Master the high-resolution chest CT (HRCT) protocol: complete scanning parameters, precise HU reference values, mandatory expiration technique, top 10 interstitial pathologies, and a three-tier pitfall framework for radiographers, radiologists, and referring clinicians — aligned with the latest ACR, RSNA, ATS, ERS, and ICRP guidance.

At a glance: HRCT protocol snapshot

Introduction: why HRCT is the gold standard for ILD

High-resolution chest CT — the HRCT protocol — occupies a singular position in thoracic radiology. Unlike routine contrast-enhanced chest CT, which is optimised for mediastinal and vascular structures, HRCT sacrifices volumetric coverage for exquisite spatial resolution of the lung parenchyma itself, yielding 0.625 mm to 1.25 mm slices reconstructed with a sharp edge-enhancing kernel that renders interlobular septa, alveolar walls, and bronchiolar structures with a level of anatomical fidelity previously achievable only at post-mortem examination.^[1] This capability makes HRCT the internationally recommended first-line imaging investigation for suspected interstitial lung disease (ILD), pulmonary fibrosis, emphysema characterisation, and small airway pathology.

The global burden of ILD is substantial and growing. Idiopathic pulmonary fibrosis (IPF) alone carries a median survival of only two to three years from diagnosis, and its optimal management — including the timely initiation of antifibrotic therapy with nintedanib or pirfenidone — depends critically on confident HRCT characterisation of the usual interstitial pneumonia (UIP) pattern.^[2] In this context, every technical decision made during an HRCT acquisition — from the choice of reconstruction kernel to the timing of the expiratory phase — directly impacts diagnostic accuracy and, ultimately, patient outcomes.

This article presents a comprehensive, protocol-level analysis of the HRCT chest examination. It covers the precise scanning parameters established by the Day 8 protocol matrix, detailed Hounsfield Unit (HU) reference values for normal and abnormal lung structures, a seven-step scanning workflow with scanner generation comparisons, the rationale for non-contrast acquisition, radiation dose management aligned to European Commission RP 185 and AAPM guidance, the ten most clinically significant pathologies with their HRCT signatures, and a structured three-tier pitfall framework covering radiographers, radiologists, and referring physicians. AI-assisted detection tools with regulatory clearance are also reviewed.

Anatomy and HU values in the thorax

A thorough command of thoracic anatomy at the resolution that HRCT affords is the single most powerful tool a radiologist or reporting radiographer possesses. At sub-millimetre slice thickness, the lung parenchyma reveals its architectural hierarchy in extraordinary detail — from the lobular level down to individual alveolar walls — and every pathological process manifests as a deviation from the precisely defined HU baseline of aerated, healthy lung tissue.

Master HU reference table for thoracic HRCT

The secondary pulmonary lobule: the unit of HRCT interpretation

The secondary pulmonary lobule (SPL), measuring approximately 1–2.5 cm in diameter, is the fundamental anatomical unit visible on HRCT and the framework for pattern-based ILD diagnosis. The SPL is bounded by interlobular septa — fibrous partitions containing lymphatics and pulmonary venules — that appear as fine linear structures of approximately −50 to +50 HU on HRCT when normal and as clearly thickened lines (1–2 mm width) when oedematous or infiltrated by tumour.

At the centre of the SPL runs the centrilobular artery and its accompanying core bronchiole. The dot-like centrilobular artery (typically +30–50 HU) is normally visible on thin-section HRCT at 3–5 mm from the pleural surface. Its abnormal appearance — tree-in-bud opacities — indicates endoluminal material such as mucus, pus, or tumour cells in small airways, a pattern with a robust differential diagnosis including endobronchial infection, aspiration, and panbronchiolitis.

The lung periphery: subpleural anatomy and pathological predilection

The subpleural region, within 1–2 cm of the visceral pleural surface, has a preferential vulnerability to fibrotic processes including IPF, asbestosis, and connective tissue disease (CTD)-related ILD. The visceral pleura itself measures less than 1 mm and is not directly visible on standard HRCT, though pleural tags — linear fibrous strands connecting a lung nodule to the pleura — are reliably seen and carry diagnostic significance in the evaluation of peripheral lung adenocarcinoma.

