RADS
RADS Systems Explained: A Guide to Structured Radiology Reporting
An overview of the six ACR RADS systems — TI-RADS, Lung-RADS, BI-RADS, LI-RADS, PI-RADS, and O-RADS — with a free structured report calculator for each.
What "RADS" Actually Means
Open any modern radiology report for a thyroid nodule, a screening chest CT, or a prostate MRI and you will likely encounter a label like "TI-RADS 4" or "PI-RADS 3." These come from a family of classification systems known collectively as RADS — Reporting and Data Systems — developed largely under the American College of Radiology.
The idea behind every RADS is the same. Imaging generates findings that vary in how worrying they are, and radiologists historically described those findings in free-form prose. A narrative report describing a "somewhat irregular, mildly hypoechoic lesion" leaves the referring clinician guessing about what to do next. A RADS replaces that ambiguity with a defined vocabulary, a small set of ordered risk categories, and an explicit management recommendation attached to each category. The result is a report that communicates risk consistently, reduces unnecessary procedures, and can be audited later.
Each organ system has its own RADS because the relevant imaging features, reference standards, and management decisions differ. What they share is the underlying logic: describe the finding with standardized descriptors, map those descriptors to a numbered category, and let the category drive the recommendation. This guide walks through the six systems most widely used in practice, each of which now has a free calculator in the Aperivue RADS suite.
The Six Systems at a Glance
| System | Organ and use | Imaging | Current version | Category scale |
|---|---|---|---|---|
| TI-RADS | Thyroid nodules | Ultrasound | ACR 2017, K-TIRADS 2021, EU-TIRADS 2017 | TR1–TR5 (1–5) |
| Lung-RADS | Lung cancer screening | Low-dose CT | v2022 | 0, 1, 2, 3, 4A, 4B, 4X |
| BI-RADS | Breast | Mammography, US, MRI | ACR Atlas 5th ed. (2013) | 0–6 (incl. 4A/4B/4C) |
| LI-RADS | Liver (HCC high-risk) | CT, MRI | v2018 | LR-1 to LR-5, LR-M, LR-TIV |
| PI-RADS | Prostate | Multiparametric MRI | v2.1 (2019) | 1–5 |
| O-RADS | Ovarian-adnexal | US, MRI | US v2022, MRI 2022 | 0–5 |
TI-RADS — Thyroid Nodules
TI-RADS (Thyroid Imaging Reporting and Data System) stratifies the malignancy risk of thyroid nodules seen on ultrasound. It is unusual among RADS systems in that three major versions coexist: ACR TI-RADS (2017), which assigns numerical points across composition, echogenicity, shape, margin, and echogenic foci to reach levels TR1 through TR5; the 2021 Korean K-TIRADS, which classifies nodules by overall pattern; and the European EU-TIRADS, which uses a simplified feature set. All three end in a category that, combined with nodule size, sets the threshold for fine-needle aspiration.
Thyroid nodules are extraordinarily common and overwhelmingly benign — only roughly 5% to 15% prove malignant — so the entire purpose of TI-RADS is to decide which nodules justify biopsy and which can be watched. Because the three systems diverge in their thresholds, a calculator that supports all three is useful for direct comparison.
Try the free TI-RADS calculator, and for a deeper walkthrough of the three systems see our step-by-step TI-RADS guide.
Lung-RADS — Lung Cancer Screening
Lung-RADS standardizes the interpretation of low-dose CT performed for lung cancer screening. The current version, v2022, sorts each scan into categories 0, 1, 2, 3, 4A, 4B, and 4X. Category 0 means the study is incomplete or a prior comparison is needed; categories 1 and 2 carry a malignancy likelihood under roughly 1% and return the patient to annual screening; category 3 (about 1% to 2%) shortens the interval to a 6-month LDCT; category 4A (roughly 5% to 15%) prompts a 3-month LDCT or PET/CT; and categories 4B and 4X (above 15%) move toward tissue sampling and definitive workup.
