A dedicated CT renal mass protocol guide for radiographers and radiologists: the triple-phase technique, HU thresholds that separate cysts from tumors, and the pitfalls that lead to missed cancers.

CT Renal Mass Protocol: 7 Steps to Nail the Triple-Phase Scan

Introduction: why renal mass CT is unforgiving

A CT renal mass protocol is one of the few abdominal CT examinations where a single missed acquisition phase can permanently destroy diagnostic value. Unlike a routine abdomen-pelvis scan, where a slightly mistimed portal venous phase is merely suboptimal, a dedicated renal mass study depends entirely on a true baseline. If the non-contrast phase is skipped, rushed, or improperly windowed, there is no way to retroactively calculate enhancement — the entire physiologic logic of the exam collapses, and the patient may need to return for a repeat scan with additional radiation and contrast exposure.

The clinical stakes are high. Renal masses are detected incidentally on a substantial proportion of abdominal CT scans performed for unrelated indications, and the radiologist’s job is to separate the overwhelming majority that are benign simple cysts from the minority that are renal cell carcinoma (RCC) or another malignancy requiring urological referral. That separation is made almost entirely on the basis of quantitative HU enhancement measured across three carefully timed phases: non-contrast, corticomedullary, and nephrographic.

The remainder of this guide is organized to follow the natural workflow of a renal mass study from first principles through to final report: the anatomy and physiology that explain why the protocol is built the way it is, the step-by-step scanning technique itself, the contrast and dose considerations that shape day-to-day execution, the pathologies the protocol is designed to characterize, and finally the three distinct categories of pitfall — scanning, interpretation, and clinical — that determine whether all of that careful technical work actually translates into a correct diagnosis and an appropriate clinical outcome for the patient.

This guide walks through the full triple-phase acquisition, the Hounsfield Unit values that anchor every diagnostic decision, the top ten pathologies a radiographer and radiologist will encounter, and — critically — the distinct failure modes that trip up scanning technologists, interpreting radiologists, and the non-radiology physicians who order the study and act on the report. Each of these three groups makes characteristically different mistakes, and understanding all three is what separates a department that catches early-stage RCC from one that doesn’t.

It is worth pausing on why this protocol exists as a distinct entity at all, separate from a routine abdomen-pelvis CT. Most abdominal CT protocols are designed to answer a binary or categorical question — is there a perforation, an abscess, a bowel obstruction — where a single well-timed phase usually suffices. Renal mass characterization is fundamentally a quantitative, comparative task: the radiologist is not simply looking for a lesion, but measuring how much a specific region of tissue changes in attenuation across time. That comparative requirement is precisely what drives every design decision in the protocol, from the mandatory non-contrast baseline to the fixed (rather than bolus-tracked) delay timings for the two contrast phases.

The downstream consequences of getting this protocol right are significant both clinically and operationally. Clinically, a correctly executed triple-phase study allows many incidentally detected renal lesions to be confidently classified as Bosniak I simple cysts requiring no further action, sparing patients unnecessary anxiety, additional imaging, and invasive workup. Operationally, every study that has to be repeated because a phase was skipped or mistimed consumes scanner time, contrast inventory, and — most importantly — exposes the patient to avoidable additional radiation. A 2021 survey of academic abdominal radiologists and urologists found persistent variability in how renal mass imaging findings are reported and acted upon, underscoring that the technical and interpretive components of this protocol cannot be considered in isolation from the clinical communication that follows.^[23]

Renal anatomy and HU reference values

The kidney is organized into three radiologically distinct zones that behave differently across the contrast bolus, and recognizing each zone on every phase is the foundation of accurate interpretation. The renal cortex is the outer, highly vascular layer containing the glomeruli; it enhances earliest and most intensely because it receives the lion’s share of renal blood flow. The renal medulla, composed of the pyramids and loops of Henle, enhances later and less intensely, creating the characteristic corticomedullary differentiation seen on early post-contrast images. The renal sinus contains the collecting system, major vessels, and peripelvic fat, and is the zone most often confused with pathology when fat-containing or fluid-containing structures are misread as masses.

Gross anatomy in cross-section

On axial imaging, each kidney sits in the retroperitoneum bounded posteriorly by the psoas and quadratus lumborum muscles and anteriorly by the peritoneal reflection. The right kidney typically sits slightly lower than the left due to hepatic mass effect. The Gerota’s fascia envelopes the kidney and perinephric fat, forming a contained compartment that is clinically important for staging renal malignancy — invasion through Gerota’s fascia upstages a tumor and changes surgical planning. The renal artery and vein emerge from the renal sinus at the hilum, with the vein typically anterior to the artery on the left and the relationship more variable on the right; venous tumor thrombus extending into the renal vein or inferior vena cava is a key finding that must be specifically searched for in any suspected RCC.

The corticomedullary and nephrographic phases explained

The corticomedullary phase (acquired around 25–70 seconds post-injection, fixed at 50 seconds in this protocol) captures peak cortical enhancement while the medulla remains relatively unenhanced. This phase is optimal for detecting hypervascular tumors such as clear cell RCC, which often enhance more avidly than the surrounding cortex, and for mapping the renal arterial anatomy prior to partial nephrectomy or ablation. The nephrographic phase (75–120 seconds, fixed at 100 seconds here) is acquired once contrast has filtered through the glomeruli and both cortex and medulla enhance homogeneously. This homogeneous background is what makes the nephrographic phase the single best phase for detecting a renal mass — a non-enhancing or hypoenhancing lesion stands out sharply against uniformly enhanced parenchyma in a way it would not against the heterogeneous cortex-medulla pattern of the corticomedullary phase.

Relevant clinical anatomy: the renal sinus fat plane

The fat plane surrounding the renal sinus structures is a frequently overlooked but diagnostically important landmark. Loss of this fat plane, or stranding within it, can indicate either inflammatory disease (pyelonephritis) or local tumor extension. Careful evaluation of the sinus fat on the nephrographic phase, where surrounding parenchyma is brightly and uniformly enhanced, makes subtle infiltration far easier to detect than on the non-contrast or corticomedullary series alone.

The physiology behind the three-phase timing

Understanding why the protocol uses 50 and 100 seconds — rather than any other pair of delays — requires understanding renal blood flow physiology. The kidneys receive approximately 20–25% of cardiac output despite representing under 1% of total body mass, and roughly 90% of that renal blood flow is directed to the cortex. This disproportionate cortical perfusion is why the corticomedullary phase, timed to the first circulatory pass of the contrast bolus through the renal artery, shows such a dramatic cortex-medulla enhancement gradient. By 75–120 seconds, contrast has been filtered by the glomeruli, passed into the tubules, and recirculated systemically enough that cortex and medulla enhance more uniformly — the nephrographic phase. Because this physiological transition happens at a fairly predictable rate across patients with normal cardiac output and renal function, fixed time delays produce reproducible results without requiring per-patient bolus tracking, which is reserved for angiographic protocols where peak intravascular opacification (rather than tissue-phase enhancement) is the target.

