Training-Free Medical Image Segmentation vs Trained Networks

A multi-domain data-efficiency study — frozen-foundation-model assembly (an LLM polygon drawer in the deployed path on 1/6 domains, else SAM / op_sam / SegGPT) + a numpy genome, zero gradient training on the target task.

6 modalitiesmeasured, from scratch GT-free at inferenceLLM-agnostic skills updated · 2026-07-07

Data-efficiency

The crossover — one panel per domain

Frozen assembly vs the strongest compatible, complete three-seed trained NN at each K, selected from U-Net, SegFormer and nnU-Net only. Filled comparator dot + whisker = exactly seeds 0/1/2 and a 95% df=2 t interval; absent dot = no eligible comparator. The tuned SAM decoder GT-box oracle is kept separate in the expanded figure below.

▬ pipeline (strict first-K curve; Spleen and Prostate updated to NN-assisted SkillOpt) · □/■ strongest trained NN

ISIC 2018 · dermoscopyassembly ahead

Kvasir · endoscopy polypnear-tie

MSD Spleen · CTNN-assisted pipeline · strict first-K

MSD Prostate · MRINN-assisted pipeline · strict first-K

Chest X-ray · lungassembly ahead

Strict leakage-free Spleen data efficiency

K Training supports	K+24 adaptation labels	NN-assisted pipeline Dice [95% volume CI]	Strongest compatible trained NN Dice [95% df=2 t CI]
5	29	0.759 [0.682, 0.828]	U-Net 0.380 [0.219, 0.541]
10	34	0.828 [0.759, 0.891]	U-Net 0.577 [0.530, 0.623]
25	49	0.845 [0.760, 0.911]	SegFormer 0.772 [0.687, 0.856]
50	74	0.814 [0.732, 0.901]	U-Net 0.855 [0.717, 0.992]
100	124	0.905 [0.860, 0.947]	U-Net 0.912 [0.864, 0.959]

The green Spleen curve is the fresh NN-assisted pipeline at every K. It uses three training seeds and its NN-specific crop scales, routing, agreement gates, morphology, margins and selectors were independently tuned at every K; baseline intervals are across exact seeds 0/1/2. Resize-224 nnU-Net and historical cached-mask SAM scores are excluded.

Strict leakage-free Prostate data efficiency

K Training supports	K+20 adaptation labels	NN-assisted pipeline Dice [95% volume CI]	Strongest compatible trained NN Dice [95% df=2 t CI]
5	25	0.321 [0.184, 0.472]	U-Net 0.596 [0.533, 0.658]
10	30	0.505 [0.341, 0.672]	U-Net 0.700 [0.599, 0.800]
25	45	0.687 [0.584, 0.789]	nnU-Net 0.791 [0.783, 0.799]
50	70	0.789 [0.711, 0.860]	nnU-Net 0.795 [0.780, 0.810]
100	120	0.792 [0.728, 0.860]	nnU-Net 0.846 [0.834, 0.858]

The green Prostate curve uses fresh masks from three NN seeds. Its NN-specific preprocessing, localization, crop-zoom, gates, margins and selectors were independently tuned at every K. Green intervals bootstrap the six complete Test volumes; baseline intervals use exact seeds 0/1/2.

The green Spleen pipeline curve strictly uses the first K Training supports. Its selector was calibrated with 24 separate Validation labels, so the disclosed total is K+24 adaptation labels. Green error bars are 95% complete-volume bootstrap intervals over the seven held-out Test volumes.

The green Prostate pipeline curve also strictly uses the first K Training supports. Its selector used 20 separate Validation labels, so the disclosed total is K+20 adaptation labels. Green error bars bootstrap the six held-out Test volumes.

Coverage is intentionally incomplete. No uncertainty bar is synthesized for one/two-seed cells or the single-run tuned-decoder oracle. The expanded figure below shows every cached method curve and makes the missing-seed cells explicit.

HC18 FetalHead remains withheld pending its leakage-free rerun; these updates do not restore any prior FetalHead result.