The branching logic depends on nodule type. Solid, part-solid, and ground-glass nodules each have their own size thresholds, and for part-solid nodules it is the solid-component size — not the total size — that drives the category. Whether a nodule is new, stable, or growing shifts the threshold further, and an "S" modifier flags clinically significant findings unrelated to the screening nodule itself.
Try the free Lung-RADS calculator, which mirrors this branching logic and produces a screening-ready report, and for a deeper walkthrough see our Lung-RADS guide.
BI-RADS — Breast Imaging
BI-RADS (Breast Imaging Reporting and Data System), now in its fifth edition, is the oldest and most established RADS, applied across mammography, ultrasound, and breast MRI with a single shared assessment scale. The categories run from 0 (incomplete — additional imaging or prior studies needed) through 1 and 2 (essentially 0% malignancy, routine screening), 3 (greater than 0% but no more than 2%, short-interval follow-up), and the 4 subcategories — 4A (more than 2% to 10%), 4B (more than 10% to 50%), and 4C (more than 50% to under 95%) — all of which warrant tissue sampling. Category 5 carries a 95% or higher malignancy risk, and category 6 denotes biopsy-proven cancer.
Unlike the point-based ACR TI-RADS, BI-RADS is fundamentally a lexicon paired with an assessment framework: the radiologist selects standardized descriptors and then assigns an integrated category. Its precise probability bands are what make breast imaging follow-up so reproducible across institutions.
Try the free BI-RADS calculator, and for the full lexicon walkthrough see our BI-RADS complete guide.
LI-RADS — Liver Observations in High-Risk Patients
LI-RADS (Liver Imaging Reporting and Data System) applies specifically to patients at high risk for hepatocellular carcinoma — those with cirrhosis, chronic hepatitis B, or a current or prior HCC — and should not be used in the general population. The CT/MRI v2018 algorithm assigns categories from LR-1 (definitely benign) and LR-2 (probably benign) through LR-3 (intermediate probability of malignancy), LR-4 (probably HCC), and LR-5 (definitely HCC). Two special categories sit outside the numeric ladder: LR-M, for observations that are probably or definitely malignant but not specific for HCC, and LR-TIV, for tumor in vein.
The diagnostic algorithm combines major features — nonrim arterial phase hyperenhancement, nonperipheral washout, an enhancing capsule, and threshold growth — with observation size. Ancillary features can nudge a category up or down by one step, but they cannot raise an observation to LR-5. When image quality prevents categorization, the observation is labeled LR-NC (non-categorizable).
Try the free LI-RADS calculator, which encodes the v2018 precedence rules and major-feature table, and for a deeper walkthrough see our LI-RADS guide.
PI-RADS — Prostate Multiparametric MRI
PI-RADS (Prostate Imaging Reporting and Data System), version 2.1, scores the likelihood of clinically significant prostate cancer on multiparametric MRI, from PI-RADS 1 (clinically significant cancer highly unlikely) to PI-RADS 5 (highly likely). Its defining feature is that it is zone-aware: in the peripheral zone, diffusion-weighted imaging is the dominant sequence, while in the transition zone, T2-weighted imaging dominates.
Dynamic contrast enhancement plays a narrow but specific role — it upgrades a peripheral-zone DWI score of 3 to PI-RADS 4 only when positive; when DCE is negative or unavailable, that lesion stays PI-RADS 3. In the transition zone, a T2W score of 3 with a DWI of 5 upgrades to PI-RADS 4. PI-RADS deliberately does not prescribe management; biopsy decisions are made clinically, informed by PSA density and local protocol.
Try the free PI-RADS calculator, which follows the v2.1 zone-aware decision rules including the DCE upgrade logic, and for a deeper walkthrough see our PI-RADS guide.