This physiological basis also explains why patient-specific factors matter so much for protocol accuracy. A patient with reduced ejection fraction, significant aortic stenosis, or peripheral vascular disease will have a slower circulation time, meaning the bolus arrives at the kidney later relative to the injection start. If the fixed delays are applied uniformly without clinical adjustment, the corticomedullary phase in such a patient may actually be acquired during what would normally be an early nephrographic-type enhancement pattern, blurring the distinction the protocol depends on. Reviewing the corticomedullary scout image in real time — rather than blindly trusting the timer — is the technologist’s primary defense against this scenario.

Differentiating renal sinus fat from a fat-containing mass

Because angiomyolipoma is defined by macroscopic fat, and the renal sinus itself is a fat-filled structure, there is a recurring source of ambiguity at the corticomedullary junction where sinus fat abuts the medulla. A true AML will show a discrete, rounded fat-density focus arising from the renal parenchyma itself, often with adjacent vessels and smooth muscle components visible as soft-tissue density within the fat. Renal sinus fat, by contrast, follows the expected branching distribution of the collecting system and vasculature and lacks a discrete parenchymal origin. Multiplanar reformats are particularly useful for resolving this distinction, since a single axial slice can make sinus fat appear to “bulge” into the parenchyma in a way that mimics a mass.

The Bosniak classification system in detail

No discussion of renal mass HU values is complete without a working knowledge of the Bosniak classification, the risk-stratification system that converts the appearance of a cystic renal lesion into an actionable management category. The 2019 update refined the original system to better incorporate modern multiphase CT and MRI technique and to reduce interreader variability that had been documented with the earlier 2005 version.^[1]

The practical value of this system for the renal mass protocol described in this guide is direct: every Bosniak category boundary depends on the radiologist’s ability to confidently determine whether true enhancement is present in a wall, septum, or solid component, which in turn depends entirely on having a reliable non-contrast baseline and at least one well-timed post-contrast phase to compare against it. A protocol that fails to deliver a clean non-contrast series does not just make the radiologist’s job harder in the abstract — it can mechanically prevent accurate Bosniak categorization for any cystic lesion identified on that study.

It is also worth noting that the Bosniak system, despite its widespread adoption, was developed and refined primarily using CT data, and its translation to MRI — while increasingly well validated — still carries some category-specific caveats around septal counting and enhancement perception that differ subtly between modalities. For a department that occasionally needs to cross-reference a CT-based Bosniak category against a follow-up MRI, awareness of these modality-specific nuances helps avoid over-interpreting a minor category shift between studies as true lesion progression when it may simply reflect a difference in how each modality visualizes thin septa or perceives low-level enhancement.

A representative case walkthrough

Consider a 58-year-old patient referred for a dedicated renal mass protocol after a 2.3 cm right renal lesion was noted incidentally on a single-phase CT performed for unrelated abdominal pain. On the non-contrast phase, the lesion measures 38 HU — already higher than the <20 HU expected for a simple cyst, raising suspicion that this could be a hyperdense cyst or a solid lesion. On the corticomedullary phase at 50 seconds, the same ROI measures 96 HU, a rise of 58 HU from baseline — well above the 15–20 HU threshold and far higher than would be expected from pseudo-enhancement alone. On the nephrographic phase at 100 seconds, the lesion measures 71 HU, showing partial washout relative to the corticomedullary peak. This avid-enhancement-then-washout pattern, in a solid (non-cystic) lesion, is classic for clear cell RCC and would prompt urological referral, with the imaging report explicitly stating the three-phase HU values to support the radiologist’s conclusion. Without the non-contrast baseline of 38 HU, the post-contrast values alone could not have distinguished this lesion from a hyperdense but entirely benign cyst that simply appeared bright on a single contrast-enhanced phase.

Now contrast this with a second, far more common scenario: a 1.4 cm left renal lesion identified incidentally on the same kind of single-phase scan, measuring 12 HU on the dedicated protocol’s non-contrast phase. On the corticomedullary phase, the same ROI measures 14 HU; on the nephrographic phase, 13 HU. The total variation across all three phases is well within measurement noise, and the lesion shows a sharply marginated, homogeneous, water-attenuation appearance with no septa or wall thickening on any phase. This is the signature of a Bosniak I simple cyst, and the report can confidently state that no further imaging follow-up is required. The contrast between these two cases illustrates exactly why the protocol’s value lies less in any single phase and more in the relationship between all three: an isolated 71 HU or 13 HU measurement on its own would be far less informative than the trajectory each lesion traces from non-contrast through corticomedullary to nephrographic phase.

Scanning technique: the 7-step protocol

Executing a dedicated renal mass CT reliably requires a fixed sequence that does not vary by operator. The following 7 steps reflect the workflow used across the protocol parameters in this series. While the individual parameters — kVp, pitch, mA, rotation time — are standard CT acquisition variables familiar from any abdominal protocol, what distinguishes this examination is the rigid sequencing and timing discipline layered on top of them. A technologist who is excellent at routine abdominal CT but unfamiliar with the specific rationale behind this protocol’s structure can still make the most common and most damaging error in this exam: treating the non-contrast phase as optional or skippable.