Detailed data-efficiency figure — five settled domains with updated Spleen and Prostate NN-assisted pipelines

Data-efficiency curves with strict first-K Spleen and Prostate pipelines updated to independently tuned NN-assisted SkillOpt — Five settled domains, five K budgets; **FetalHead remains withheld**. Green is the strict first-K pipeline curve. Spleen and Prostate use **fresh NN-assisted pipelines with independently tuned NN-specific SkillOpt at every K**; their green whiskers bootstrap seven and six complete Test volumes. The separate Validation selector labels remain disclosed as K+24 and K+20 total adaptation labels. U-Net, SegFormer and nnU-Net use protocol-compatible cached score artifacts; filled squares and whiskers appear only for exact seeds 0/1/2 and show 95% df=2 t intervals. Resize-224 Spleen nnU-Net and historical cached-mask SAM scores are excluded. Exact values, label counts, seed counts and source paths are in `dataeff_grid.json`.

Results — best trained NN across N vs the frozen-assembly pipeline

Strongest trained NN per cell — max of SegFormer (mit_b2), pretrained U-Net (resnet34), and nnU-Net, 2-seed means — at matched budgets N = 10 / 25 / 50 / 100, against the LLM pipeline (re-measured from scratch via eval_all_domains.py). The LLM's only label-fit module is a ~5–8-param genome; localization and SAM are GT-free/frozen — so the LLM number is ~flat in N while the NN climbs. Caveat: the N=100 column is a set of within-noise near-ties — best-NN@100 sits inside the LLM bootstrap 95% CI on 5/6 domains (stats.md), and 0/4 apparent N=100 leads survive as significant. The honest, CI-separated lead is at N≤25. Amber = within-CI near-tie; a green cell reflects the point estimate only, not significance.

Domain	best NN N=10	N=25	N=50	N=100	training-free LLM	few-shot decoder N=100 (oracle)
ISIC 2018 (derm)	0.815	0.829	0.832	0.868	0.880	0.911
MSD Spleen (CT)	0.633	0.802	0.859	0.907	0.920	0.953
Kvasir-SEG (endo)	0.756	0.824	0.833	0.889	0.8646	0.944
ChestXray (X-ray)	0.938	0.948	0.952	0.9546	0.9497 SegGPT in-context; fs-deploy 0.9488	0.9535
HC18 Fetal-Head (US) ·new	0.923	0.958	0.965	0.970	0.9667 retrieval SegGPT; was 0.9328 fs / 0.9225 tf	0.9665 tuned-decoder oracle
MSD Prostate (MRI) ·new	0.679	0.720	0.772	0.815	0.7877 SegGPT⊕op_sam ensemble; was 0.760	0.938

Two LLM columns: the training-free pipeline (zero gradient training — the study's headline) and the few-shot tuned decoder (generic-SAM init, encoder frozen, ~4M-param mask decoder tuned on N=100 labels, GT-box oracle). The few-shot decoder beats the from-scratch NN on four of six domains given a good box (ISIC/Spleen/Kvasir/Prostate) and ties the other two (CXR, HC18 — near-ceiling NNs at ~0.95–0.96) — the residual between it and the deployed training-free pipeline is mostly localization (the exception was CXR, where the frozen SAM decoder was the wall — now moved training-free by swapping in a different frozen decoder, the SegGPT in-context drawer, to 0.9497), detailed below.

CXR is now training-free again at 0.9497 — a swapped 4th frozen drawer backend (SegGPT in-context). The old CXR training-free headline (0.922, frozen box-SAM) is superseded by a different frozen decoder: SegGPT (BAAI/seggpt-vit-large, generic → leak-free, non-SAM) paints the whole lung mask in-context from K=8 spread-sampled train (image,mask) support pairs — no box, no SAM, no localizer, ZERO gradient training. Deployed 0.9497 (median 0.9616, n=107), within 0.005 of the from-scratch NN's 0.9546 — a near-tie, still not a win. This is the 4th backend at the model-coupled drawer seam (LLM polygon · frozen SAM · few-shot tuned decoder · in-context SegGPT): the seam is genuinely swappable, and CXR's gap was neither architectural nor requiring our labels — it was the particular frozen decoder (SAM). A separate, non-training-free few-shot lane remains documented: tuning only the ~4M-param SAM mask decoder on N=100 CXR labels (generic-vit_l init, encoders frozen, box-jittered, leak-free) + mask-bbox re-decode reaches 0.9488 (median 0.9608, 102/107 improve) — but the deployed CXR is now the training-free SegGPT 0.9497, which supersedes it. Frozen-decoder study (leak-free generic): among frozen alternatives the generic in-context SegGPT wins; SAM2 / SAM1-vit_h merely tie frozen SAM-HQ (no gain), and text-prompted CLIPSeg fails on grayscale X-ray (0.00) — the interesting frozen alternative is a generic in-context segmenter (few-shot at inference, zero gradient), not another SAM variant.