Understanding that honeycombing is, by definition, a subpleural, basal-predominant phenomenon provides the interpretive logic underpinning the UIP classification in IPF guidelines.^[4] Equally, the peribronchovascular distribution of sarcoidosis — hugging the bronchovascular bundles along their length — and the diffuse, random distribution of miliary nodules in disseminated tuberculosis or metastatic disease each reflect the specific routes by which disease spreads through the lung’s anatomical compartments.

The airways: bronchial anatomy on HRCT

HRCT routinely resolves airways down to the fifth or sixth generation of branching. The normal bronchial wall-to-lumen ratio is less than 0.18; bronchiectasis is defined radiologically as a bronchial internal diameter exceeding the diameter of the adjacent pulmonary artery (signet-ring sign). The anatomical basis of traction bronchiectasis — where progressive fibrosis physically distorts and dilates the bronchial lumen — is readily appreciated on coronal reformats and is one of the most reliable indicators of established, irreversible fibrosis in all ILD subtypes.

Scanning technique: 7 essential steps

HRCT chest is a protocol with minimal tolerance for compromise. Unlike conventional chest CT — where generous scan coverage and moderate spatial resolution serve mediastinal and vascular evaluation well — HRCT demands that every technical parameter be optimised specifically for pulmonary parenchymal resolution. The following seven steps represent the evidence-based standard of practice endorsed by the ACR, the ESR Chest Subcommittee, and the technical guidance documents of major CT manufacturers.

1. Patient positioning and breath-hold instruction. The patient lies supine, arms elevated above the head to minimise beam-hardening artefact from the humeri across the chest wall. Before scanning, a dedicated breath-hold instruction is given and practised — ideally with an audio coaching system. Patients inspire to full lung capacity and hold for the inspiratory acquisition. For the expiratory phase, they exhale as completely as possible, which in practice requires a verbal cue (“breathe all the way out and hold”) and a minimum of two rehearsal attempts. Failed or incomplete breath-holds are the single most common technical cause of diagnostic degradation on HRCT.^[5]
2. Scout and field of view planning. Acquire a low-dose PA digital radiograph (scout) to set the scan range from the lung apices to the costophrenic recesses. The FOV should encompass the full thoracic cage without wasting coverage on subdiaphragmatic structures. Use a body-to-skin FOV on acquisition, then reconstruct to a lung-FOV (typically 350–400 mm) to maximise in-plane spatial resolution.
3. Inspiratory acquisition. Perform the primary acquisition at full inspiration using the parameters specified in the Day 8 protocol matrix: 120 kVp, pitch 1.2, 100–150 mA, 0.5 s rotation time. Use a high-spatial-frequency (sharp/bone) reconstruction kernel — typically B70f on Siemens Healthineers, LUNG on GE Healthcare, FC51 on Canon Medical, or SHARP on Philips. Slice thickness of 0.625–1.25 mm with reconstruction interval of 0.5–1.0 mm. Window the lung images at W:1500 / L:−600 for parenchymal assessment and W:350 / L:40 for mediastinal review.
4. Expiratory acquisition — the critical addition. A second, low-dose scan series is acquired immediately following the inspiratory series, during maximal breath-hold exhalation. This expiratory series is acquired from the carina to the costophrenic recesses (not necessarily the full lung). At lower radiation burden (80–100 mA is acceptable given the high native contrast of air-trapping), it provides the definitive evidence for air trapping, confirming small airway obstruction. In normal individuals, lung attenuation increases by more than 100 HU from inspiration to expiration. Regions that fail to densify — remaining below −900 HU — represent air trapping. This finding is the imaging hallmark of hypersensitivity pneumonitis, constrictive bronchiolitis, and sarcoidosis with small airway involvement, and cannot be inferred from inspiratory images alone.^[6]
5. Prone series (if required). When dependent atelectasis in the posterior lung bases mimics ground-glass opacity in a supine patient, a limited prone acquisition of 20–30 slices through the lower lobes quickly resolves the ambiguity. Dependent atelectasis resolves completely in the prone position; true pathological GGO persists. This is particularly valuable in the evaluation of early ILD where differentiating minimal physiological dependency from early UIP is clinically decisive.
6. Reconstruction kernels and image post-processing. Reconstruct all HRCT series with a high-frequency (sharp) kernel to maximise edge contrast for fine lung structures. If the scanner has deep learning reconstruction (DLR) capability, apply vendor-specific DLR (see dedicated DLR section below). Additionally, reconstruct a medium-soft kernel series (B30f, STANDARD, FC12, or B) for mediastinal assessment. Minimum intensity projections (MinIP) over 5–10 mm slabs are invaluable for visualising the tree-in-bud pattern and centrilobular emphysema in their spatial context. Maximum intensity projections (MIP) over 5 mm slabs help highlight micronodules in sarcoidosis and lymphangitic carcinomatosis.
7. Slice review sequence and structured reporting. Review on thin-section lung windows (W:1500/L:−600), then mediastinal windows (W:350/L:40). Systematically evaluate: (a) distribution of parenchymal abnormality — upper versus lower, central versus peripheral, subpleural versus peribronchovascular; (b) dominant pattern — GGO, consolidation, reticulation, nodules, honeycombing, air trapping; (c) secondary findings — pleural disease, lymphadenopathy, cardiac size, pulmonary artery calibre. Dictate using a structured ILD template with mandatory comment on UIP pattern presence, air trapping on expiratory series, and comparison with any prior imaging.