1. Patient preparation and positioning. Confirm renal function (eGFR) and contrast allergy history before proceeding. Position the patient supine, arms raised above the head to eliminate streak artifact across the kidneys, and center the coronal isocenter at the level of the xiphoid to mid-iliac crest to capture both kidneys and the bladder base. Patients who cannot raise their arms due to shoulder pathology or frailty should have at least the contralateral arm raised, with awareness that some streak artifact may remain.
2. Acquire the non-contrast baseline. Scan from the diaphragm through the iliac crests at 120 kVp, 200–300 mA with a pitch of 1.0 and 0.5 s rotation time. This phase is non-negotiable — review every lesion’s baseline HU before injecting contrast. If a suspicious lesion is identified on this phase, place and record the ROI coordinates so the same anatomical location can be re-measured precisely on subsequent phases.
3. Establish IV access and load the injector. Use an 18–20 gauge antecubital line capable of sustaining a 3.5 mL/s flow rate. Load 100 mL of low- or iso-osmolar contrast medium followed by a 100 mL saline chaser to maximize bolus geometry and minimize streak artifact from undiluted contrast in the SVC. Test the line with a small saline flush before connecting to the power injector to confirm patency and rule out extravasation risk, particularly in older or fragile veins.
4. Acquire the corticomedullary phase at 50 seconds. This fixed delay (rather than bolus tracking) is standard for renal mass protocols because the goal is reproducible cortical timing across patients rather than aortic peak opacification. Confirm strong cortical enhancement on the scout image before proceeding. If cortical enhancement appears weak or absent — suggesting a slow circulation time — consider a brief delayed acquisition rather than proceeding immediately to the nephrographic phase on schedule.
5. Acquire the nephrographic phase at 100 seconds. Allow the full 100-second delay to elapse — this is the single most common point of technologist-driven error (see the scanning pitfalls section). Confirm homogeneous parenchymal enhancement on review. A nephrographic phase acquired even 15–20 seconds early can still show visible corticomedullary differentiation rather than the expected homogeneous blush, which will undermine the HU comparison the radiologist depends on.
6. Reconstruct thin sections in multiple planes. Generate 1.0–1.5 mm axial reconstructions for all three phases plus coronal and sagittal reformats to assist with surgical planning and to detect subtle exophytic lesions at the renal poles that can be missed on axial-only review. Thin sections also reduce partial-volume averaging error when placing ROIs on small lesions, directly improving the reliability of the enhancement calculation.
7. Quality-check enhancement measurements before the patient leaves the table. Place matched regions of interest (ROIs) on any suspicious lesion across all three phases while the patient is still positioned, so that if HU measurements are ambiguous or a phase appears mistimed, a delayed image can be acquired immediately rather than requiring a recall visit. This single quality-control habit, performed consistently, is one of the most effective ways to eliminate repeat-visit renal mass studies in a department.

Common scanning scenarios that complicate the standard workflow

Several recurring clinical scenarios require deviation from the textbook protocol. Patients with a solitary kidney, a transplant kidney in the iliac fossa, or a horseshoe kidney all require adjusted field-of-view planning to ensure complete coverage, since these anatomical variants do not sit in the expected retroperitoneal location. Patients on dialysis with minimal residual renal function present a different challenge: contrast administration decisions must weigh diagnostic necessity against the (generally lower, but not zero) risk profile in this population, typically following institutional protocols and direct radiologist or nephrology consultation rather than a one-size-fits-all rule. Pediatric and young adult patients undergoing renal mass workup for suspected hereditary RCC syndromes (such as von Hippel–Lindau disease) may also require protocol modifications to balance cumulative lifetime radiation exposure against the diagnostic yield of repeated surveillance imaging.

Scanner comparison: 16-slice to 320-slice

Detector configuration materially affects renal mass CT quality, primarily through z-axis coverage and temporal resolution rather than raw image noise.

Dual-energy and photon-counting protocols

Dual-energy CT (DECT) and photon-counting detector (PCD) CT have meaningfully changed renal mass workup by enabling virtual non-contrast (VNC) reconstruction from a single contrast-enhanced acquisition. This can reduce the need for a dedicated non-contrast phase in some workflows, although most departments still acquire a true non-contrast series for medicolegal and diagnostic certainty given that VNC HU values can differ from true non-contrast values, particularly in small lesions.

Deep learning reconstruction (DLR)

Deep learning reconstruction algorithms denoise CT data more aggressively than traditional iterative reconstruction while preserving edge detail, which allows departments to reduce mA on the corticomedullary and nephrographic phases without degrading the conspicuity of small renal lesions. For renal mass protocols specifically, DLR’s ability to suppress noise without smoothing lesion margins is valuable because margin characteristics — smooth versus irregular, well-defined versus infiltrative — directly inform the radiologist’s malignancy assessment.

A practical caveat for departments adopting DLR for renal mass protocols: vendor algorithms are trained on specific noise textures and reconstruction kernels, and switching between DLR strength levels mid-series (for example, applying a different denoising strength to the non-contrast phase than to the contrast-enhanced phases) can subtly alter measured HU values in small lesions. For a protocol whose entire diagnostic logic rests on comparing HU across phases, this is not a trivial concern — departments should lock a single reconstruction kernel and DLR strength setting across all three phases of every renal mass study to preserve measurement consistency.

Protocol harmonization across multiple scanner platforms

Many radiology departments and hospital networks operate a mixed fleet of CT scanners across different sites, vendors, and generations. A frequently underappreciated source of variability in renal mass reporting is not the protocol itself but the fact that the “same” protocol executed on a 16-slice scanner at one site and a 320-slice wide-detector system at another can produce subtly different image noise, spatial resolution, and even effective phase timing due to differences in table speed and acquisition duration. Departments running renal mass protocols across multiple sites should periodically audit phase timing accuracy and image quality metrics across their full scanner fleet, rather than assuming a protocol validated on one system will perform identically on another.

Documentation and quality assurance habits

Beyond the seven acquisition steps themselves, a small number of documentation habits materially improve the diagnostic reliability of every renal mass study. Recording the exact injection start time, flow rate actually achieved, and any deviation from the planned 50/100-second delays in the technologist’s notes gives the radiologist crucial context when HU measurements look unexpected — a borderline enhancement value is interpreted very differently if the technologist’s notes confirm the nephrographic phase ran 12 seconds late versus if no such note exists. Many departments now capture this metadata automatically through injector-scanner integration, but in mixed-fleet environments where older injectors are not networked to the scanner console, a brief manual note remains an effective and low-cost safeguard.

Equally important is a habit of closing the loop with the ordering physician when a technical problem occurs during acquisition — for example, if IV access failed partway through the injection and only a partial bolus was delivered. Rather than simply completing the study with a footnote buried in the technical comments, a direct verbal or electronic flag to the radiologist before final interpretation, and where appropriate to the ordering physician, ensures the limitation is weighed appropriately rather than discovered after a confident-sounding report has already been issued.

How this protocol differs from related genitourinary CT studies

It is useful to situate the dedicated renal mass protocol against its closest relatives in the genitourinary CT family, since technologists and trainees frequently confuse the phase structures of these similarly named studies. A 3-phase adrenal washout protocol shares the non-contrast-plus-multiphase logic but extends the final delayed phase out to 15 minutes specifically to calculate absolute and relative washout percentages for adrenal adenoma characterization — a much longer delay than the 100-second nephrographic phase used here, reflecting the different washout kinetics of adrenal versus renal tissue. A CT urogram, by contrast, adds a delayed excretory phase (often 10 minutes or more, or achieved via a split-bolus technique) specifically to opacify the collecting system and ureters for hematuria or urothelial cancer evaluation — a goal this dedicated renal mass protocol does not pursue, since its corticomedullary and nephrographic phases are timed for parenchymal lesion characterization rather than collecting-system opacification. A technologist who defaults to a urogram-style delayed phase when a renal mass protocol was ordered, or vice versa, will produce a technically competent but clinically mismatched study that fails to answer the question actually being asked.

