FSS-UBrain: Multi-region Few-Shot Brain Tumor MRI Segmentation

Truong Viet Vu; Nguyen Phuc Nguyen; Dang Thi Thu Hang; Tran Thien Thanh; Vo Nguyen Quoc Bao; Nguyen Thai Anh; Ngo Hoang Tu

cs.CVSep 26, 2026

FSS-UBrain: Multi-region Few-Shot Brain Tumor MRI Segmentation

Authors: Truong Viet Vu, Nguyen Phuc Nguyen, Dang Thi Thu Hang, Tran Thien Thanh, Vo Nguyen Quoc Bao, Nguyen Thai Anh, Ngo Hoang Tu

Organizations: Faculty of Information Technology, Van Lang School of Technology, Van Lang University, Ho Chi Minh City 70000, Vietnam · Institute of Information Technology and Electrical-Electronics Engineering, Ho Chi Minh City University of Transport, Ho Chi Minh City 70000, Vietnam

Abstract

Accurate delineation of whole tumor (WT), tumor core (TC), and enhancing tumor (ET) from multimodal magnetic resonance imaging remains challenging under limited annotation, cross-cohort variation, and severe target sparsity. We propose FSS-UBrain, a region-wise one-shot segmentation framework that uses a labeled positive support slice to condition binary query segmentation separately for WT, TC, and ET. Support-derived foreground and background descriptors guide query-feature adaptation, bottleneck interaction, decoder-side reconstruction, and boundary refinement. Episodic training additionally incorporates hard-negative and fully negative queries with empty-query regularization to suppress spurious foreground activation when the selected region is absent. Although inference operates on two-dimensional support--query slice pairs, checkpoint selection, threshold calibration, and final evaluation are performed after volumetric reconstruction. FSS-UBrain is evaluated on a held-out BraTS 2020 split and under target-supported cross-cohort protocols on BraTS 2023 and BraTS-Africa. Cases used as target support are excluded from the query cohorts, and no target-domain fine-tuning or test-time parameter updates are performed. On BraTS 2020, FSS-UBrain achieves volumetric Dice scores of 89.82%, 82.14%, and 77.42% for WT, TC, and ET, respectively, with corresponding 95th-percentile Hausdorff distance (HD95) values of 11.12, 9.01, and 4.46 mm. It also achieves the highest mean Dice and lowest finite-pair mean HD95 point estimates on BraTS 2023 and BraTS-Africa among the compared few-shot methods. These findings support target-conditioned few-shot segmentation while highlighting sensitivity to support selection and cohort-specific variation.

Figures & tables

Panel of Figure 1 — Figure 1: Representative BraTS slices illustrating ET sparsity and severe slice-level target-presence imbalance.

Figure 2: Overview of the proposed FSS-UBrain model architecture.

Figure 3: Support-conditioned descriptor adaptation at encoder level $\ell$ . Masked support features form FG and BG descriptors that are query-refined, matched to query features, and used for descriptor gating and FiLM modulation.

Preprocessing or Augmentation Step and Procedure	Purpose	Applied During
Modality organization. Each axial input is formed by stacking FLAIR, T1, T1ce, and T2 as a four-channel slice.	Preserves the standard multimodal BraTS input representation.	Training, validation, and testing
Region-mask construction. Binary WT, TC, and ET masks are generated using the unified region definitions described in Section 3.1 .	Ensures a consistent clinical-region interpretation across datasets.	Training, validation, and testing
Per-case normalization. For each modality, the mean and standard deviation are computed from nonzero voxels across the complete case volume. Only nonzero voxels are standardized, while zero-valued BG remains zero.	Reduces inter-case intensity variation without altering the zero-valued BG convention.	Training, validation, and testing
Spatial resizing. Images are resized to $256\times 256$ using bilinear interpolation; binary masks are resized using nearest-neighbor interpolation.	Provides a fixed input size while avoiding interpolation-induced label mixing.	Training, validation, and testing
Horizontal and vertical flips. Each image–mask pair is independently flipped horizontally and vertically with probability $0.5$ .	Increases invariance to in-plane orientation variation.	Training only
Right-angle rotation. With probability $0.5$ , an image–mask pair is rotated by $0^{\circ}$ , $90^{\circ}$ , $180^{\circ}$ , or $270^{\circ}$ .	Improves robustness to in-plane orientation variation while preserving voxel-grid alignment.	Training only

Table 1: Preprocessing and training-time augmentation applied to support and query inputs.

