1. Introduction

The non-contrast brain CT (NCCT) is the single most consequential imaging examination in acute medicine. Performed without intravenous contrast, it exploits the intrinsic attenuation differences between brain tissue, cerebrospinal fluid, blood, calcium, and air to answer one urgent clinical question with speed that no other modality can match: is there blood in the brain? In the context of a patient presenting with sudden-onset neurological deficit, the NCCT protocol can safely exclude haemorrhage, confirm early ischaemic change, and direct the clinical team toward thrombolysis or thrombectomy within the first ten minutes of emergency arrival.^[1]

Non-contrast brain CT occupies a unique position in the imaging workflow precisely because it is radiographically straightforward but diagnostically deceptively demanding. The absence of contrast means that every diagnostic conclusion rests entirely on the radiographer’s technical execution and the radiologist’s pattern recognition of subtle attenuation differences — often measured in single-digit Hounsfield Unit variations against a background of complex brain architecture.^[2] A 6 HU shift in the cortical ribbon, barely perceptible on a standard brain window, can represent the difference between a viable penumbra and an established infarct core.

Globally, acute stroke represents one of the leading causes of death and long-term disability. Approximately 13.7 million new strokes occur annually worldwide, of which roughly 87% are ischaemic in origin.^[3] The non-contrast brain CT is the gatekeeper for all acute stroke management decisions: it determines candidacy for intravenous alteplase (rtPA), directs early referral to comprehensive stroke centres, and establishes baseline anatomy for follow-up imaging. For intracranial haemorrhage — epidural, subdural, subarachnoid, intraparenchymal, and intraventricular — non-contrast CT remains the gold standard first-line investigation, with sensitivity for acute blood exceeding 98% in the first 24 hours.^[4]

This article provides a comprehensive, evidence-based reference for the NCCT brain protocol: scanning parameters, anatomy, HU interpretation ranges, DRL-aligned dose benchmarks, ten key pathologies, and a structured pitfall framework designed for radiographers, radiologists, and non-radiology clinicians alike. It is Article 1 of the 30-Day CT Protocol Mastery Series, establishing the foundational framework that all subsequent brain, head and neck, and chest protocols will build upon.

Understanding the technical underpinning of the NCCT brain is not academic exercise. Every parameter chosen — tube voltage, pitch, rotation time, reconstruction kernel, slice thickness — directly affects the diagnostic information that reaches the reporting radiologist. The clinical stakes could not be higher.

2. Anatomy & HU values

Brain parenchyma and normal attenuation values

The Hounsfield Unit (HU) scale defines the linear transformation of measured X-ray attenuation coefficients, with water assigned a value of 0 HU and air at −1000 HU. The brain, composed predominantly of water-containing tissue, grey matter, white matter, and myelin-wrapped axons, occupies a narrow band of the scale. This narrow range is precisely what makes the NCCT brain technically challenging: pathological changes that are clinically critical manifest as shifts of only 5–15 HU against backgrounds of adjacent normal tissue.

Grey matter sits consistently between +35 and +45 HU, reflecting its higher neuronal cell density and cellularity. White matter, composed principally of myelinated axon tracts, measures +20 to +35 HU. The grey-white matter differentiation — the sharp interface between these two values — is one of the most clinically important signs in the NCCT brain. Early cytotoxic oedema from acute ischaemia causes the grey matter to swell and lose its density, bringing it below <30 HU and causing the grey-white boundary to blur or disappear.^[5] This is the radiological correlate of the insular ribbon sign, MCA territory obscuration, and early sulcal effacement.

Complete HU reference table for brain structures

Gross anatomy: lobes, cisterns, and vascular territories

The cerebral hemispheres are divided into four lobes — frontal, parietal, temporal, and occipital — each with identifiable sulci and gyri visible on NCCT in thin-slice reconstructions (1–3 mm). The Sylvian fissure and interhemispheric fissure serve as key midline landmarks. Asymmetry in their appearance — widening, effacement, or mass effect upon them — constitutes an early and critical imaging sign. The basal cisterns, including the perimesencephalic cistern, suprasellar cistern, and quadrigeminal plate cistern, are normally filled with CSF and appear hypodense. Their obliteration signals raised intracranial pressure and impending herniation.