The pipeline wins the EXTREME scarce-label regime (N≤25) where the target is GT-free-localizable — ISIC and Spleen are CI-separated leads at N≤25 (fewshot_matched.md); at N=100 the numbers below (ISIC 0.880 vs 0.868, Spleen 0.920 vs 0.907) are within-CI near-ties, not significant leads (stats.md) — and on Kvasir polyp a frozen-DINOv2 dense-correspondence localizer (OP-SAM CPG + peak-CC box) lifts it 0.659→0.8646 — the gap was localization (frozen GT-box oracle 0.914 > NN), and this closes 89% of it; the median 0.9484 now beats the NN mean, though the mean stays just below best-NN 0.889. On Chest X-ray the deployed training-free path is now a swapped 4th frozen drawer — the SegGPT in-context drawer (K=8 support pairs, zero gradient) — reaching 0.9497 ≈ NN 0.9546 (−0.005, a near-tie), superseding the earlier box-SAM 0.922 and the separate few-shot-tuned decoder 0.9488 lane; the old "frozen-decoder cap ~0.93" was specific to SAM, not to architecture. On HC18 fetal-head the same story now plays out training-free: a swapped retrieval-prompted SegGPT drawer (K=8 query-retrieved DINOv2-similar in-context supports, zero gradient, SAM-free) reaches 0.9667 ≈ NN 0.9703 (−0.004, a near-tie), superseding the op_sam+SAM 0.9225 and the few-shot decoder 0.9328 — so FetalHead is no longer a few-shot lane. Honest headline: at N=100 these are within-CI near-ties — 0/4 apparent leads survive bootstrap 95% CIs (stats.md), and the Spleen "+0.019" is a seed-averaging artifact (best single U-Net seed 0.9323 > LLM 0.9262). The defensible lead is at N≤25 on localizable targets (ISIC / Kvasir / Spleen / Prostate), budget-matched and CI-separated. All 6 deployed paths are gradient-free frozen-model assembly, but the LLM drawer is out of the mask-producing path on 5/6 (llm_ablation.md) — Prostate, the last domain lifted off its floor, is a complementary 2nd frozen drawer (a SegGPT⊕op_sam edge-gated keep-best ensemble). SegFormer is the strongest NN in 10 of 12 sweep cells.

Segmentation galleries

Per-case GT-vs-prediction overlays for the deployed pipeline on every test image (all 465, sorted worst-Dice first) — open the 6-domain gallery hub → (ISIC · Kvasir · Spleen · Prostate · CXR · FetalHead). Green = ground truth, red = deployed prediction; overlays rendered from the committed mask_cache/.

Held-out validation-set galleries (the deployed pipeline on each domain's calibration split, distinct from the test cases and the reported test numbers) — open the 5-domain validation gallery hub → (ISIC · Kvasir · Spleen · Prostate · CXR; FetalHead withheld pending the HC18 subject-level leak fix).