Contrast media protocol

The dedicated renal mass protocol is fundamentally a contrast-timing exercise, and getting the injection mechanics right is just as important as getting the scan delays right.

Full injection protocol

• Contrast volume: 100 mL of low-osmolar or iso-osmolar iodinated contrast (typically 300–370 mg I/mL concentration, adjusted to patient body habitus).
• Flow rate: 3.5 mL/s, delivered through a power injector via an 18–20 gauge antecubital catheter.
• Saline chaser: 100 mL administered immediately following the contrast bolus to push the trailing contrast column and reduce streak artifact in the brachiocephalic and subclavian veins.
• Delay/trigger: Fixed time delays are used rather than bolus tracking — corticomedullary phase at 50 seconds, nephrographic phase at 100 seconds post-injection start.
• Patient-specific adjustment: In patients with reduced cardiac output or significant cardiomyopathy, consider extending both delays by 10–15 seconds, since slower circulation delays peak cortical and parenchymal enhancement.

Fixed delays are deliberately chosen over bolus tracking in this protocol because the diagnostic goal is reproducible parenchymal phase timing rather than peak arterial opacification — the corticomedullary and nephrographic phases are defined by tissue enhancement physiology, not by a trigger threshold in a single vessel. This is a key conceptual difference from angiographic CT protocols where bolus tracking against an aortic or pulmonary trunk threshold is preferred.

Contrast concentration and viscosity considerations

Higher-iodine-concentration contrast media (350–370 mg I/mL) deliver more iodine per unit volume, which can improve corticomedullary enhancement, but also carry higher viscosity that can challenge smaller-gauge IV catheters at the 3.5 mL/s flow rate this protocol requires. Departments should match contrast concentration and injection rate to the catheter gauge actually in place rather than to an idealized standard, and should have a documented fallback flow rate (typically 2.5–3.0 mL/s) for patients with smaller or more fragile venous access, accepting that corticomedullary phase timing may need slight adjustment in these cases.

Premedication and allergy management

Patients with a documented prior moderate or severe reaction to iodinated contrast should follow institutional premedication protocols (typically a corticosteroid regimen beginning several hours before the study, with or without an antihistamine) per ACR Manual on Contrast Media guidance. For patients with a known severe allergy where premedication is insufficient or contraindicated, alternative non-contrast renal mass evaluation strategies (MRI, or in select cases ultrasound with contrast-enhanced ultrasound) should be discussed with the referring physician before the CT appointment, rather than discovered as a problem on the day of the scan.

Why the saline chaser volume matters

The 100 mL saline chaser specified in this protocol is not an arbitrary round number — it is sized to fully clear the contrast column from the injection tubing, antecubital vein, and central venous system before the corticomedullary acquisition begins, ensuring that no residual undiluted contrast remains to cause beam-hardening streak artifact through the superior vena cava and right atrium on the chest portion of the scan field. A chaser volume that is too small can leave a “tail” of contrast in the line that arrives at the heart as a separate, smaller secondary bolus, subtly altering the expected enhancement curve and occasionally creating a visible second peak in dynamic enhancement curves on more sophisticated reconstruction software. Conversely, an excessively large or fast saline chaser in a patient with fragile venous access raises the risk of triggering extravasation at the injection site, particularly if the chaser is delivered at the same 3.5 mL/s rate as the contrast bolus itself.

Some departments adjust chaser rate slightly downward relative to the contrast injection rate as a routine safety measure in older or more fragile-veined patients, accepting a marginally less sharp bolus front in exchange for reduced extravasation risk. This kind of judgment call — balancing image quality against patient safety on an individual basis — is exactly the sort of protocol nuance that benefits from being documented in departmental policy rather than left to ad hoc decision-making at the injector console.

Extravasation recognition and management

Given the relatively high flow rate (3.5 mL/s) used in this protocol, technologists should monitor the injection site closely during the bolus, particularly in the first 10–15 seconds when extravasation is most likely to occur and most amenable to early intervention. Signs include localized swelling, pain, or a visible change in skin appearance at the injection site. If extravasation is suspected, the injection should be stopped immediately, the affected limb elevated, and the area assessed per institutional extravasation protocol — most small-volume extravasations resolve without sequela, but documentation and patient monitoring remain standard practice. A significant extravasation will also compromise the diagnostic bolus delivered to the kidneys, meaning the resulting corticomedullary and nephrographic phases may show artificially reduced or delayed enhancement that should be flagged to the interpreting radiologist rather than treated as a true physiological finding.

Radiation dose and diagnostic reference levels

Because a dedicated renal mass protocol requires three full acquisitions of the same anatomy, cumulative dose is meaningfully higher than a single-phase abdominal CT, making dose optimization strategies especially important for this exam.

These figures are consistent with published national diagnostic reference levels (DRLs) for multiphase renal CT and should be benchmarked locally against institutional DRL audits, in line with EC Radiation Protection guidance, AAPM practice recommendations, and ICRP Publication 135 on diagnostic reference levels.^[14][15]

It is worth emphasizing that the cumulative effective dose figure for a full three-phase study — roughly 9–13.5 mSv — places this examination meaningfully above a single-phase abdominal CT, and in a similar range to some oncologic staging protocols. This is the direct cost of the diagnostic benefit the protocol provides: three full acquisitions are what make HU-based lesion characterization possible at all. Communicating this trade-off clearly to referring physicians is part of responsible stewardship, since not every incidentally noted renal lesion requires a full dedicated characterization study — many small, simple-appearing cysts identified on a routine single-phase scan can be confidently called benign without ever needing the triple-phase protocol, based on their non-contrast or single-phase appearance alone.

Benchmarking against institutional and national DRLs

Diagnostic reference levels are not dose limits but benchmarking tools — a department whose typical renal mass protocol dose consistently exceeds the relevant national or regional DRL for this examination type should review its acquisition parameters, not simply accept the higher dose as a cost of better image quality. Conversely, a department operating well below the DRL should confirm that diagnostic image quality, particularly the conspicuity of sub-centimeter lesions, has not been compromised in pursuit of dose reduction. Regular DRL audits, ideally reviewed at the protocol level rather than only in aggregate across all CT studies, are the mechanism by which departments catch this kind of unintentional drift in either direction.

