Prostate mpMRI & PI-RADS
- Multiparametric MRI (mpMRI) of the prostate combines three ingredients: anatomy (T2), water-movement (diffusion), and blood-flow (dynamic contrast). Each answers a different question.
- The peripheral zone is the prostate's "outer crust" where most clinically significant cancers live, and DWI/ADC is the dominant sequence there. In the transition zone, T2 takes over.
- PI-RADS v2.1 turns the whole study into one category from 1 (very low) to 5 (very high). The 4-versus-5 line is size: 1.5 cm in greatest dimension, or definite extraprostatic extension.
- The scanner has to earn the score: a high b-value of at least 1400 s/mm² and DCE temporal resolution of 15 s or better are the technical minimums that make the rules valid.
- Pooled across a large meta-analysis, PI-RADS v2 was 89% sensitive and 73% specific for prostate cancer. It is a risk dial, not a yes/no diagnosis.
Imagine you're trying to spot a rotten patch in an apple without cutting it open. One light shows you its shape, another shows where the flesh has gone dense and waterlogged, and a third shows where extra little blood vessels have sprouted to feed the rot. Stack those three views and you can point straight at the bad spot. That's exactly what prostate mpMRI does — three complementary looks at one walnut-sized gland, fused into one confident answer.
The three sequences and what each one actually tells you
The "multiparametric" part just means we don't trust any single picture. We deliberately gather three, because each has a blind spot the others cover.
First, T2-weighted imaging is the anatomy map — the high-resolution architectural drawing. It shows the zones, the capsule, and the seminal vesicles. (If "T2-weighted" still feels like radiology Klingon, the foundations live over at MRI basics: T1, T2 and weighting.) On T2, normal peripheral zone is bright and healthy, like a clean sponge. Cancer shows up as a smudgy, ill-defined dark blot, as if someone pressed a thumb of charcoal into that sponge.
Second, diffusion-weighted imaging (DWI), with its sidekick the ADC map, measures how freely water molecules wander. Tumor cells pack together tightly, like commuters cramming a subway car, so water can't move — it "restricts." That shows up as bright on the high b-value DWI and dark on the ADC map. This is the single most important sequence for the peripheral zone. (How DWI and ADC actually work.)
Third, dynamic contrast-enhanced (DCE) imaging watches a bolus of gadolinium wash through. Tumors grow greedy, leaky vessels, so they tend to light up early. (Gadolinium has its own quirks and cautions — see gadolinium agents.) DCE is the tie-breaker, not the headliner.
Findings by modality
MRI
This is the whole show, so it gets the detail. Where a lesion sits decides which sequence runs the scoring.
| Zone | Where it is | Dominant sequence | Cancer's usual disguise |
|---|---|---|---|
| Peripheral zone (PZ) | Outer back portion | DWI/ADC | Round or wedge-shaped focus, bright on high b-value DWI, dark on ADC, with a matching low-T2 smudge |
| Transition zone (TZ) | Central, around the urethra | T2-weighted | "Erased charcoal": a homogeneous, lens-shaped, poorly marginated dark lesion that blurs into its neighbors, unlike the crisply encapsulated round nodules of BPH |
On T2, look for the hypointense focus and then immediately look at its edges: is it round and encapsulated (benign nodule), or smeared and lenticular (cancer)? Then look at the capsule for bulge, irregularity, or frank tumor signal in the periprostatic fat, because that changes the category and the stage.
On DWI/ADC, the tell is concordance: the same spot bright on the high b-value image and dark on the ADC map. Bright on DWI alone can be T2 shine-through; dark on ADC alone can be a benign stromal nodule. You want both. For the rules to apply, the high b-value image must be at least 1400 s/mm², acquired or calculated.
On DCE, the only thing you need is focal early enhancement that matches a suspicious T2 or DWI finding. Diffuse enhancement is more in keeping with prostatitis than with a focal tumor. DCE can only be scored if the temporal resolution is 15 s or better.