Where is the wall — the decoder, or localization? A fair few-shot probe

Is the residual to the trained NN a decoder limit or a localization limit? MedSAM (SAM fine-tuned on ~1M medical masks) beats box-prompted SAM — but it is data-leaky (its training set likely includes these test distributions), so it is really "import a trained medical net." The fair, leak-free counterpart of the same operation: init from the generic SAM (no medical data), freeze the image + prompt encoders, and fine-tune only the ~4M-param mask decoder on our own N-label train split — GT never touches test. Given a good box (the GT-box oracle, which isolates the decoder), this few-shot-tuned decoder beats the from-scratch NN at every budget on 4 of 6 domains (ISIC/Spleen/Kvasir/Prostate — Prostate by +0.12) and ties on the other two (CXR, HC18 — both near-ceiling NNs at ~0.95–0.96), and it saturates at N=10 (pretrained features + a tiny adapted head), so its curve is flat-high while the NN climbs:

Domain (tuned decoder, oracle)	N=10	N=25	N=50	N=100	best NN N=100
ISIC 2018 (derm)	0.905	0.904	0.907	0.911	0.868
MSD Spleen (CT)	0.957	0.960	0.960	0.953	0.907
Kvasir-SEG (endo)	0.936	0.934	0.937	0.944	0.889
ChestXray (X-ray)	0.933	0.946	0.955	0.9535	0.9546
HC18 Fetal-Head (US) ·new	0.918	0.954	0.957	0.9665	0.970
MSD Prostate (MRI) ·new	0.906	0.931	0.932	0.938	0.815

Mostly localization — with CXR the telling exception. The oracle hands the decoder a good box; on most domains the gap between it and the deployed pipeline is localization. Kvasir is the proof: tuned-decoder oracle 0.944 vs the original deployed 0.659 — 0.285 of pure localization headroom, zero decoder; the frozen dense-correspondence localizer has since realized most of it (0.659→0.8646), exactly as this predicted, leaving only ~0.05 to the oracle box. Where localization is already solved, the tuned decoder is a clean deployable win over both production and the NN: CXR few-shot decoder 0.9488 > the earlier box-SAM production 0.922 ≈ NN 0.9546 (though the deployed training-free CXR is now the SegGPT in-context drawer 0.9497, which supersedes both); Spleen (3D volume-prior box) 0.942 > production 0.920 > NN 0.907. This is few-shot (N=100 labels = the NN's budget), not training-free — an explicit, leak-free counterpart to MedSAM. CXR is the exception that sharpens the rule: its wall was the particular frozen decoder (SAM), not localization and not architecture — swapping in a different frozen decoder, the SegGPT in-context drawer, moves it training-free to 0.9497 (≈ NN, zero gradient), and a few-shot-tuned SAM decoder independently reaches 0.9488. So of the residual "structural / architectural ceilings," the polyp boundary was a localization gap in disguise, while the CXR "decoder wall" was a SAM-specific wall breachable by another frozen decoder with none of our labels.

Is the medical pretraining (MedSAM) worth it? — a leaky benchmark

MedSAM is SAM already fine-tuned on ~1M medical masks — a leaky upper reference (its training likely saw these test distributions). Arch-matched (both vit_b, GT-box oracle, N=100), few-shot tuning of the GENERIC SAM matches or beats the leaky MedSAM on 3 of 6 domains, and stays within 0.015 on the other 3 — its own pretraining distributions (derm/US/MRI). Medical pretraining buys no consistent edge once the decoder is adapted, and where generic SAM is already strong (Kvasir/Spleen) it starts above MedSAM even frozen. Bold = better of the two few-shot columns.

Domain (N=100 oracle, vit_b)	from-scratch NN	generic frozen	generic few-shot (fair)	MedSAM frozen (leaky)	MedSAM few-shot (leaky)
ISIC 2018 (derm)	0.868	0.870	0.915	0.921	0.943
MSD Spleen (CT)	0.907	0.941	0.956	0.909	0.946
Kvasir-SEG (endo)	0.889	0.917	0.936	0.888	0.918
ChestXray (X-ray)	0.9546	0.925	0.954	0.935	0.952
HC18 Fetal-Head (US) ·new	0.970	0.769	0.965	0.904	0.972
MSD Prostate (MRI) ·new	0.815	0.895	0.928	0.807	0.941