Dose considerations in special populations

Younger patients undergoing renal mass workup — whether for a hereditary RCC syndrome, a sporadic incidental finding, or post-transplant surveillance — carry a longer remaining lifespan over which cumulative radiation exposure can theoretically contribute to secondary malignancy risk, even though that absolute risk from any single diagnostic CT remains low. For these patients, departments should be especially diligent about applying the dose reduction strategies above and about avoiding unnecessary repeat triple-phase studies when a single non-contrast or single-phase contrast-enhanced follow-up would adequately answer the clinical question (for example, confirming stability of a previously characterized Bosniak I cyst). Pregnant patients require particular care: iodinated contrast crosses the placenta, and a dedicated renal mass protocol should only proceed in pregnancy when the clinical urgency clearly outweighs the radiation and contrast exposure, with documented multidisciplinary discussion and, where feasible, consideration of MRI or ultrasound alternatives first.

Patients undergoing serial surveillance of a known small renal mass under an active surveillance management strategy represent another population worth specific institutional attention. Because these patients may undergo several renal mass protocol studies over a multi-year surveillance period, departments should track cumulative dose for this patient population specifically and consider whether a less-than-full-triple-phase follow-up protocol (for example, omitting the corticomedullary phase on stable surveillance studies where only nephrographic-phase size and enhancement confirmation is needed) is appropriate per institutional and radiologist judgment.

5 dose reduction strategies

1. Automatic exposure control (AEC) with tube current modulation across all three phases, adjusted to patient cross-sectional area rather than a fixed mA for every body habitus.
2. Iterative or deep learning reconstruction to permit lower mA without sacrificing the conspicuity of small lesions, particularly valuable on the noise-sensitive non-contrast phase.
3. Eliminate redundant phases when prior imaging or clinical context already provides a reliable non-contrast baseline — for example, a recent unenhanced CT performed for renal colic within an appropriate timeframe may sometimes substitute for a repeat non-contrast acquisition, per departmental protocol.
4. Virtual non-contrast reconstruction on dual-energy or photon-counting platforms to potentially reduce the number of true acquisitions required, where validated locally.
5. Tight z-axis collimation limited to the kidneys and immediate surrounding structures rather than a full abdomen-pelvis field of view, when the clinical question is strictly lesion characterization rather than staging.

Top 10 renal mass pathologies

The following ten entities account for the overwhelming majority of renal masses encountered on a dedicated protocol, ranging from entirely benign to highly malignant. They are organized roughly from the most clinically significant solid malignant tumors through to benign cystic and infectious processes, mirroring the diagnostic priority a radiologist applies when reviewing an unfamiliar renal lesion: first rule out an aggressive solid malignancy, then consider the benign solid mimics, then characterize cystic lesions by Bosniak category, and finally consider inflammatory or infectious processes that can transiently mimic a mass.

Why enhancement pattern, not just enhancement magnitude, matters

A recurring theme across these ten entities is that the radiologist is not simply asking “does this lesion enhance?” but “how does this lesion enhance over time, and how does that pattern compare to the surrounding parenchyma?” Clear cell RCC’s hallmark avid-enhancement-then-washout pattern reflects its hypervascular, often disorganized tumor vasculature. Papillary RCC’s more sluggish, progressive enhancement reflects a comparatively hypovascular tumor biology. Chromophobe RCC and oncocytoma can both show a centripetal or spoke-wheel enhancement pattern tied to their characteristic vascular architecture. None of these patterns can be appreciated from a single post-contrast phase — they require at minimum two contrast phases compared against each other and against the non-contrast baseline, which is exactly the structure this protocol provides.

Lesion size also interacts with all of the pathology-specific patterns above in clinically important ways. Small renal masses (generally defined as ≤4 cm) are disproportionately likely to be low-grade or indolent even when histologically RCC, and active surveillance has become an accepted management option for select small, asymptomatic masses in older or comorbid patients, particularly when growth rate on serial imaging remains slow.^[4] This makes accurate, reproducible serial measurement — both of size and of enhancement characteristics — a clinically meaningful task in its own right, not merely an academic exercise in classification.

Multidisciplinary tumor boards increasingly rely on the specific HU values and enhancement percentages documented in the renal mass CT report — not merely a categorical “suspicious for RCC” impression — to guide decisions between partial nephrectomy, radical nephrectomy, thermal ablation, and active surveillance. A report stating that a 2.1 cm exophytic lesion rose from 32 HU to 95 HU on the corticomedullary phase with washout to 68 HU on the nephrographic phase gives the urologic surgeon and tumor board far more actionable information than a report stating only that the lesion “demonstrates enhancement consistent with RCC.” This level of quantitative detail in reporting is not a stylistic preference; it is what allows downstream clinical decision-making to engage directly with the same evidence the radiologist used to reach a conclusion, rather than simply trusting a categorical label.

Staging-relevant findings beyond the primary mass

Characterizing the primary renal lesion is only part of the radiologist’s task on a dedicated renal mass protocol. Because the corticomedullary and nephrographic phases also clearly depict the renal vein, IVC, perinephric fat, Gerota’s fascia, and regional lymph nodes, every renal mass study should be systematically reviewed for staging-relevant findings even when the primary lesion itself appears straightforward. Extension of tumor through the renal capsule into perinephric fat, invasion through Gerota’s fascia into adjacent organs or the abdominal wall musculature, renal vein or IVC tumor thrombus, and enlarged or necrotic regional lymph nodes all upstage a presumed RCC and materially change the surgical approach — from a minimally invasive partial nephrectomy to a more extensive radical resection, potentially with vascular surgery involvement if caval thrombus is present.

This systematic staging review is easy to under-prioritize when a radiologist is focused on the more cognitively demanding task of characterizing the primary lesion’s enhancement pattern, but skipping it represents a meaningful gap in the overall diagnostic value of the study. A structured reporting template that explicitly prompts for renal vein and IVC assessment, perinephric fat stranding or nodularity, and regional lymph node size on every renal mass report helps ensure this step is not silently omitted, particularly on studies read late in a shift or under heavy worklist pressure.

Pitfalls for radiographers

The single most consequential scanning pitfall in this protocol is inadvertently skipping the baseline non-contrast scan. Once contrast has been injected, there is no way to retroactively obtain a true unenhanced HU value for a lesion — the opportunity is permanently lost for that visit, and it becomes impossible to verify baseline HU numbers needed to distinguish a hyperdense benign cyst from a tumor. This typically happens under time pressure, when a technologist assumes a recent prior CT provides an adequate baseline, or when a protocol is selected from a template that omits the non-contrast series by default.