O-RADS — Ovarian and Adnexal Lesions
O-RADS (Ovarian-Adnexal Reporting and Data System) is the newest of the six and, like TI-RADS, comes in two flavors that share a 0-to-5 scale but rest on different evidence. O-RADS US v2022 scores lesions from their sonographic lexicon — lesion type, solid components, papillary projections, and a color score for vascularity — with malignancy risk rising from essentially 0% at category 1, under 1% at category 2, 1% to under 10% at category 3, 10% to under 50% at category 4, and 50% or higher at category 5. O-RADS MRI 2022 instead centers on enhancing solid tissue and its dynamic contrast-enhanced time-intensity curve, with approximate positive predictive values of about 0.5% at score 2, 5% at score 3, 50% at score 4, and 90% at score 5.
Because the two systems answer different clinical questions — ultrasound for initial characterization, MRI for problem-solving indeterminate lesions — a calculator that handles both is particularly useful.
Try the free O-RADS calculator, which supports both the US v2022 and MRI 2022 systems, and for a deeper walkthrough see our O-RADS guide.
What the Six Systems Share
Despite covering different organs, all six RADS systems are built on the same three-part scaffold: a standardized lexicon of descriptors, an ordered set of risk categories, and a management recommendation tied to each category. Three principles follow from this design.
First, a RADS category is a communication tool, not a diagnosis. "PI-RADS 4" tells the urologist how suspicious the lesion is and what step is reasonable next; it does not establish cancer. Second, consistency matters more than the choice of system. Studies repeatedly show that the variability introduced by applying a system inconsistently exceeds the performance differences between competing systems. Third, the category drives but does not dictate management. Size thresholds, follow-up intervals, and biopsy recommendations are built on population-level risk-benefit analysis and must still be weighed against the individual patient's history and preferences.
These shared principles are also why a structured report generator helps. Applying the precise criteria of any one system during a busy reading list is error-prone; a tool that encodes the rules reduces the bookkeeping burden and standardizes the documentation of a judgment the radiologist has already made.
The Aperivue RADS Suite
Aperivue RADS provides a free calculator and structured report generator for each of the six systems described here:
- TI-RADS — thyroid (ACR, K-TIRADS, EU-TIRADS)
- Lung-RADS — lung cancer screening (v2022)
- BI-RADS — breast (ACR Atlas 5th edition)
- LI-RADS — liver, CT/MRI (v2018)
- PI-RADS — prostate mpMRI (v2.1)
- O-RADS — ovarian-adnexal (US v2022 and MRI 2022)
Every calculator runs directly in the browser, requires no login, and produces a report you can copy into your reporting workflow. Each page also shows the assessment criteria — categories, risk, and key decision rules — transcribed directly from the same scoring logic the calculator uses, so what you read matches what the tool computes.
Frequently Asked Questions
How many RADS systems are there? More than six. This guide covers the six used most often in routine practice, but others exist for specific contexts — for example C-RADS for CT colonography, NI-RADS for head and neck post-treatment surveillance, and CAD-RADS for coronary CT angiography.
Is a higher RADS number always worse? Within a single system, yes — the categories are ordered by increasing risk. But the numbers are not comparable across systems. An O-RADS 4 and a PI-RADS 4 describe different organs, different risk magnitudes, and different next steps.
Do RADS categories replace biopsy? No. A RADS category stratifies risk and indicates whether, and how soon, biopsy or follow-up is appropriate. The histological diagnosis still requires tissue or, in some settings such as LR-5 in LI-RADS, a noninvasive diagnostic pathway defined by the system itself.
Are the Aperivue calculators free? Yes. All six run directly in the browser, require no login, and generate a report you can copy into your workflow.
Disclaimer
The information provided in this article and through the Aperivue RADS calculators is intended for educational purposes and as a reference for healthcare professionals. It does not constitute medical advice, diagnosis, or treatment recommendations. RADS categories are risk-stratification tools; clinical decisions should be made by qualified physicians based on the complete clinical context. These calculators are support tools — not a replacement for professional medical judgment, and not regulatory-cleared medical devices.