The report also states gland volume, using the ellipsoid formula on T2: maximum anteroposterior diameter × longitudinal diameter (measured on the mid-sagittal T2 image) × transverse diameter (measured on the axial T2 image) × 0.52.
Ultrasound
Transrectal ultrasound (TRUS) is the biopsy-guidance and volume tool, not a detection tool. Classic cancer is a hypoechoic peripheral-zone focus, but plenty of cancers are isoechoic and plenty of hypoechoic foci are benign. Its real modern job is fusing with the MRI so the needle goes where the MRI pointed.
CT
CT cannot see cancer inside the gland. It shows an enlarged prostate, calcification, and, in staging, bulky nodes or sclerotic bone metastases. A "normal prostate on CT" sentence tells the urologist nothing about cancer.
Nuclear medicine
Bone scan and PSMA PET hunt for spread, not the primary. When the question shifts from "is there cancer here?" to "where has it gone?", the work moves to PSMA PET, and to the staging page linked below.
The numbers
| What | Threshold / value | Why it matters |
|---|---|---|
| PI-RADS v2.1 categories | 1 very low; 2 low; 3 intermediate (equivocal); 4 high; 5 very high likelihood of clinically significant cancer | The single number the urologist acts on |
| Dominant sequence | PZ: DWI/ADC. TZ: T2-weighted | Decides which score becomes the category |
| Score 4 versus 5 | Score 4: <1.5 cm in greatest dimension. Score 5: ≥1.5 cm, or definite extraprostatic extension or invasive behavior | Size and invasion, not "how dark," separate high from very high |
| PZ upgrade | DWI score 3 becomes category 4 if DCE shows focal early enhancement; otherwise stays 3 | The one job DCE has |
| TZ upgrades | T2 score 3 becomes category 4 if DWI is 5, otherwise stays 3; T2 score 2 becomes category 3 if DWI is ≥4 | DWI can rescue an equivocal or bland-looking TZ lesion |
| High b-value DWI | ≥1400 s/mm², acquired or calculated | Below this the DWI score is not trustworthy |
| DCE temporal resolution | ≤15 s | Slower acquisitions miss early enhancement |
| Gland volume | AP × longitudinal (mid-sagittal T2W) × transverse (axial T2W) × 0.52 | Feeds PSA density and treatment planning |
Memorize the pairing: diffusion for the PZ, T2 for the TZ. Then memorize the one size that matters: 1.5 cm is the line between PI-RADS 4 and 5, and definite extraprostatic extension makes it a 5 regardless of size.
How good is the test
In a systematic review and meta-analysis of 21 studies and 3857 patients, PI-RADS version 2 (the immediate predecessor of v2.1, with the same architecture) had a pooled sensitivity of 89% and a pooled specificity of 73% for the detection of prostate cancer. Read that as: a good MRI misses few cancers, but it also flags a fair share of men who turn out not to have one, because the specificity is only 73%. That is exactly why PI-RADS 3 exists and why the score is a probability, not a verdict. There is no robust pooled accuracy figure for v2.1 specifically; the v2 figures are the ones to quote.
PI-RADS v2.1 at a glance
Abbreviated. Version: PI-RADS v2.1 (2019 update of v2). The full staging logic, including what happens once tumor leaves the gland, lives on prostate cancer staging with mpMRI.
| Category | Likelihood of clinically significant cancer | How you get there |
|---|---|---|
| 1 | Very low (highly unlikely) | Dominant sequence score 1 |
| 2 | Low (unlikely) | Dominant sequence score 2 (a TZ T2 score 2 becomes 3 only if DWI is ≥4) |
| 3 | Intermediate (equivocal) | Dominant sequence score 3 without an upgrade |
| 4 | High (likely) | Dominant sequence score 4 (lesion <1.5 cm in greatest dimension), or a PZ DWI 3 upgraded by focal early DCE enhancement, or a TZ T2 3 upgraded by DWI 5 |
| 5 | Very high (highly likely) | Dominant sequence score 5: ≥1.5 cm in greatest dimension, or definite extraprostatic extension or invasive behavior |
The "3" deserves a special mention: it's the radiology equivalent of a shrug, the honest "I'm not sure," and it's exactly where DCE enhancement or careful follow-up can nudge things one way or the other.