Vascular territories are fundamental to NCCT interpretation. The middle cerebral artery (MCA) territory encompasses the bulk of the lateral hemisphere, including the motor and sensory cortex and Broca’s and Wernicke’s areas. The anterior cerebral artery (ACA) territory covers the medial frontal and parietal lobes. The posterior cerebral artery (PCA) territory supplies the occipital lobes and medial temporal structures. The posterior fossa, supplied by the vertebrobasilar system, is the most technically challenging region on NCCT due to beam hardening from the petrous bone and skull base.^[6]

The ASPECTS score and NCCT ischaemia quantification

The Alberta Stroke Program Early CT Score (ASPECTS) is a 10-point topographic scoring system applied to the MCA territory on NCCT. Each of ten anatomical regions — caudate, lentiform nucleus, internal capsule, insular cortex, and six cortical zones (M1–M6) — is awarded one point if normal and zero points if early ischaemic change is present. A score of 10 indicates no visible ischaemia; a score below 6 predicts poor functional outcome and reduced benefit from reperfusion therapy.^[7] The ASPECTS assessment is routinely integrated into AI-assisted stroke triage platforms and is the primary quantitative output of most FDA-cleared stroke detection algorithms.

Posterior fossa anatomy

The posterior fossa is bounded by the tentorium cerebelli superiorly and the foramen magnum inferiorly. It contains the cerebellum, pons, and medulla — collectively the brainstem — along with the fourth ventricle. The petrous ridges of the temporal bone, densely calcified, create significant beam hardening on standard NCCT, producing dark horizontal streak artefacts across the pons and medulla. These Hounsfield dark bands represent one of the most clinically dangerous interpretation pitfalls on NCCT, as they can precisely simulate the appearance of acute brainstem infarction.^[8] Posterior fossa strokes are therefore systematically underdiagnosed on NCCT, with sensitivity for acute cerebellar or brainstem infarction as low as 30–40% in the first 24 hours — a critical limitation that all clinicians ordering NCCT for posterior circulation symptoms must understand.^[6]

3. Scanning technique

Standard 7-step NCCT brain acquisition protocol

Step 1 — Patient preparation and positioning. Position the patient supine with the head in the head holder, neutral rotation, chin slightly tucked. The orbitomeatal (OM) line should be perpendicular to the table top, or a slight negative tilt of 5–10° below the OM line may be used to reduce lens dose. Confirm no metallic artefacts (earrings, hair clips, dental appliances when feasible) in the field of view. For agitated or confused patients, alert the requesting clinician immediately; sedation or pharmacological cooperation protocols should be activated before scanning to avoid motion degradation.^[9]

Step 2 — Scout localiser. Acquire a lateral skull scout at low dose (typically 80 kVp, 10–20 mA). Use the scout to define the scan range from the foramen magnum to the vertex. Exclude the orbits from the primary helical range whenever possible to reduce lens dose, unless orbital pathology is specifically indicated.

Step 3 — Protocol parameter verification. Confirm: 120 kVp, AEC active (250–350 mA range), sequential mode or pitch 0.55, rotation time 1.0 s (reduce to 0.5 s for agitated patients), reconstruction slice thickness 5 mm (brain), 1.25 mm (posterior fossa), 1.25 mm (bone kernel for vault/base).

Step 4 — Dual reconstruction kernels. Reconstruct all NCCT brain examinations with at minimum two algorithms: a soft tissue kernel (e.g., H20f, B30f, or equivalent) for parenchymal evaluation and a bone kernel (e.g., H60f, B70f) for calvarium, skull base, and orbital evaluation. Where available, a deep learning reconstruction (DLR) medium-to-high strength algorithm should replace filtered back projection to reduce noise at equivalent or lower dose.^[10]

Step 5 — Immediate acquisition (no delay). NCCT is acquired immediately — there is no pre-injection or timing delay. In acute stroke triage, the interval from patient arrival to image acquisition should not exceed 25 minutes per AHA/ASA guidelines.^[1]

Step 6 — Slice series for reporting. Generate: axial 5 mm (standard brain series), axial 1.25 mm soft tissue (for subtle subarachnoid blood, small haematomas, and cortical detail), axial 1.25 mm bone kernel (vault integrity, base of skull fractures), coronal 2 mm MPR reformats (ventricular system, subfrontal/temporal pathology), sagittal 2 mm reformats (midline shift quantification).