Fair few-shot generic SAM vs leaky MedSAM benchmark, N=100 oracle — Fair few-shot tuning of **generic** SAM (dark blue) vs the leaky **MedSAM** (orange), each frozen + few-shot-tuned, vs the from-scratch NN (gray) — arch-matched vit_b, N=100 GT-box oracle. The fair generic few-shot ≥ leaky MedSAM few-shot on Spleen/Kvasir/ChestXray (**3/6**); MedSAM leads on its own pretraining distributions — derm (ISIC), US (HC18), MRI (prostate) — but by **<0.015** once the generic decoder is tuned, and generic-frozen still beats MedSAM-frozen on prostate (0.895 vs 0.807). Medical pretraining helps most exactly where MedSAM saw the data, yet few-shot tuning of the generic decoder nearly erases it — decoder adaptation, not medical pretraining, is the lever.

The MedSAM attack inverts once it is prompted fairly and tested OOD (ood_medsam.md). The oracle-box column above is a leaky upper reference. Give MedSAM a fair box (learned from N GT, not the oracle) and it drops +0.16..+0.59 Dice below the deployed pipeline in-distribution; take it out-of-distribution (DRIVE retinal vessels — a distribution MedSAM never trained on) and the gap widens to +0.51 while MedSAM collapses to 0.220 even handed the ORACLE box (vs ~0.90 in-distribution). MedSAM's in-distribution edge was its decoder recognizing box-shaped targets it had trained on — memorization that does not transfer; the pipeline's few-shot-refit GT-free localization is distribution-robust.

Can it be ONE domain-agnostic pipeline? — few-shot skills + a router, SAM frozen

Each domain has a distinct setup — so can it be a single domain-agnostic pipeline that few-shot-fits per domain WITHOUT tuning SAM? We test the transferable skill — a frozen-SAM-feature localizer prototype fit from N labeled images — with identical code across all six domains, SAM never touched.

Domain-agnostic few-shot skill transfer: skills few-shot-fit, topology is routed not tuned — **(A)** One fixed pipeline (prototype-sim localizer): the skill **few-shot-fits** — saturates by ~N=10 on every domain — and scores high where the target is prototype-localizable (ISIC 0.71, CXR 0.62, HC18 US 0.72), mid where it is a central structure (Prostate MRI 0.51), low where it needs a specialized localizer (Kvasir 0.35, Spleen 0.12). **(B)** **Routing** the per-domain topology (genome / static-box / op_sam dense-correspondence / volume-prior / ellipse-snap / volume-trajectory) + the few-shot skill recovers most of the deployed number (Kvasir 0.35→0.77→0.86, HC18 0.69→0.92, Prostate 0.51→0.76, Spleen 0.14→0.66→0.83@N=50).

Yes — a domain-agnostic pipeline = frozen SAM + a GT-free router + few-shot skills. The learning (localizer prototype + gate weights + genome) is domain-agnostic, few-shot (~10 labels; ~50 for the isodense spleen), and SAM-frozen — it transfers with identical code. The per-domain "distinct setup" you'd notice is which localizer topology to route to — a GT-free routing choice (learned config, the project's route.py), not SAM tuning. So onboarding a new domain is ~10 labels + a routing decision, zero gradient training. Robust localizer skill: the deployed default is now the op_sam frozen-DINOv2 dense-correspondence localizer (kvasir/HC18/prostate) — it transfers 10-shot support masks through the full patch cross-correlation, and superseded the earlier SAM+DINOv2 AMG-ranker ensemble on Kvasir (0.655→0.8646). The lesson holds and is now stronger: the polyp/fold and faint-skull walls yield to the right frozen feature space — dense correspondence, plus feature-diverse supports — not to distractor-verification tricks. The AMG-ranker ensemble (at/near the best single feature space on all three, Kvasir 0.65 / Spleen 0.64 / ISIC 0.67, never catastrophic where SAM-alone fails the isodense spleen 0.40) remains the LLM-agnostic fallback where a support set is too small for correspondence.

agenticSeg · training-free frozen-foundation-model assembly for medical segmentation (LLM drawer in the deployed path on 1/6 domains) · github.com/R9-Hu/agenticSeg
Skills are LLM-agnostic numpy/config; every result reproducible via eval_all_domains.py. AI-assisted (Claude Code).