Spectral CT and Photon-Counting CT in Neuroimaging

Spectral computed tomography (CT) and photon-counting detector (PCD) CT are extending CT beyond conventional anatomical imaging through energy-resolved acquisition, improved tissue characterisation and reduced artefacts. A recent review in Radiology: Imaging Cancer assesses their physical principles and applications in neuroradiology, with particular attention to head and neck cancer, primary central nervous system tumours and brain metastases. Spectral CT acquires data at two or more energy levels, while PCD CT counts individual x-ray photons and measures their energies. These approaches can improve spatial resolution, contrast-to-noise performance, material differentiation and quantitative imaging while supporting dose optimisation in selected settings. 

Energy-Resolved Imaging Expands CT Capabilities 

Spectral CT uses differences in x-ray attenuation across energy levels to distinguish materials and generate quantitative information. Dual-energy approaches acquire data at two energy spectra, while PCD CT records individual photons across multiple energy bins. PCD systems also use smaller detector pixels and reduce the influence of electronic noise, supporting higher spatial resolution and more detailed spectral separation than conventional energy-integrating detector CT. 

Several reconstruction tools extend these capabilities. Iodine maps can increase lesion conspicuity by showing iodine uptake and can provide quantitative information on vascularity and perfusion. Virtual noncontrast images can be generated from contrast-enhanced scans, potentially removing the need for a separate unenhanced acquisition in appropriate indications. Virtual monoenergetic imaging can increase iodine contrast at lower energy levels or reduce beam-hardening artefacts at higher energies. Material decomposition can also separate substances such as iodine, calcium, blood and soft tissue. 

These features support functional and quantitative assessment alongside anatomical imaging. Low-energy images can improve visualisation of small lesions or lesions in complex anatomy, although greater enhancement may be accompanied by more image noise. Dose-saving potential is also reported, but dedicated evidence in neuro-oncologic populations remains limited. Replacement of true noncontrast imaging with virtual noncontrast images also requires validation for the specific indication and scanner platform. Prospective work is still needed to establish optimised acquisition protocols, diagnostic performance and safe dose thresholds. 

Neuro-Oncology Gains from Detail and Material Mapping 

In head and neck imaging, spectral CT and PCD CT can improve tumour delineation, tissue contrast and assessment of small anatomical structures. Higher spatial resolution can help define tumour margins and extension into cartilage or bone. PCD CT can also reduce artefacts from dental implants and fillings, improving visualisation in areas where conventional CT may be degraded by metal. 

Must Read: Photon-Counting CT Improves ARDS Chest Imaging 

Quantitative spectral information adds another layer to tumour assessment. Iodine density mapping can distinguish tumour from normal mucosa and provide information related to tumour vascularity and perfusion. Effective atomic number mapping can improve material discrimination and has been used to differentiate tumour from normal tissue, distinguish tumour subtypes and separate recurrent disease from post-treatment changes. Spectral measurements may also help differentiate metastatic from benign or inflamed lymph nodes. 

For primary central nervous system tumours, early-phase evidence indicates roles in tumour characterisation, grading, treatment-response assessment and radiation therapy planning. Iodine and spectral attenuation measurements can help distinguish recurrent glioma from treatment-related changes by reflecting differences in vascularity. Spectral and PCD CT can also quantify features related to tumour composition and microvascularity. Evidence for brain metastases is more limited, but low-energy imaging can improve the conspicuity of small enhancing lesions, while material decomposition can help separate enhancement from haemorrhage or calcification. In current practice, these techniques are positioned as advanced CT options when MRI is limited or as complementary imaging alongside MRI. 

Technical and Workflow Barriers Limit Wider Adoption 

Wider clinical implementation remains constrained by detector, reconstruction and operational challenges. At high photon flux, photon-counting detectors can misregister closely arriving photons, while charge sharing and cross talk can reduce spectral accuracy and material differentiation. Current detectors also have limits in separating photons with very similar energies. Electronic noise at low x-ray flux and beam-hardening near dense structures remain additional concerns, requiring advanced reconstruction and correction methods. 

PCD CT also brings greater hardware complexity, with advanced semiconductor materials and high-speed electronics increasing cost and maintenance requirements. Reliable quantitative imaging requires calibration, validation and artefact correction, while clinical use may require changes to imaging protocols, radiologist training and contrast administration. Integration of quantitative outputs into picture archiving and radiology information systems, accessible postprocessing tools, reimbursement pathways and standardised reporting are also identified as requirements for broader use. Structured training programmes and consensus guidelines may also support reproducibility across institutions. User-friendly postprocessing tools are needed to avoid adding interpretive burden. 

Development is continuing in detector design, reconstruction and artificial intelligence-based correction. Deep learning approaches are being investigated for denoising, artefact reduction, material decomposition and improvement of virtual monoenergetic and virtual noncontrast images. However, much of the current artificial intelligence evidence in neuro-oncology comes from MRI-based work, so validation on spectral CT and PCD CT datasets remains necessary. 

Spectral CT and PCD CT extend neuroradiologic CT through improved tissue characterisation, higher spatial resolution, quantitative material mapping and artefact reduction. Their applications span head and neck tumours, central nervous system neoplasms, brain metastases, vascular imaging and paediatric oncology, with potential benefits for lesion detection, treatment assessment and dose optimisation. Evidence remains uneven across applications, particularly for clinical outcomes and neuro-oncologic dose reduction. Technical detector limitations, reconstruction demands, hardware costs and workflow integration continue to restrict widespread adoption, making further clinical validation and technology development central to broader implementation. 

Source: Radiology: Imaging Cancer 

Image Credit: iStock