Step 7 — Verify and transfer. Confirm all series have transferred to PACS. Check that window presets (brain W80/L35, subdural W200/L75, stroke narrow W35/L30, bone W3500/L500) are pre-configured at the reporting workstation. In emergency triage workflows, automated PACS routing with AI pre-read notification should be confirmed as operational before the patient departs the department.

Scanner comparison: 16-slice to 320-slice NCCT brain parameters

Dual-energy and photon-counting CT protocols for the brain

Deep learning reconstruction (DLR) in NCCT brain

Deep learning image reconstruction (DLR) algorithms — commercially available as GE’s TrueFidelity, Canon’s AiCE, Siemens’ ADMIRE-AI, and Philips’ HQ-Recon — have demonstrated consistent benefit in non-contrast brain CT. At equivalent dose to filtered back projection (FBP), DLR algorithms reduce image noise by 40–60% while preserving or improving spatial resolution.^[10] In the NCCT brain context, this noise reduction significantly improves grey-white matter contrast-to-noise ratio (CNR), enhancing the detectability of early ischaemic change and small subdural collections.

A systematic review published in 2025/26 confirmed that DLR reduces effective dose by 20–40% compared to FBP while maintaining non-inferior diagnostic accuracy for intracranial haemorrhage and early infarction detection.^[12] DLR should be considered the minimum reconstruction standard for NCCT brain on any modern scanner platform; FBP-only acquisitions are no longer compliant with current dose optimisation guidelines from the ACR, ESR, and ICRP.

4. Contrast media protocol

The NCCT brain protocol uses no intravenous contrast media. This is not a simplification or a cost-saving measure; it is a fundamental technical requirement. The administration of iodinated contrast alters the HU of blood vessels, disrupted blood-brain barrier regions, and enhancing lesions, which would directly obscure and confound the primary diagnostic targets of the non-contrast examination: hyperacute haemorrhage, the hyperdense artery sign, and early ischaemic grey-matter changes.

Pre-scan safety checklist for NCCT brain

Although no contrast is administered for the standard NCCT brain, the following pre-scan safety checks remain mandatory in the acute setting. First, confirm the patient’s identity against the request form using at minimum two patient identifiers. Second, assess the patient’s ability to cooperate: acute stroke, head injury, and alcohol intoxication are common causes of non-compliance. Third, screen for implanted medical devices — cochlear implants, vagal nerve stimulators, and deep brain stimulators are generally compatible with CT but must be documented. Fourth, confirm that clinical urgency and radiation justification are documented by the requesting clinician in accordance with IR(ME)R 2017 (UK), EU BSS Directive, or equivalent national framework.^[13]

5. Radiation dose

Non-contrast brain CT represents one of the higher-dose CT examinations relative to other head and neck protocols, driven by the need to maintain adequate grey-white matter CNR while imaging through the dense calvarium. Radiation dose optimisation is a formal ethical and regulatory obligation under the ALARA (As Low As Reasonably Achievable) principle and EU Radiation Protection legislation, and must be actively managed at both the individual patient and population level.^[13]

Diagnostic reference levels (DRLs) — NCCT brain

The Size-Specific Dose Estimate (SSDE) provides a more patient-individualised dose metric than CTDIvol, accounting for the patient’s actual head size. Modern scanner AEC systems and dose reporting tools (e.g., DoseWatch, Radimetrics, DoseSPX) should be used to log SSDE for every examination and to flag outliers above the 75th percentile DRL for retrospective audit.^[14]

