prostateview
Teaching concept only — not for diagnosis, PI‑RADS scoring, biopsy planning, or intra-operative navigation. How to use it safely.
Research & education prototype · UroRef

A GPU pipeline for a mobile 3D prostate-zone teaching library from public MRI

Nity G, Urology ST4 · built on Prostate158 (Adams et al. 2022, CC BY 4.0)

AIiH 2026 RISE-UK 2026 NHS CEP
Not a medical device. Not for diagnosis, PI-RADS, staging or biopsy targeting. Not clinically validated. Educational orientation only.

Problem

Trainees learn prostate zonal anatomy and lesion localisation from 2D MRI, which is hard to internalise spatially. Interactive 3D helps, but bespoke models are slow to make and too heavy for a phone at the bedside.

Approach

An automated, GPU-accelerated pipeline turns public Prostate158 MRI into a library of 100 mobile-optimised 3D zonal models (gland, peripheral and transition zones, lesion), each a single Draco-compressed GLB that loads on a mid-range phone and in AR.

ProstateView pipeline from public MRI to mobile 3D
Pipeline: public T2W → expert/MONAI zonal labels → mesh → decimate + smooth → GLB + Draco → mobile + AR.

By the numbers

100
mobile 3D models
~60,694
mean triangles / model
~0.12 MB
mean GLB size (Draco)
11.5 MB
whole 100-model set
1.21s
GPU validation / case
10,246 MB
peak VRAM (of 16 GB)

Technical validation

Automated MONAI Prostate158 segmentation vs the dataset's expert ground-truth labels, on the exact 100 cases shipped (CUDA, mixed precision). Model-vs-expert agreement — not a clinical validation.

StructureDice (mean ± SD)
Whole gland0.91 ± 0.03
Transition / central zone0.87 ± 0.05
Peripheral zone0.77 ± 0.07

Device: GPU (CUDA, mixed precision) · n = 100. PZ is hardest, as expected.

Honest QA gate

The build runs an automated geometric orientation check (affine vs mesh centroid vs metadata) on every case. Of 102 candidates, 0 are auto-promoted to “orientation verified”: 88 inconclusive, 14 flagged for review.

The library is presented as orientation QA pending — the system flags what it cannot yet certify rather than overclaiming. Per-case left/right and base/apex review is the next step before any clinical-teaching promotion.

Try it on a phone

Scan to open the touch + AR-capable viewer. Each model lazy-loads only when its 3D view is opened.

Hand off to phone

/v2/cases Open viewer here

Scan from your phone to launch the AR-capable viewer.

What this is not

  • Not a diagnostic tool and not a medical device.
  • Does not assign PI-RADS, stage disease, or guide biopsy.
  • Not clinically validated; 3D spatial orientation not yet verified per case.
  • Built only from public, consented research data (Prostate158, CC BY 4.0).