop metrological equipment and procedures for the spectral characterisation of sCT radiation fields. This includes development of approaches for sCT imaging and radiation dosimetry (e.g. measurement setups for spectrometry at clinical CTs) and application of developed procedures on a representative set of sCT scanners.
Objective 4 — Personalised dosimetry software framework Develop a harmonised and validated software framework to facilitate personalised dosimetry in sCT and CT in clinical practice. This includes:
- Algorithms for AI-assisted data-driven personalised dose assessment (target uncertainty: 20%)
- Algorithms for AI-based organ segmentation and matching
- Calculations of organ-dose volume histograms based on real patient anatomies
- Testing in a multi-centre study
Objective 5 — Dissemination and uptake Facilitate the take-up of technology and measurement infrastructure by the measurement supply chain, standards-developing bodies (IEC TC 62), professional and scientific societies, international and regulatory organisations, industry, and end users.
Expected Outcomes
- Calibrated physical phantoms for sCT with traceable uncertainty budgets
- Validated digital (virtual) phantom models
- Good Practice Guide on biomarker variability assessment and correction
- Harmonised clinical imaging protocols for multi-centre sCT studies
- Validated catalogue of sCT-specific x-ray spectra
- AI-based personalised dosimetry framework with full uncertainty evaluation
- Input to IEC 63483, IEC 61267, IEC 61674 and IAEA TRS-457