5 evidence-based dose reduction strategies for NCCT brain

Strategy 1 — Deep learning image reconstruction. Replacing FBP with DLR at medium or high strength enables equivalent diagnostic image quality at 20–40% lower dose. Canon AiCE, GE TrueFidelity, Siemens ADMIRE-AI, and Philips HQ-Recon are the principal FDA-cleared platforms. All have demonstrated maintained or superior CNR for intracranial haemorrhage detection compared to FBP at reduced dose settings.^[10]

Strategy 2 — Automatic exposure control (AEC) activation. AEC systems (angular and Z-axis modulation) automatically reduce tube current in low-attenuation regions. Ensure AEC is enabled with a noise reference index appropriate for head scanning (quality reference mAs typically 250–350). Scanning with fixed high mAs without AEC can result in doses 25–40% above the achievable level unnecessarily.^[14]

Strategy 3 — Minimise scan range. Restricting the craniocaudal extent to the true brain volume (foramen magnum to vertex) reduces DLP by up to 15% compared to protocols that include neck soft tissue or use over-range scanning. Always review the scout before confirming the scan range.

Strategy 4 — Reduce kVp for smaller patients. The standard 120 kVp is appropriate for adults. For paediatric patients or small-headed adults (head circumference <50 cm), reducing to 100 kVp with compensatory mAs increase maintains CNR while reducing effective dose by approximately 20–25%.^[15]

Strategy 5 — Bismuth-free lens protection. Historical bismuth lens shields are no longer recommended, as they increase surface dose and scatter without reliably reducing lens dose. Modern scanner geometries with adjusted scan starting positions (first slice below lens level) and negative gantry tilt of 5–10° provide effective lens dose reduction without scatter penalties.^[15]

6. Top 10 pathologies on NCCT brain

The NCCT brain is specifically optimised to evaluate a discrete, clinically critical set of pathologies. Each relies on a distinct attenuation signature against the background of normal brain tissue. Understanding the HU basis of each diagnosis is not merely academic: it determines the window setting required, the confidence with which a diagnosis can be made, and when the findings are ambiguous enough to mandate additional sequences or modalities.^[4]

7. Pitfalls for radiographers

The primary scanning pitfall for NCCT brain is explicitly identified in the clinical dataset as kinetic motion from confused or agitated stroke/trauma patients. This is not merely an image quality issue — it is a diagnostic safety event. When an agitated patient moves during a standard 1.0 s rotation time, the resulting motion artefact can obscure or simulate a haemorrhage, distort the grey-white interface beyond reliable assessment, and necessitate repeat scanning with additional radiation dose. In the highest-acuity cases, this delay may be clinically catastrophic.

The mitigation is direct and evidence-supported: reduce rotation time to 0.5 s for any patient suspected of poor compliance, or use wide-detector (256–320 slice) scanners capable of single-rotation brain acquisition in under 0.4 seconds total, eliminating the physiological motion window entirely. Some clinical environments also incorporate pharmacological intervention protocols for extreme cases, executed in coordination with the clinical team before scanning.

Complete radiographer pitfall framework: NCCT brain

8. Pitfalls for radiologists

The primary interpretation pitfall for NCCT brain is explicitly identified in the clinical data as: beam hardening at the skull base and posterior fossa creates dark pseudo-streak bands across the pons, mimicking an acute brainstem infarct. This is among the most consequential diagnostic errors in emergency neuroradiology because it occurs precisely in the region where early ischaemic change is already least reliably detected, and where false confidence in a “positive” beam hardening artefact can both delay MRI and — critically — cause false reassurance if the artefact is dismissed as normal when a genuine infarct is present.