This pitfall is particularly insidious because it often produces a study that still looks complete to a casual review — two well-timed contrast phases are present, the images are diagnostic quality, and nothing about the acquisition itself appears to have gone wrong. The failure only becomes apparent when the radiologist attempts to calculate enhancement and realizes there is no valid reference point. By that stage, the patient has already left the department, and the only remedies are an inferior workaround (comparing against typical published baseline ranges rather than the patient’s actual lesion) or a recall visit. Building institutional awareness that this is a “silent” failure mode — one that does not announce itself through obviously poor image quality — is part of what makes consistent protocol adherence so important, and is a recurring topic worth revisiting during technologist competency reviews rather than addressing only once during initial onboarding.

Pitfalls for radiologists

The defining interpretation pitfall in renal mass CT is the fat-poor angiomyolipoma masquerading as clear cell RCC. A small angiomyolipoma with minimal or no visible macroscopic fat can demonstrate rapid, intense enhancement and subsequent washout that is visually and quantitatively indistinguishable from an aggressive clear cell RCC on CT alone. Because the entire triple-phase protocol is built around enhancement kinetics, this single overlap case is the protocol’s most important diagnostic blind spot.

The clinical stakes of this particular overlap cut in both directions. Misclassifying a fat-poor AML as clear cell RCC can lead to unnecessary partial or radical nephrectomy on a tumor that would never have caused harm. Misclassifying a small clear cell RCC as a fat-poor AML — perhaps because a prior imaging report mentioned a “probable AML” without histologic confirmation — can lead to inappropriate surveillance of an actively malignant lesion. Because CT enhancement kinetics alone cannot always resolve this ambiguity, many academic radiology departments now have an explicit threshold (often based on lesion size, growth, and morphology) below which biopsy is recommended for indeterminate solid enhancing renal masses, rather than relying on imaging alone to make a final call.

Pitfalls for non-radiology physicians

Emergency physicians, internal medicine clinicians, and primary care providers frequently encounter renal mass findings as an incidental result on a scan ordered for an entirely different indication — flank pain, hematuria workup, or trauma evaluation. Because these clinicians are not reading the images themselves and are relying entirely on the written report, the specific wording of that report and how it is acted upon becomes the critical link in the diagnostic chain. The pitfalls below are less about misreading an image and more about misreading — or failing to act fully on — the language of a radiology report.

One further communication pattern worth flagging for non-radiology physicians is the difference between a radiologist recommending “follow-up imaging” versus “urology referral” versus “multidisciplinary discussion” — these are not interchangeable phrases, and each implies a different urgency and a different next step in the care pathway. A busy referring physician scanning a long report for the bottom-line impression can sometimes conflate these recommendations, particularly when a report contains multiple findings of varying significance. Taking the extra moment to confirm the specific recommended action — and who is responsible for arranging it — closes a communication gap that contributes to delayed care more often than any single misread image ever does.

Pitfall comparison summary

These three failure modes are not equally visible within a department’s existing quality assurance structures. Scanning pitfalls are usually caught quickly, because a missing or mistimed phase is often apparent on technical review shortly after acquisition. Interpretation pitfalls like the AML-RCC overlap are harder to catch in real time and often only surface in retrospect, when surgical pathology returns a result that does not match the pre-operative imaging impression. Clinical pitfalls — an indeterminate result quietly becoming a closed loop — are the hardest of all to catch, because by definition they involve a process failure (a missed follow-up appointment, an unscheduled recall) rather than a single identifiable error in a single encounter. Departments serious about improving renal mass outcomes need distinct quality improvement strategies for each of these three failure types, since a solution designed for one (such as a stricter scanning checklist) will do little to address the others.

An implementation checklist for departments

• Lock the non-contrast phase as mandatory in the scanner protocol template, requiring explicit supervisor authorization to bypass it.
• Standardize injector settings for flow rate, volume, and saline chaser across every technologist and shift to minimize bolus-geometry variability between patients.
• Adopt a structured reporting template for renal mass studies that explicitly states the non-contrast, corticomedullary, and nephrographic HU values for every measured lesion, rather than a free-text summary alone.
• Build automatic follow-up tracking into the reporting system for any Bosniak IIF or indeterminate solid lesion, so that recommended imaging intervals generate a trackable task rather than relying on the ordering physician to remember independently.
• Audit dose and phase-timing accuracy on a recurring schedule, especially in departments running this protocol across multiple scanner platforms or sites.
• Establish a clear biopsy referral threshold for indeterminate solid enhancing masses where the AML-versus-RCC distinction cannot be resolved on imaging alone.

AI and automation in renal mass CT

Artificial intelligence is increasingly embedded in renal mass workflows at several distinct points: automated lesion detection on screening or incidental scans, quantitative HU and volumetric tracking across serial studies, and decision-support tools that flag enhancement patterns warranting Bosniak reclassification or urology referral.

Several CE-marked and FDA-cleared software platforms now offer automated renal lesion segmentation and longitudinal HU tracking, reducing the manual ROI-placement variability that contributes to the interreader disagreement long documented in Bosniak classification studies.^[2] Deep-learning detection algorithms trained on large multiphase CT datasets have demonstrated the ability to flag small renal lesions that might otherwise be under-characterized on a busy worklist, functioning as a second-reader safety net rather than a replacement for radiologist judgment.

For departments running high volumes of incidental renal lesion follow-up, automation also helps close the “indeterminate to benign” reporting gap discussed above — structured reporting templates with embedded follow-up scheduling triggers ensure that a recommendation for a 6- or 12-month repeat CT is not lost in a free-text impression line. The evidence base for these tools continues to mature, and current professional society guidance still positions AI as an assistive layer under radiologist oversight rather than an autonomous diagnostic step.

Beyond detection and tracking, automation is also beginning to reach the scanning side of the workflow. Some modern CT platforms now offer automated phase-timing verification, flagging when an acquisition has started meaningfully earlier or later than the protocol-specified delay based on real-time bolus tracking data, even in fixed-delay protocols like this one. This kind of technical safety net directly targets the most common radiographer-side pitfall described earlier — a mistimed nephrographic acquisition — by surfacing the discrepancy to the technologist before the patient leaves the table, rather than leaving it for the radiologist to discover, and possibly question, after the fact.

A practical adoption consideration for hospital administrators is that AI-assisted renal lesion tools generally fall into two distinct categories with different value propositions: detection-support tools, which flag potentially missed lesions on studies originally ordered for other indications, and characterization-support tools, which assist with quantitative tracking of known lesions across serial studies. The former addresses the risk of an incidental renal mass being overlooked entirely on a busy general radiology worklist; the latter addresses the risk of inconsistent manual measurement introducing noise into longitudinal HU and size comparisons. Departments evaluating a purchase should be clear about which problem they are trying to solve, since a tool validated for one use case is not automatically appropriate for the other.