Scanner comparison table: 16-slice to 320-slice

Dual-energy and photon-counting HRCT

Dual-energy CT (DECT) of the lung offers two complementary applications relevant to ILD practice. First, virtual monoenergetic images (VMI) can suppress beam-hardening artefacts from ribs and mediastinal structures, improving visibility at the lung bases and posterior costophrenic recesses. Second, iodine maps derived from DECT can quantify perfusion abnormalities in conditions such as hypersensitivity pneumonitis, where mosaic attenuation reflects a combination of air trapping and perfusion redistribution.

Photon-counting CT (PCCT), now entering clinical deployment at major academic centres, represents the next threshold in HRCT capability. By detecting individual photons without electronic noise, PCCT achieves in-plane spatial resolution of 0.2 mm (ultra-high-resolution mode), sufficient to resolve intralobular reticular changes that represent the earliest phase of fibrotic ILD — changes that fall below the resolution threshold of conventional MDCT.^[7] Emerging data suggest that PCCT may change the threshold for biopsy avoidance by enabling earlier confident UIP diagnosis.

Deep learning reconstruction (DLR) for HRCT

Conventional iterative reconstruction algorithms — iDose (Philips), AIDR 3D (Canon), ASIR-V (GE), and ADMIRE (Siemens) — reduce noise at lower mA settings by smoothing the image, which is counterproductive for the edge-preserving requirements of HRCT. Deep learning reconstruction (DLR), validated by multiple independent studies in thoracic CT, addresses this trade-off directly.^[8]

Vendor DLR packages — TrueFidelity (GE), AiCE (Canon), Precise Image (Philips), and Deep Resolve (Siemens Healthineers) — train convolutional neural networks on paired high-dose/low-dose image datasets and learn to restore fine structural detail rather than merely suppress noise statistics. In HRCT practice, DLR allows equivalent diagnostic quality at 30–50% lower radiation dose compared to conventional filtered back projection (FBP) at the same reconstruction sharpness. For surveillance HRCT in patients with established ILD — who may receive scans annually or biannually over many years — this dose compounding benefit is clinically meaningful.

Contrast media protocol: why HRCT is non-contrast

The Day 8 HRCT protocol specifies zero contrast volume, zero flow rate, and zero saline chaser — and this is not a default or an omission. The deliberate decision to perform HRCT without intravenous contrast media is grounded in the diagnostic logic of the examination itself and is fully aligned with ACR Appropriateness Criteria and ESR guidance.

The pathological signatures targeted by HRCT — honeycombing, GGO, traction bronchiectasis, interlobular septal thickening, air trapping, tree-in-bud opacities, and centrilobular emphysema — are purely parenchymal phenomena defined by their attenuation relative to aerated lung and adjacent fibrous tissue. Contrast enhancement adds no diagnostic value to these features and, critically, obscures them by raising the overall mediastinal and vascular attenuation, which degrades the window level settings needed for optimal lung parenchymal review.