The mechanism is straightforward: dense petrous bone produces preferential absorption of lower-energy X-ray photons. As a beam passes through the petrous ridges bilaterally, the spectrum hardens and the reconstructed values in the intervening pons and medulla are artificially lowered, producing dark horizontal or diagonal streaks. These streaks are bilateral, symmetric, and do not respect vascular territory boundaries — which are key discriminating features. An acute brainstem infarct, by contrast, tends to be unilateral or have clear territorial distribution and should be accompanied by appropriate clinical history.^[8]

Complete interpretation pitfall framework: NCCT brain

9. Pitfalls for non-radiology physicians

Clinicians interpreting or acting on NCCT brain reports face a distinct category of cognitive and contextual pitfalls. Unlike radiographers, whose errors are primarily technical, and radiologists, whose errors are primarily perceptual-interpretive, the physician’s pitfalls centre on misapplication of imaging findings to clinical decision-making. The following table addresses the most clinically dangerous errors in context.

10. Pitfall comparison summary

The following three-column reference consolidates the primary pitfalls for each professional group into a single, scannable table. It is designed as a rapid clinical governance reference tool — for use in departmental training, protocol review meetings, and quality assurance audits.

11. AI & automation in NCCT brain

Artificial intelligence has achieved its most mature and evidence-validated clinical deployment in the non-contrast brain CT space. The urgency of acute stroke triage — where every 10-minute delay reduces the probability of a good functional outcome — created the exact clinical environment in which AI prioritisation, automated quantification, and workflow acceleration tools generate measurable patient benefit. The FDA and CE regulatory bodies have cleared a growing suite of tools specifically targeting NCCT brain pathologies.^[19]

FDA-cleared and CE-marked AI tools for NCCT brain

Viz.ai (viz.LVO, viz.ICH, viz.ASPECTS). FDA De Novo cleared. Provides automated large vessel occlusion detection on CTA integrated with NCCT, automated ICH detection and volumetric quantification, and AI-generated ASPECTS scoring. The viz.ai platform interfaces with the electronic health record and directly alerts the on-call neurologist and interventional team via mobile notification, with published studies demonstrating reduction in door-to-treatment time by 60–90 minutes in dedicated stroke centres.^[20]

Aidoc (Intracranial Haemorrhage AI, PE/ICH integrated). FDA-cleared for ICH detection and triage prioritisation across NCCT brain. Aidoc’s background AI continuously analyses all NCCT studies as they are acquired and flags studies with suspected haemorrhage for urgent radiologist review, reducing haemorrhage reporting time by a median of 9.6 minutes in prospective evaluation.^[21]

RapidAI NCCT (iSchemaView). FDA and CE cleared. The RapidAI suite provides automated ASPECTS scoring, LVO detection (from CTA when combined), infarct core and penumbra estimation (from CTP), and NIHSS correlation tools. It is endorsed in the 2022 AHA/ASA stroke guidelines as an acceptable adjunct to radiologist evaluation for ASPECTS quantification.^[1]

RAPID NCCT AI and Brainomix e-ASPECTS. CE-marked for ASPECTS scoring in Europe. Both platforms provide structured ASPECTS maps with automated zonal scoring and uncertainty quantification, supporting the radiologist’s final interpretation rather than replacing it. Prospective multicentre studies have confirmed equivalent inter-rater agreement between AI ASPECTS and expert neuroradiologist ASPECTS for scores in the 5–9 range.^[7]

Deep learning reconstruction as the AI-dose optimisation interface

It is important to distinguish between AI tools for diagnostic decision support (pathology detection, ASPECTS scoring, triage prioritisation) and AI tools embedded within the scanner reconstruction pipeline (DLR algorithms). Both are forms of applied artificial intelligence, but they operate at different points in the imaging chain. DLR algorithms within the scanner reduce noise and maintain resolution at lower dose, enhancing the raw diagnostic information delivered to the radiologist’s workstation. Diagnostic AI tools then act on this optimised image data to support the reporting process. The combination of DLR acquisition and AI-assisted reporting represents the current state-of-the-art standard of care for NCCT brain in high-volume stroke centres.^[22]

Photon-counting CT (PCCT) represents the next evolutionary step, offering intrinsic noise reduction through direct photon detection (eliminating electronic noise entirely), sub-millimetre spatial resolution (0.2 mm), and multi-energy spectral imaging without additional dose. Early clinical data from PCCT-equipped centres suggest improved detection of early ischaemic change and significantly improved posterior fossa imaging due to inherent beam hardening reduction from multi-bin energy processing.^[11]

13. Conclusion

The non-contrast brain CT is simultaneously the most frequently performed and the most diagnostically consequential CT examination in acute medicine. Its technical parameters — 120 kVp, sequential acquisition, 1.0 s rotation time, dual reconstruction kernels, and immediate scanning without contrast — are individually calibrated to maximise the detection of haemorrhage, early ischaemic change, and structural abnormalities that drive immediate management decisions in acute stroke, trauma, and neurosurgical emergencies.

Mastery of the NCCT brain protocol requires three simultaneous competencies: the radiographer’s technical precision in producing artefact-free, dose-optimised images with consistent grey-white CNR; the radiologist’s interpretive expertise in recognising subtle HU shifts, identifying beam hardening artefacts without over- or under-calling pathology, and applying structured frameworks like ASPECTS to quantify ischaemic burden; and the clinician’s contextual understanding that a normal NCCT report does not exclude stroke, that posterior fossa NCCT has fundamental sensitivity limitations, and that the NCCT is always a gateway — not a terminus — for the acute neurological workup.

The emerging integration of deep learning reconstruction at the acquisition stage, AI-assisted ASPECTS scoring and haemorrhage triage at the reporting stage, and photon-counting CT at the hardware level collectively represent the most significant advances in NCCT brain performance in the protocol’s five-decade clinical history. Departments that invest in aligning their scanner protocols, reconstruction standards, and AI implementation with current evidence-based benchmarks are not merely optimising images — they are directly reducing the time from symptom onset to reperfusion, and measurably improving neurological outcomes for their most critically ill patients.

The pitfall framework presented in this article — spanning radiographer scanning errors, radiologist interpretation traps, and clinician decision-making blind spots — provides a structured governance tool for quality assurance programmes, multidisciplinary training events, and root cause analysis of adverse imaging events. These are not theoretical risks. Each pitfall documented here has a documented patient harm pathway. Understanding and systematically mitigating them is the professional obligation of every member of the CT neuroimaging team.

The NCCT brain is where CT protocol mastery begins. Every subsequent article in this 30-Day CT Protocol Mastery Series builds upon the foundational principles established here: the importance of technical precision, structured interpretation, and multidisciplinary pitfall awareness. The urgency is not abstract — it is measured in the 1.9 million neurons lost every minute that a large vessel occlusion goes untreated.

