AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

K-Prism: A Knowledge-Guided and Prompt Integrated Universal Medical Image Segmentation Model

Bangwei Guo, Yunhe Gao, Meng Ye, Difei Gu, Yang Zhou, Leon Axel, Dimitris N. Metaxas

ICLR 2026 · Detection and segmentation › Open-set and few-shot detection · approach: Interactive and referring segmentation · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add MoE cross-attention (ablation)K-Prism decoder · semantic, ACDC+BUSDice: 77.38 -> 81.28Improved
add MoE FFN (ablation)K-Prism decoder · semantic, ACDC+BUSDice: 78.57 -> 81.28Improved
add both MoE CA and FFN (ablation)K-Prism decoder · semantic, ACDC+BUSDice: 76.77 -> 81.28Improved
add MoE CA and FFN (ablation)K-Prism decoder · in-context, ACDC+BUSDice: 75.10 -> 79.21Improved
add MoE CA and FFN (ablation)K-Prism decoder · interactive, ACDC+BUSDice(5): 92.22 -> 93.79Improved
add 2-D dense fusion (ablation)K-Prism in-context mode · in-context, ACDC+BUSDice: 54.65 -> 80.84Improved
add 1-D sparse queries (ablation)K-Prism in-context mode · in-context, ACDC+BUSDice: 77.19 -> 80.84Improved
add 2-D dense fusion (ablation)K-Prism interactive mode · interactive, ACDC+BUSDice(5): 79.21 -> 93.68Improved
add 1-D click queries (ablation)K-Prism interactive mode · interactive, ACDC+BUSDice(1): 85.44 -> 86.55Improved
initialize with Mode-1 semantic mask then refine with clicks (ablation)K-Prism interactive · ACDCNoC90: 2.67 -> 1.67Improved
initialize with Mode-2 in-context mask then refine with clicks (ablation)K-Prism interactive · BUSNoC90: 1.67 -> 1.17Improved
use in-context mode instead of semantic modeK-Prism · KiTSDice: 78.70 -> 79.12Mixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy