ProtoCouncil: Multi-Encoder Anatomical Prototypes for Prostate Cancer Detection in Micro-Ultrasound
Authors: Mohammad Mahdi Abootorabi, Nooshin Maghsoodi, Armin Saadat, Obed Dzikunu, Zhuoxin Guo, Hannes Cash, Ferdinand Luger, Adam Kinnaird, Brian Wodlinger, Parvin Mousavi, Purang Abolmaesumi
Organizations: The University of British Columbia, Vancouver, BC, Canada. · Vector Institute, Toronto, ON, Canada. · Queen’s University, Kingston, ON, Canada. · ProUro, Berlin, Germany. · Ordensklinikum Linz, Linz, Austria. · University of Alberta, Edmonton, AB, Canada. · Exact Imaging, Markham, ON, Canada.
Purpose: Micro-ultrasound (μUS) is a promising modality for prostate cancer (PCa) detection, but identifying suspicious tissue depends heavily on clinical experience. Models trained on sparse, noisy core-level pathology offer little insight into their decisions and can fail under domain shift. Clinical use needs a traceable decision process and a flag for uncertain predictions. Methods: We introduce ProtoCouncil, in which three foundation encoders (MedSAM, DINOv3 and V-JEPA 2.1) learn case-based prototypes at three anatomical levels: whole gland, needle track and gland sectors. Prototypes are projected onto real training-core regions, and label-free fusion keeps every class logit as a bias plus an exact sum of prototype similarity contributions. Encoder disagreement triggers selective prediction. Results: In five-fold patient-level cross-validation (6,588 cores, 693 patients), ProtoCouncil reaches 82.9% AUROC for clinically significant PCa (csPCa, Grade Group ≥3), 4.3 points above the best previously published method. In zero-shot evaluation on two external cohorts, ProtoCouncil achieves higher all-cancer and csPCa AUROC than all previous methods. Nearest same-class exemplars match test-core tumor involvement better than random same-class cases. Externally, csPCa accuracy falls from 79% when encoders agree to 27% when two oppose the fused call. Abstaining on disagreement raises balanced accuracy by 7 and 15 points, while retaining 58% and 54% of the cores. Conclusion: ProtoCouncil grounds core-level predictions in exact similarity-based evidence from real, anatomically indexed training cases, which makes its reasoning interpretable. It achieves the highest csPCa AUROC across three cohorts and identifies unreliable predictions through encoder disagreement.
Figures & tables
Cohort
Sites
Biopsy
Patients
Cores
Benign
GG1–2
GG ≥ 3
csPCa patients
A
5
Sys., TR
693
6,588
5,709 (86.7%)
480 (7.3%)
399 (6.1%)
117
B
2
Sys. + targ., TR
111
1,460
982 (67.3%)
364 (24.9%)
114 (7.8%)
28
C
1
Sys. + targ., TP
42
558
433 (77.6%)
105 (18.8%)
20 (3.6%)
6
Table 1: Characteristics of development cohort A and external cohorts B and C. Pathology categories show core counts (percentages). csPCa: GG ≥ 3; sys./targ.: systematic/targeted; TR/TP: transrectal/transperineal.
Figure 1: ProtoCouncil overview. Three encoders feed anatomical prototype heads whose logits are fused without labels. The model reports the fused prediction only when all three encoders agree; otherwise, it abstains. Bottom: each head compares pooled gland, needle-track, and gland-relative sector embeddings with prototypes from training regions.
All
High-inv.
csPCa
Patient
Method
AUROC
Sens@80
AUROC
AUROC
Sens@60
Sens@80
AUROC
Cohort A (development, five-fold cross-validation)
MedSAM-UNETR Alzate-Grisales et al. (2023)
70.1±2.1
49.4±4.6
79.7±1.1
71.4±4.0
71.4±6.1
50.7±7.5
67.4±9.2
MedSAM-FT Ma et al. (2024)
72.2±3.0
49.9±5.1
80.0±0.8
72.5±3.5
73.4±4.0
48.4±6.4
67.0±6.2
MicroSegNet Jiang et al. (2024)
76.3±1.3
57.7±3.1
83.2±1.2
77.6±4.6
78.4±7.0
62.7±6.7
73.7±7.5
ProstNFound Wilson et al. (2024)
77.1±2.9
60.3±4.6
82.6±1.9
77.7±4.8
82.3±7.1
62.1±9.2
72.2±3.3
Table 2: Prostate cancer detection in micro-ultrasound (%, mean ±SD over the five fold models). All: all cancer (GG ≥ 1); High-inv.: benign vs. cancer cores with > 40% involvement; csPCa: clinically significant cancer (GG ≥ 3). Best and second-best are shown in bold and underlined .
Figure 2: Results on the combined external zero-shot cohorts. (a) All-cancer AUROC by minimum cancer involvement. (b) Comparison with clinician-recorded PRI-MUS scores (1,919 cores). (c–e) Head, anatomical-family and encoder ablations: AUROC change relative to ProtoCouncil, with pointwise 95% bootstrap CIs. (f) Effect of clinical prompts.