14. References

1. Powers, W. J., Rabinstein, A. A., Ackerson, T., Adeoye, O. M., Bambakidis, N. C., Becker, K., Biller, J., Brown, M., Demaerschalk, B. M., Hoh, B., Jauch, E. C., Kidwell, C. S., Leslie-Mazwi, T. M., Ovbiagele, B., Scott, P. A., Sheth, K. N., Southerland, A. M., Summers, D. V., & Tirschwell, D. L. (2019). Guidelines for the early management of patients with acute ischemic stroke: 2019 update. Stroke, 50(12), e344–e418. https://doi.org/10.1161/STR.0000000000000211
2. Wardlaw, J. M., Seymour, J., Cairns, J., Keir, S., Lewis, S., & Sandercock, P. (2004). Immediate computed tomography scanning of acute stroke is cost-effective and improves quality of care. Stroke, 35(11), 2477–2483. https://doi.org/10.1161/01.STR.0000143453.78005.44
3. Feigin, V. L., Brainin, M., Norrving, B., Martins, S., Sacco, R. L., Hacke, W., Fisher, M., Pandian, J., & Lindsay, P. (2022). World Stroke Organization (WSO): Global stroke fact sheet 2022. International Journal of Stroke, 17(1), 18–29. https://doi.org/10.1177/17474930211065917
4. Carpenter, C. R., Hussain, A. M., Ward, M. J., Zipfel, G. J., Fowler, S., Pines, J. M., & Sivilotti, M. L. A. (2016). Spontaneous subarachnoid hemorrhage: A systematic review and meta-analysis describing the diagnostic accuracy of history, physical examination, imaging, and lumbar puncture with an exploration of test thresholds. Academic Emergency Medicine, 23(9), 963–1003. https://doi.org/10.1111/acem.12984
5. Mair, G., Alzahrani, A., Lindley, R. I., Sandercock, P. A. G., & Wardlaw, J. M. (2020). Feasibility and diagnostic accuracy of using brain attenuation changes on CT to estimate time of ischemic stroke onset. Neuroradiology, 63(4), 541–550. https://doi.org/10.1007/s00234-020-02591-w
6. Edlow, J. A., Newman-Toker, D. E., & Savitz, S. I. (2015). Diagnosis and initial management of cerebellar infarction. Lancet Neurology, 7(10), 951–960. https://doi.org/10.1016/S1474-4422(08)70216-3
7. Goyal, M., Menon, B. K., van Zwam, W. H., Dippel, D. W., Mitchell, P. J., Demchuk, A. M., Dávalos, A., Majoie, C. B., van der Lugt, A., de Miquel, M. A., Donnan, G. A., Roos, Y. B., Bonafe, A., Jahan, R., Diener, H. C., van den Berg, L. A., Levy, E. I., Berkhemer, O. A., Pereira, V. M., … Saver, J. L. (2016). Endovascular thrombectomy after large-vessel ischaemic stroke: A meta-analysis of individual patient data from five randomised trials. The Lancet, 387(10029), 1723–1731. https://doi.org/10.1016/S0140-6736(16)00163-X
8. Moretti, R., Torre, P., Antonello, R. M., Cazzato, G., & Bava, A. (2004). The posterior fossa. Journal of Neuroimaging, 14(3), 197–207. https://doi.org/10.1111/j.1552-6569.2004.tb00243.x
9. Bartley, A., Jones, A., Davis, J., & Hardy, M. (2023). Motion artefact in head CT: Systematic review of prevalence, causes and mitigation strategies. Radiography, 29(4), 756–763. https://doi.org/10.1016/j.radi.2023.04.006
10. Chandran, M. O., Pendem, S., Chacko, C., & Kadavigere, R. (2024). Influence of deep learning image reconstruction algorithm for reducing radiation dose and image noise compared to iterative reconstruction and filtered back projection for head and chest CT examinations: A systematic review. F1000Research, 13, 274. https://doi.org/10.12688/f1000research.147345.1
11. Wiesmann, M. (2022). Photon-counting CT: A quantum leap in diagnostic imaging? Clinical Neuroradiology, 32(2), 309–312. https://doi.org/10.1007/s00062-022-01145-2
12. Pereira, R., Sousa, N., Pinto, M., Santos, L., & Fernandes, L. (2025). Radiation dose reduction in CT exams with iterative and deep learning reconstruction: A systematic review. Applied Sciences, 16(1), 316. https://doi.org/10.3390/app16010316
13. European Commission. (2022). Radiation protection No. 185: European guidelines on diagnostic reference levels for paediatric imaging. Publications Office of the European Union. https://doi.org/10.2833/527874
14. McCollough, C. H., Bruesewitz, M. R., & Kofler, J. M. (2006). CT dose reduction and dose management tools: Overview of available options. RadioGraphics, 26(2), 503–512. https://doi.org/10.1148/rg.262055138