Looking ahead, the most promising near-term direction for AI in this space is probably not autonomous diagnosis but tighter integration between the acquisition and interpretation stages — software that confirms phase timing accuracy at the moment of scanning, automatically populates structured HU values into the report draft, and flags indeterminate enhancement patterns for radiologist double-read, all within a single connected workflow. This kind of integration directly targets the three-tier pitfall framework described throughout this guide by reducing the manual handoff points where scanning errors, measurement inconsistency, and reporting language ambiguity are most likely to creep in.

Frequently asked questions

Why is the non-contrast phase mandatory in a dedicated CT renal mass protocol?
The non-contrast phase establishes the baseline Hounsfield Unit value of a renal lesion. Without it, radiologists cannot calculate true enhancement on later phases, making it impossible to reliably separate a hyperdense benign cyst from a solid enhancing tumor.

What HU enhancement threshold indicates a solid renal mass?
A rise of 15 to 20 Hounsfield Units or more between the non-contrast phase and the corticomedullary or nephrographic phase is generally accepted as indicating a solid, enhancing mass requiring further characterization, though smaller lesions may need a higher threshold to account for pseudo-enhancement.

What is the most common scanning pitfall in renal mass CT?
Inadvertently skipping or rushing the baseline non-contrast acquisition is the most consequential scanning pitfall, since it removes the only reference point needed to confirm true enhancement, and the omission cannot be corrected after contrast has already been injected.

Can MRI replace CT for renal mass characterization?
MRI can be a valuable alternative or adjunct, particularly for patients who cannot receive iodinated contrast or when CT findings remain indeterminate, but CT generally remains the first-line modality due to its speed, availability, and well-established HU-based diagnostic criteria.

How often should an indeterminate (Bosniak IIF) renal cyst be followed up?
Most departments follow Bosniak IIF lesions with repeat imaging at approximately 6 and 12 months and then annually if stable, though exact intervals should follow institutional protocol and the individual patient’s risk factors.

What CT alone cannot tell you

It is important to be explicit about the diagnostic ceiling of even a perfectly executed triple-phase CT. CT cannot reliably distinguish low-grade from high-grade clear cell RCC, cannot always distinguish chromophobe RCC from oncocytoma, and — as emphasized throughout this guide — cannot always distinguish fat-poor angiomyolipoma from clear cell RCC. These limitations are not failures of technique; they reflect genuine overlap in the macroscopic enhancement behavior of biologically distinct tissue types. MRI, with its superior soft-tissue contrast and chemical-shift fat-sensitive sequences, can resolve some but not all of this overlap. Percutaneous biopsy, where appropriate and feasible, remains the only way to achieve histologic certainty for genuinely indeterminate lesions. A well-run renal mass CT service should be confident in what the triple-phase protocol can answer, equally clear about what it cannot, and have an established referral pathway — to MRI, biopsy, or urology — for the subset of cases that fall into that residual uncertainty.

Further reading

1. Multi-Phase Liver CT Protocol: 7 Critical HCC Steps
2. Abdomen Pelvis CT Protocol: 7 Proven Scan Steps
3. Dual-Phase Pancreatic CT Protocol: 7 Critical Steps
4. 5 Male Pelvic CT Protocol Tactics for Radiologists
5. 10 Essential Tips for Female Pelvic CT Protocols

Conclusion

The dedicated CT renal mass protocol succeeds or fails on a single design principle: every diagnostic decision depends on comparing HU values across three precisely timed phases against a true non-contrast baseline. Get the non-contrast acquisition, the 50-second corticomedullary delay, and the 100-second nephrographic delay right, and the protocol reliably separates simple Bosniak I cysts from the ten pathologies covered here — from indolent oncocytomas to aggressive clear cell RCC.

The three-tier pitfall framework matters because each professional group fails differently. Radiographers most often lose diagnostic value by skipping or mistiming a phase. Radiologists are most vulnerable to the fat-poor angiomyolipoma–clear cell RCC overlap that no amount of careful technique can fully resolve on CT alone. Non-radiology physicians most often break the diagnostic loop by treating an “indeterminate” result as a closed, benign finding. Departments that train explicitly against all three failure modes — not just the scanning technique — are the ones that consistently catch renal malignancy at a curable stage.

As detector technology, dual-energy and photon-counting platforms, and AI-assisted detection and tracking tools continue to mature, the core physiological logic of this protocol is unlikely to change: a true non-contrast baseline, a corticomedullary phase capturing peak cortical perfusion, and a nephrographic phase providing a homogeneous detection background will remain the three pillars of confident renal mass characterization. The technologies built around that logic will keep improving the speed, dose efficiency, and reproducibility of the exam — but they cannot substitute for disciplined execution of the fundamentals at every single step, from the first non-contrast slice to the final signed report.

Ultimately, the quality of a renal mass CT program is not determined by any single piece of equipment or any single radiologist’s expertise, but by whether the entire chain — technologist, radiologist, and ordering physician — treats this protocol with the procedural discipline its diagnostic logic demands. A department that builds that discipline into its templates, training, and reporting habits will consistently deliver the clear, actionable answers that patients with an incidental renal lesion are counting on.