Contrast safety check — when contrast IS subsequently indicated

Following HRCT, the clinical team may request an additional contrast-enhanced series in the same session — for example, to evaluate a dominant mass lesion, assess pleural enhancement, or characterise mediastinal lymphadenopathy. In this circumstance, the following safety framework applies:

Radiation dose and dose reduction strategies

HRCT presents a particular radiation stewardship challenge. The high-spatial-frequency reconstruction kernels used to resolve lung microstructure are inherently noise-amplifying, which historically forced radiologists to accept either high tube current (and therefore high dose) or suboptimal noise. Modern dose optimisation strategies have transformed this trade-off, but the fundamental principle remains: in patients who will undergo repeat HRCT surveillance over years to decades, cumulative dose warrants structured, protocol-level management.

Diagnostic reference levels for chest HRCT

5 evidence-based dose reduction strategies for HRCT

1. Automatic exposure control (AEC) with angular modulation. Vendor AEC systems — CARE Dose4D (Siemens), Auto mA/Smart mA (GE), DoseRight (Philips), SureExposure (Canon) — modulate tube current in real time based on attenuation feedback from the scout image. In the thorax, the lung field offers minimal attenuation, so AEC naturally reduces mA in the lung relative to the denser shoulders and mediastinum, achieving 20–40% dose saving without compromising parenchymal image quality.
2. Deep learning reconstruction (DLR) in place of FBP. As discussed in the scanning technique section, DLR achieves equivalent diagnostic quality at 30–50% lower mA than FBP. For ILD surveillance programmes, this compounds to clinically significant lifetime dose reductions.
3. Tube voltage optimisation: 100 kVp for patients ≤90 kg. Reducing kVp from 120 to 100 decreases dose by approximately 25–30% (dose scales approximately with kVp²). In the lung — where the high inherent contrast between aerated parenchyma and soft tissue means photon energy selection has less impact on contrast-to-noise ratio than in abdominal CT — 100 kVp is diagnostically equivalent to 120 kVp for most patients.^[9]
4. Limited expiratory acquisition coverage. The expiratory series need not replicate the full inspiratory coverage. Limiting the expiratory scan to the carina–costophrenic angle range (approximately 15 cm coverage) captures the regions where air trapping is clinically relevant and reduces the expiratory series DLP by 30–40% compared to full-chest expiration.
5. Elimination of clinically redundant series. Review the clinical referral carefully. If the question is purely one of ILD pattern characterisation in a known case, the mediastinal reconstruction is informational only and the dose contribution of a dedicated mediastinal series can be eliminated by reconstructing multiple kernel variants from a single raw dataset. Modern CT allows simultaneous reconstruction of lung-kernel, mediastinal-kernel, and DLR-processed series from a single acquisition at no additional radiation cost.

Top 10 pathologies detected on HRCT

The HRCT protocol is optimised for parenchymal lung pathology, and the conditions below represent the most clinically significant diagnoses for which HRCT provides decisive or diagnostic-grade imaging evidence. Each entry specifies the characteristic HU range, dominant HRCT pattern, and the specific protocol adaptations that maximise detection.

Pitfalls for radiographers

The primary scanning pitfall for the HRCT protocol is explicitly defined in the Day 8 protocol matrix: scanning purely in inspiration. Air trapping due to small airway disease can only be accurately confirmed by adding mandatory end-expiratory scans — and when this acquisition is omitted, one of the most diagnostically significant findings in ILD — and the principal discriminator between HP and IPF — is rendered invisible.

Pitfalls for radiologists

The interpretation pitfall column for Day 8 in the protocol matrix does not carry a named entry — an important observation in itself. The Illustration_Pitfall field for HRCT is: Dependent atelectasis — normal, unventilated lung bases in a supine patient can mimic ground-glass changes. Mitigation: rescan in a prone position. This is the most operationally significant interpretation error in HRCT, with a high frequency of occurrence in daily practice and profound consequences for patient management.

Pitfalls for non-radiology physicians

Pitfall comparison summary

AI and automation in HRCT

Artificial intelligence has achieved its most mature and clinically validated applications in thoracic CT among all radiology subspecialties, and the assessment of interstitial lung disease on HRCT is a domain where AI tools have moved decisively from research to regulated clinical practice. The evidence base is substantial and the regulatory landscape is evolving rapidly.