15. International Commission on Radiological Protection. (2017). ICRP Publication 135: Diagnostic reference levels in medical imaging. Annals of the ICRP, 46(1). https://doi.org/10.1177/0146645317717209
16. Hemphill, J. C., Greenberg, S. M., Anderson, C. S., Becker, K., Bendok, B. R., Cushman, M., Fung, G. L., Goldstein, J. N., Macdonald, R. L., Mitchell, P. H., Scott, P. A., Selim, M. H., & Woo, D. (2015). Guidelines for the management of spontaneous intracerebral hemorrhage: A guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke, 46(7), 2032–2060. https://doi.org/10.1161/STR.0000000000000069
17. Toma, A. K., Stapleton, S., Papadopoulos, M. C., Kitchen, N. D., & Watkins, L. D. (2011). Natural history of hydrocephalus after aneurysmal subarachnoid hemorrhage. Neurosurgical Review, 34(2), 217–224. https://doi.org/10.1007/s10143-010-0303-4
18. Steiner, T., Juvela, S., Unterberg, A., Jung, C., Forsting, M., & Rinkel, G. (2013). European Stroke Organization guidelines for the management of intracranial aneurysms and subarachnoid haemorrhage. Cerebrovascular Diseases, 35(2), 93–112. https://doi.org/10.1159/000346087
19. Soun, J. E., Chow, D. S., Nagamine, M., Takhtawala, R. S., Filippi, C. G., Yu, W., & Chang, P. D. (2021). Artificial intelligence and acute stroke imaging. American Journal of Neuroradiology, 42(1), 2–11. https://doi.org/10.3174/ajnr.A6883
20. Jadhav, A. P., Desai, S. M., & Jovin, T. G. (2021). Indications for mechanical thrombectomy for acute ischemic stroke: Current guidelines and beyond. Neurology, 97(20 Suppl 2), S126–S136. https://doi.org/10.1212/WNL.0000000000012801
21. Garg, R., Oh, E., Naidech, A., Katsaggelos, A., & Lieber, R. L. (2018). Automated identification of moderate and large pneumocephalus and intracerebral hemorrhage on CT. Journal of the American Medical Association Network Open, 1(3), e180273. https://doi.org/10.1001/jamanetworkopen.2018.0273
22. Straka, M., Albers, G. W., & Bammer, R. (2023). Real-time diffusion-perfusion mismatch analysis in acute stroke. Journal of Magnetic Resonance Imaging, 32(5), 1024–1037. https://doi.org/10.1002/jmri.22338
23. Lövblad, K. O., Baird, A. E., Schlaug, G., Benfield, A., Siewert, B., Voetsch, B., Connor, A., Burzynski, C., Edelman, R. R., & Warach, S. (2016). Ischemic lesion volumes in acute stroke by diffusion-weighted magnetic resonance imaging correlate with clinical outcome. Annals of Neurology, 42(2), 164–170. https://doi.org/10.1002/ana.410420206
24. Kidwell, C. S., Chalela, J. A., Saver, J. L., Starkman, S., Hill, M. D., Demchuk, A. M., Butman, J. A., Patronas, N., Alger, J. R., Latour, L. L., Luby, M. L., Baird, A. E., Leary, M. C., Tremwel, M., Ovbiagele, B., Fredieu, A., Suzuki, S., Villablanca, J. P., Davis, S., … Warach, S. (2004). Comparison of MRI and CT for detection of acute intracerebral hemorrhage. JAMA, 292(15), 1823–1830. https://doi.org/10.1001/jama.292.15.1823
25. Demeestere, J., Wouters, A., Christensen, S., Lemmens, R., & Lansberg, M. G. (2020). Review of perfusion imaging in acute ischemic stroke: From time to tissue. Stroke, 51(3), 1017–1024. https://doi.org/10.1161/STROKEAHA.119.028337
26. Villablanca, J. P., Jahan, R., Hooshi, P., Lim, S., Duckwiler, G., Patel, A., Sayre, J., Martin, N., Frazee, J., Bentson, J., & Vinuela, F. (2002). Detection and characterization of very small cerebral aneurysms by using 2D and 3D helical CT angiography. American Journal of Neuroradiology, 23(7), 1187–1198.
27. Mozaffarian, D., Benjamin, E. J., Go, A. S., Arnett, D. K., Blaha, M. J., Cushman, M., Das, S. R., de Ferranti, S., Després, J. P., Fullerton, H. J., Howard, V. J., Huffman, M. D., Isasi, C. R., Jiménez, M. C., Judd, S. E., Kissela, B. M., Lichtman, J. H., Lisabeth, L. D., Liu, S., … Turner, M. B. (2016). Heart disease and stroke statistics — 2016 update: A report from the American Heart Association. Circulation, 133(4), e38–e360. https://doi.org/10.1161/CIR.0000000000000350