References

1. Silverman, S. G., Pedrosa, I., Ellis, J. H., Hindman, N. M., Schieda, N., Smith, A. D., Remer, E. M., Shinagare, A. B., Curci, N. E., Raman, S. S., Wells, S. A., Kaffenberger, S. D., Wang, Z. J., Chandarana, H., & Davenport, M. S. (2019). Bosniak classification of cystic renal masses, version 2019: An update proposal and needs assessment. Radiology, 292(2), 475–488. https://doi.org/10.1148/radiol.2019182646
2. Bhattacharya, A., Lim, R. S., & Schieda, N. (2020). MRI-based Bosniak classification of cystic renal masses, version 2019: Interobserver agreement, impact of readers’ experience, and diagnostic performance. Radiology, 297(1), 134–142. https://doi.org/10.1148/radiol.2020200478
3. Schoots, I. G., Zaccai, K., Hunink, M. G., & Verhagen, P. C. M. S. (2017). Bosniak classification for complex renal cysts reevaluated: A systematic review. Journal of Urology, 198(1), 12–21. https://doi.org/10.1016/j.juro.2016.09.160
4. Campbell, S. C., Uzzo, R. G., Karam, J. A., Chang, S. S., Clark, P. E., & Souter, L. (2021). Renal mass and localized renal cancer: Evaluation, management, and follow-up: AUA guideline, part I. Journal of Urology, 206(2), 199–208. https://doi.org/10.1097/JU.0000000000001911
5. Herts, B. R., Silverman, S. G., Hindman, N. M., Uzzo, R. G., Hartman, R. P., Israel, G. M., Baumgarten, D. A., Berland, L. L., Pandharipande, P. V., & Wells, S. A. (2018). Management of the incidental renal mass on CT: A white paper of the ACR Incidental Findings Committee. Journal of the American College of Radiology, 15(2), 264–273. https://doi.org/10.1016/j.jacr.2017.04.028
6. Davenport, M. S., Perazella, M. A., Yee, J., Dillman, J. R., Fine, D., McDonald, R. J., Rodby, R. A., Wang, C. L., & Weinreb, J. C. (2020). Use of intravenous iodinated contrast media in patients with kidney disease: Consensus statements from the American College of Radiology and the National Kidney Foundation. Radiology, 294(3), 660–668. https://doi.org/10.1148/radiol.2019192094
7. Levin, A., & Stevens, P. E. (2024). Executive summary of the KDIGO 2024 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney International, 105(4), 684–701. https://doi.org/10.1016/j.kint.2023.10.016
8. Krishna, S., Murray, C. A., McInnes, M. D., Chatelain, R., Siddaiah, M., Sathiadoss, P., & Schieda, N. (2017). CT imaging of solid renal masses: Pitfalls and solutions. Clinical Radiology, 72(9), 708–721. https://doi.org/10.1016/j.crad.2017.05.003
9. Sasaguri, K., & Takahashi, N. (2018). CT and MR imaging for solid renal mass characterization. European Journal of Radiology, 99, 40–54. https://doi.org/10.1016/j.ejrad.2017.12.008
10. Kay, F. U., & Pedrosa, I. (2017). Imaging of solid renal masses. Radiologic Clinics of North America, 55(2), 243–258. https://doi.org/10.1016/j.rcl.2016.10.003
11. Schieda, N., Davenport, M. S., Pedrosa, I., Shinagare, A., Chandarana, H., Curci, N., Doshi, A., Hindman, N., Remer, E., & Silverman, S. G. (2020). Renal cell carcinoma versus oncocytoma: How do MR and CT differentiate? Abdominal Radiology, 45(1), 35–49. https://doi.org/10.1007/s00261-019-02244-9
12. Silverman, S. G., Israel, G. M., Trinh, Q. D., & Bredael, M. (2015). Incompletely characterized incidental renal masses: Emerging data support conservative management. Radiology, 275(1), 28–42. https://doi.org/10.1148/radiol.14122631
13. Schieda, N., Lim, R. S., Krishna, S., McInnes, M. D., Flood, T. A., & Thornhill, R. E. (2021). Diagnostic accuracy of CT and MRI for detection of clear cell renal cell carcinoma negative imaging features. American Journal of Roentgenology, 217(4), 859–870. https://doi.org/10.2214/AJR.20.23202
14. American Association of Physicists in Medicine. (2023). AAPM medical physics practice guideline: CT protocol management and review. https://doi.org/10.1002/acm2.13905
15. International Commission on Radiological Protection. (2017). ICRP Publication 135: Diagnostic reference levels in medical imaging. Annals of the ICRP, 46(1), 1–144. https://doi.org/10.1177/0146645317717209
16. European Commission, Directorate-General for Energy. (2018). Radiation Protection N° 185: European guidelines on diagnostic reference levels for medical imaging. https://op.europa.eu/en/publication-detail/-/publication/d8794ac1-d8a7-11e8-9424-01aa75ed71a1
17. Hindman, N. M. (2017). Approach to very small (<1.5 cm) cystic renal lesions: Ignore, observe, or treat? American Journal of Roentgenology, 209(6), 1182–1187. https://doi.org/10.2214/AJR.17.18667
18. Bata, P., Gyebnar, J., Tarnoki, D. L., Tarnoki, A. D., Kekesi, D., Szendroi-Kiss, D., & Lazar, A. (2015). Angiomyolipoma: Correlation of imaging features with histopathological findings. Polish Journal of Radiology, 80, 463–470. https://doi.org/10.12659/PJR.894400
19. Hodgdon, T., McInnes, M. D., Schieda, N., Flood, T. A., Lamb, L., & Thornhill, R. E. (2015). Can quantitative CT texture analysis be used to differentiate fat-poor renal angiomyolipoma from renal cell carcinoma on unenhanced CT images? Radiology, 276(3), 787–796. https://doi.org/10.1148/radiol.2015142215
20. Patel, B. N., Alexander, L. F., Allen, B. C., Galgiani, A., Brady, C., Farjat, A. E., & Marin, D. (2019). Dual-energy CT workflow for evaluation of incidental renal lesions: Added value of virtual unenhanced images. Abdominal Radiology, 44(8), 2839–2849. https://doi.org/10.1007/s00261-019-02043-4
21. U.S. Food and Drug Administration. (2024). Artificial intelligence-enabled medical devices list. https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-enabled-medical-devices
22. Bhandari, A., Ibrahim, M., Sharma, C., Liong, R., Gustafson, S., & Prior, M. (2021). Artificial intelligence and machine learning in renal cell carcinoma: Methods, applications and challenges. European Urology Focus, 7(4), 706–711. https://doi.org/10.1016/j.euf.2020.08.005
23. Davenport, M. S., Hu, E. M., Smith, A. D., Sahani, D. V., Kaffenberger, S. D., Doshi, A. M., Khalatbari, S., Wood, C. G., & Silverman, S. G. (2017). Reporting standards for the imaging-based diagnosis of renal masses on CT and MRI: A national survey of academic abdominal radiologists and urologists. Abdominal Radiology, 42(4), 1229–1240. https://doi.org/10.1007/s00261-016-1004-5
24. Moch, H., Amin, M. B., Berney, D. M., Compérat, E. M., Gill, A. J., Hartmann, A., Menon, S., Raspollini, M. R., Rubin, M. A., Srigley, J. R., Tan, P. H., Tickoo, S. K., Tsuzuki, T., Turajlic, S., Cree, I., & Netto, G. J. (2022). The 2022 World Health Organization classification of tumours of the urinary system and male genital organs—part A: Renal, penile, and testicular tumours. European Urology, 82(5), 458–468. https://doi.org/10.1016/j.eururo.2022.06.016
25. National Comprehensive Cancer Network. (2025). NCCN clinical practice guidelines in oncology: Kidney cancer (Version 2.2025). https://www.nccn.org/guidelines/category_1
26. American Cancer Society. (2025). Key statistics about kidney cancer. https://www.cancer.org/cancer/types/kidney-cancer/about/key-statistics.html