Cohort A (70%)
Cohort B (58%)
Cohort C (54%)
BAcc
AUROC
BAcc
AUROC
BAcc
AUROC
No abstention
Prost-RL Abootorabi et al. (2026)
71.2
78.6
68.0
75.4
54.3
65.1
ProtoCouncil (ours)
74.1
82.9
72.4
80.0
70.3
79.5
Abstention
Prost-RL, confidence
77.6
81.9
76.2
81.4
70.4
81.9
Table 3: Selective csPCa prediction performance. Each gate abstains on the same share of cores as the disagreement gate, which abstains unless all three encoders agree (percentage of retained cores in parentheses). Balanced accuracy and csPCa AUROC of retained cores are reported.
Figure 3: Qualitative examples showing how clinicians can trace ProtoCouncil predictions to labeled training examples and compare the matched tissue patterns. For each encoder and anatomical family, each cell shows a prototype that strongly contributes to the prediction. Blue and orange mark prototype training regions and matched test regions, respectively. Letters (M, D, V) identify the encoder associated with each matched sector prototype. Labels report whole-core Grade Group and cancer involvement. Green frames mark pairs that an expert judged to show the same PRI-MUS pattern.
Micro-ultrasound (μUS) is a new, emerging, and promising imaging modality for prostate cancer (PCa) detection, but accurate identification of suspicious tissue remains highly dependent on clinical experience, leading to substantial inter-observer variability. Machine-learning assistance can reduce this variability; however, training reliable deep models is challenging because supervision is sparse and noisy -- typically limited to core-level histopathology outcomes (e.g., cancer grade and its percentage in a biopsy core) without pixel-level lesion annotations and under severe class imbalance. We introduce Prost-RL, which reframes μUS PCa detection as a spatially aware, policy-driven inference problem by learning where to look before decoding. Prost-RL integrates a lightweight reinforcement-learning policy into a foundation-model encoder-decoder to generate interpretable spatial attention maps that act as soft prompts for both cancer-likelihood heatmap prediction and image-level classification. We further propose Adaptive Policy Optimization (APO) to stabilize hybrid supervised-RL training and a noise-robust objective combining symmetric cross-entropy with negative-entropy regularization to mitigate weak-label noise and encourage sharp localization. On a cohort of 6,607 biopsy cores from 693 patients across five clinical sites, Prost-RL achieves 79.0±3.5 AUROC with 64.6±6.3% sensitivity at 80% specificity for core-level detection (+2.1 AUROC and +4.5 sensitivity points over the strongest baseline), and 79.3±5.8 AUROC for clinically significant cancer classification. The learned policy highlights biopsy-aligned regions, providing transparent, spatially grounded evidence alongside quantitative risk predictions. Code is available at: https://github.com/DeepRCL/Prost-RL.
Mohammad Mahdi Abootorabi, Sina Namazi, Armin Saadat +7
The University of British Columbia, Vancouver, BC, Canada · Vector Institute, Toronto, ON, Canada · Queen’s University, Kingston, ON, Canada +1
Purpose: Non-invasive grading of prostate cancer (PCa) from micro-ultrasound (micro-US) could expedite triage and guide biopsies toward the most aggressive regions, yet current models struggle to infer tissue micro-structure at coarse imaging resolutions. Methods: We introduce an unpaired histopathology knowledge-distillation strategy that trains a micro-US encoder to emulate the embedding distribution of a pretrained histopathology foundation model, conditioned on International Society of Urological Pathology (ISUP) grades. Training requires no patient-level pairing or image registration, and histopathology inputs are not used at inference. Results: Compared to the current state of the art, our approach increases sensitivity to clinically significant PCa (csPCa) at 60% specificity by 3.5% and improves overall sensitivity at 60% specificity by 1.2%. Conclusion: By enabling earlier and more dependable cancer risk stratification solely from imaging, our method advances clinical feasibility. Source code will be publicly released upon publication.
Emma Willis, Tarek Elghareb, Paul F. R. Wilson +6
Queen’s University, Kingston, Canada. · University of British Columbia, Vancouver, Canada∗. · Exact Imaging, Markham, Canada.
Prostate cancer claims a life every 80 seconds. Early detection is needed to prevent disease progression, and both PSA density calculation and biopsy decisions rely on knowing the exact boundary of the gland. Conventional ultrasound at 6-12 MHz blurs this boundary, missing one in three high-risk cancers. Micro-ultrasound (29 MHz) improves resolution threefold but introduces dense acoustic speckle that obscures the outer wall; given the same image, two clinicians draw outlines differing by over 10% in area. Supervised methods are costly and generalise poorly across scanners. Can a foundation model segment the prostate with no training data? We present the first zero-shot pipeline for this modality: MedSAM, pre-trained on over 1.5 million medical images, localises the prostate; we then apply CLAHE to sharpen the outer wall, binary dilation to recover missed pixels, and Fourier smoothing (4 modes, s=1.05) to refine the boundary. MedSAM requires a spatial prompt, so we evaluate bounding-box and point-click strategies across 75 patients of the Micro-Ultrasound Prostate Segmentation dataset (2,621 slices). On the 20-patient held-out test set, the pipeline reduces mean boundary-distance error by 45% (Dice 0.749+/-0.043 to 0.865+/-0.029; HD95 217.2+/-36.9 to 120.1+/-26.1 px), reaching Dice 0.859 across the cohort. Its mean overlap shows no significant difference from the three non-expert rater groups (p>0.19), while segmenting 38-52% more consistently (lower inter-patient standard deviation). Point-click prompts fail regardless of placement (best Dice=0.350), because speckle gives no stable local contrast. Only an approximate bounding box is required, so any clinic can deploy it without data collection, annotation, or retraining.