FDA-cleared and CE-marked AI tools for HRCT analysis

AI for expiratory air trapping detection

The identification and quantification of air trapping on expiratory HRCT has traditionally relied on visual estimation — an inherently subjective and reader-dependent process. AI-driven density histogram analysis tools, including those embedded in VIDA Diagnostics and IMBIO platforms, now provide automated quantification of the percentage lung volume with expiratory attenuation below −856 HU (the published threshold for air trapping per the Fleischner Society Technical Standards).^[12]

Studies demonstrate that when this threshold is exceeded in ≥32% of total lung volume on expiratory HRCT, the sensitivity and specificity for HP-related air trapping versus IPF-type disease improves to 84% and 91%, respectively — a discriminative accuracy that exceeds expert visual assessment. Integration of these tools into clinical practice at the report generation level holds meaningful potential to reduce diagnostic uncertainty in the HP versus IPF distinction, which remains one of the most consequential diagnostic challenges in respiratory medicine.

Deep learning for ILD progression monitoring

Serial HRCT monitoring of ILD — tracking honeycombing extent, fibrosis volume, and GGO regression over treatment — is an area where AI-assisted longitudinal registration is transforming clinical trial design and individual patient follow-up. By deformably registering serial HRCT volumes to the same anatomical reference, platforms such as the Boehringer Ingelheim / VIDA collaboration quantify fibrosis change down to 2–3% of total lung volume per year — well below the threshold detectable by visual comparison.^[13] This granularity enables earlier detection of antifibrotic treatment response or failure, with implications for treatment switching decisions.

Conclusion

High-resolution chest CT occupies an irreplaceable position at the centre of interstitial lung disease diagnosis. The Day 8 protocol — 120 kVp, pitch 1.2, 100–150 mA, 0.5 s rotation, non-contrast, with mandatory inspiratory and expiratory acquisitions — distils decades of clinical evidence into a set of parameters that, when executed precisely, transform a CT scanner into the most powerful non-invasive diagnostic tool available for lung parenchymal pathology.

The anatomical precision of HRCT — resolving structures from the secondary pulmonary lobule level, through individual bronchiolar generations, down to interlobular septa — demands a correspondingly precise level of technical execution. The reconstruction kernel, the breath-hold quality, the slice thickness, and the critical inclusion of the expiratory series are not optional refinements: they are the protocol. An HRCT performed without an expiratory series is, for the purposes of HP versus IPF discrimination, an incomplete examination.

The ten pathologies reviewed — IPF/UIP, NSIP, sarcoidosis, hypersensitivity pneumonitis, lymphangitic carcinomatosis, bronchiectasis, emphysema, septal thickening, COP, and silicosis — span the full spectrum of the HRCT indication landscape. Each has a defined HU fingerprint, a characteristic architectural distribution, and a set of protocol adaptations that maximise detection. The three-tier pitfall framework — covering technical failures for radiographers, interpretive errors for radiologists, and clinical reasoning gaps for referring physicians — provides the quality governance infrastructure needed to reduce diagnostic failure across the full ILD care pathway.

Emerging AI platforms, from VIDA Diagnostics to IMBIO Lung Texture Analysis, are adding quantitative precision to subjective pattern recognition, and photon-counting CT is beginning to extend the spatial resolution envelope of what HRCT can resolve. As these technologies mature, the radiographer, radiologist, and referring physician communities must evolve together — maintaining the technical rigour and interpretive discipline that characterise expert HRCT practice as the tools that support it become increasingly capable.

Medically Reviewed by Prof. Dr. Damien O’Neil, MD, PhD

Last updated: 17 June 2025 | Reviewed for clinical accuracy and adherence to the latest guidelines of the American Thoracic Society (ATS), European Respiratory Society (ERS), European Society of Radiology (ESR), American College of Radiology (ACR), Radiological Society of North America (RSNA), the International Commission on Radiological Protection (ICRP), and the Fleischner Society.

This article is intended for healthcare professionals and hospital administration. It does not constitute individual clinical advice. Clinical decisions should be made in consultation with qualified medical practitioners and in accordance with institutional protocols.