What else looks like cancer
| Mimic | Looks similar because | Tell them apart by |
|---|---|---|
| Post-biopsy hemorrhage | Blood restricts diffusion and darkens T2 | Bright on T1 (blood), often diffuse or streaky; ask when the biopsy was |
| Prostatitis | Low T2 and restricted diffusion in the PZ | Diffuse, band-like or wedge-shaped rather than a discrete focus; diffuse rather than focal early enhancement |
| Benign stromal BPH nodule | Dark on T2 and can be dark on ADC | Round, sharply encapsulated, sitting in the TZ; cancer is lenticular and erases the margins |
| Central zone | Symmetric low T2 and low ADC at the base | Symmetric, paired, tapering toward the verumontanum; a cancer is asymmetric |
| Anterior fibromuscular stroma | Dark on T2 and ADC by nature | Midline anterior, no focal DWI brightness, no early enhancement |
| Extruded BPH into the PZ | A dark "lesion" where cancers live | Continuous with the TZ and encapsulated; benign nodules do not smear |
Recent biopsy can splatter hemorrhage through the gland, and blood restricts diffusion and distorts T2 just like tumor can — a beautiful cancer mimic. This is a big reason MRI is often done before biopsy, or after a waiting interval. When you see blotchy signal, check the T1 images and ask the timeline before you cry wolf.
Reporting
The urologist needs, in this order:
- Gland volume by the ellipsoid formula (three diameters × 0.52) so a PSA density can be calculated.
- Each lesion with its zone, its location on the sector map, its greatest dimension (say which plane), and its T2, DWI, and DCE scores.
- The PI-RADS v2.1 category for each lesion, and which one is the index lesion.
- Extraprostatic extension, seminal vesicle invasion, and neurovascular bundle involvement: present, absent, or equivocal.
- Nodes and bones in the field of view.
- Technical adequacy: high b-value used, DCE temporal resolution, and any artifact (hip prostheses, rectal gas, motion) that limits the study.
Scoring is reproducible but not magic. Readers disagree most on the 3s, and the system explicitly targets significant disease, so a quiet indolent cancer can slip under the radar. mpMRI's superpower isn't catching everything — it's confidently steering the needle toward the lesions that actually matter and sparing some men a biopsy they didn't need.
If you remember one thing: three pictures, one number, and the number means how likely rather than yes or no.
References
- Turkbey B, Rosenkrantz AB, Haider MA, et al. Prostate Imaging Reporting and Data System Version 2.1: 2019 Update of Prostate Imaging Reporting and Data System Version 2. Eur Urol 2019;76(3):340–351. Used for: the PI-RADS v2.1 categories 1–5 and their likelihood wording, the dominant-sequence rule (DWI/ADC for the peripheral zone, T2 for the transition zone), the upgrade rules (PZ DWI 3 to 4 with focal early DCE enhancement; TZ T2 3 to 4 with DWI 5; TZ T2 2 to 3 with DWI ≥4), the 1.5 cm size threshold between scores 4 and 5 and the extraprostatic-extension criterion, the ≥1400 s/mm² high b-value and ≤15 s DCE temporal-resolution minimums, and the ellipsoid volume formula (× 0.52) in "Findings by modality," "The numbers," "PI-RADS v2.1 at a glance," and "Reporting."
- Woo S, Suh CH, Kim SY, Cho JY, Kim SH. Diagnostic Performance of Prostate Imaging Reporting and Data System Version 2 for Detection of Prostate Cancer: A Systematic Review and Diagnostic Meta-analysis. Eur Urol 2017;72(2):177–188. Used for: the pooled sensitivity of 89% and pooled specificity of 73% for PI-RADS v2, from 21 studies and 3857 patients, in the Key Points and "How good is the test."
Sign in to take the quiz and track your progress