Prostate MRI — zone-aware scoring, volume, PSA density and a structured report
Anterior at the top; the patient’s right is on the left of each section. Schematic — not to scale.
PI-RADS v2.1 assigns a prostate MRI lesion a category from 1 to 5, describing how likely it is that clinically significant cancer is present. The category is not an average of the sequences. One sequence dominates, and which one depends on where the lesion sits.
In the peripheral zone, diffusion-weighted imaging drives the assessment. In the transition zone, T2-weighted imaging does. Lesions in the central zone and the anterior fibromuscular stroma have no table of their own — they are assessed with the criteria of the zone they abut, and the report should say so.
Contrast has exactly one job in v2.1: a positive study raises a peripheral zone lesion scoring DWI 3 to an overall PI-RADS 4. It cannot upgrade anything else, and it plays no part at all in the transition zone. In the transition zone the equivalent tie-break is a DWI score of 5 against a T2 score of 3.
The descriptors for 4 and 5 are the same words. What separates them is measurement: a lesion of 15 mm or more scores 5, as does one of any size showing definite extraprostatic extension. Measure the greatest dimension on the axial image where the lesion is most conspicuous, usually the ADC map.
Volume is estimated as AP × transverse × craniocaudal × 0.52, the ellipsoid approximation. PSA density is the serum PSA divided by that volume. It matters most at PI-RADS 3, where the finding is equivocal and density is one of the things commonly used to decide between biopsy and surveillance; 0.15 ng/mL/cc is the threshold in widest use.
Every PI-RADS v2.1 assessment reduces to one of these two tables. The calculator above applies them, but they are short enough to be worth knowing by heart.
| DWI | DCE | PI-RADS |
|---|---|---|
| 1 | any | 1 |
| 2 | any | 2 |
| 3 | − | 3 |
| 3 | + | 4 |
| 4 | any | 4 |
| 5 | any | 5 |
T2-weighted imaging does not appear in this table at all. In the peripheral zone it contributes nothing to the category unless DWI is non-diagnostic, in which case it carries the assessment and the report should say why.
| T2W | DWI | PI-RADS |
|---|---|---|
| 1 | any | 1 |
| 2 | ≤ 3 | 2 |
| 2 | 4 or 5 | 3 |
| 3 | ≤ 4 | 3 |
| 3 | 5 | 4 |
| 4 | any | 4 |
| 5 | any | 5 |
The two upgrades in this table have different thresholds, which is the detail most often got wrong. A T2 score of 2 — an atypical nodule — rises to an overall 3 on a DWI of 4 or 5. A T2 score of 3 rises to an overall 4 only on a DWI of 5. The first of those was introduced in v2.1 and did not exist in v2.
Version 2.1 kept the structure of v2 and altered the details. The transition zone gained the atypical-nodule upgrade above. The T2 criteria for scores 1 and 2 in the transition zone were rewritten around encapsulation rather than appearance alone. DWI acquisition guidance was tightened, and the definition of a positive DCE study was clarified. Reporting of the central zone and the anterior fibromuscular stroma was addressed explicitly for the first time. The practical effect for most readers is a modest change in transition zone categorisation and a clearer account of what a technically adequate study looks like.
A PI-RADS category is a statement about the probability of clinically significant cancer on imaging. It is not a diagnosis, it is not a grade, and it does not incorporate PSA, age, family history or prior biopsy. Interreader agreement is moderate at best and is consistently worse in the transition zone than in the peripheral zone. A category 3 in particular carries a wide range of reported detection rates between centres, which is precisely why it is the category where other information is brought to bear.
Nor does the system claim to be a substitute for reading experience. There is evidence that a structured calculator helps: a 2021 study in European Journal of Radiology Open found that non-specialised radiologists using a browser-based PI-RADS score calculator read prostate MRI faster than with the published document alone, without any loss of interreader agreement or lesion-based accuracy. What a calculator removes is arithmetic and lookup error, not the need to recognise what is on the image.
Wang X, et al. Value of an online PI-RADS v2.1 score calculator for assessment of prostate MRI. Eur J Radiol Open. 2021;8:100332. · Turkbey B, et al. Prostate Imaging Reporting and Data System Version 2.1. Eur Urol. 2019;76(3):340–351.
PI-RADS asks for lesions to be localised on a sector map: 38 prostate sectors across base, mid-gland and apex, plus the two seminal vesicles and the external urethral sphincter, 41 in all. The map matters because a targeted biopsy has to find the lesion again. Naming the sector is more use to the urologist than a description of where it sits, and it makes the report comparable with the next study. The calculator above lets you either click sectors by name or draw the lesion directly onto the map, and puts whichever you choose into the report.
PI-RADS v2.1 also asks that no more than four lesions be reported, and that the highest-scoring one be designated the index lesion. Where two share the same category, extraprostatic extension and then size are the usual tie-breaks.