Evaluation Metric and Formula	Main Advantages	Main Limitations
Dice coefficient. This overlap measure is defined as ( Milletari et al., 2016 ; Taha and Hanbury, 2015 ) $\displaystyle\operatorname{Dice}(M_{r},\widehat{M}_{r})=\frac{2TP+\epsilon_{3}}{2TP+FP+FN+\epsilon_{3}},$ (10) where $\epsilon_{3}=10^{-6}$ is the small constant used for numerical stability. Dice is the voxel-wise $F_{1}$ score of the predicted FG region and is reported as a percentage.	Primary overlap measure for imbalanced binary segmentation; weights $FP$ and $FN$ equally while excluding $TN$ , so it is not inflated by the large BG volume; comparable in scale with the BraTS literature.	Invariant to error location; merges over- and under-segmentation into one value; unstable for sparse regions such as ET; degenerate for jointly empty pairs.
HD95. The 95th-percentile Hausdorff distance is defined as ( Taha and Hanbury, 2015 ) $\displaystyle\operatorname{HD95}(M_{r},\widehat{M}_{r})=Q_{0.95}\!\left(\mathcal{D}(\partial M_{r},\partial\widehat{M}_{r})\uplus\mathcal{D}(\partial\widehat{M}_{r},\partial M_{r})\right),$ (11) where $\uplus$ denotes concatenation with multiplicities preserved, $Q_{0.95}$ returns the 95th percentile of the resulting distance collection, whereas $\partial M_{r}$ and $\partial\widehat{M}_{r}$ denote the surfaces of the reference and predicted masks, respectively. For two surfaces $A$ and $B$ , $\mathcal{D}(A,B)$ denotes the collection of directed nearest-surface distances from $A$ to $B$ .	Robust surface-distance measure in physical space that reduces sensitivity to isolated outlier voxels; interpretable as a clinical margin; sensitive to the geometric placement of errors that overlap metrics cannot express.	Undefined when exactly one mask is empty, yielding a finite-pair summary over a varying subset; conveys no volumetric information; depends on voxel spacing; the discarded $5\%$ tail may hide extreme deviations.
Intersection over union (IoU). The overlap between prediction and reference is ( Taha and Hanbury, 2015 ) $\displaystyle\operatorname{IoU}(M_{r},\widehat{M}_{r})=\frac{TP}{TP+FP+FN}.$ (12) Also known as the Jaccard index, it satisfies $\operatorname{IoU}=\operatorname{Dice}/(2-\operatorname{Dice})$ for an individual volume–region pair, with both quantities expressed as fractions.	Supplementary overlap measure that penalizes false positives and false negatives symmetrically; stricter than Dice, hence separates methods whose Dice values are compressed near the upper end of the scale.	Monotone transform of Dice per pair and therefore largely redundant; shares the insensitivity of Dice to error location; absolute values are not comparable with Dice-based results reported elsewhere.
Precision. The reliability of predicted FG voxels is given by ( Taha and Hanbury, 2015 ) $\displaystyle\operatorname{Precision}(M_{r},\widehat{M}_{r})=\frac{TP}{TP+FP}.$ (13) It is equivalently the positive predictive value.	Proportion of predicted FG voxels that are correct; isolates the $FP$ component that Dice merges with $FN$ , thereby quantifying over-segmentation.	Driven close to its maximum by conservative predictions, hence uninformative unless read jointly with recall; degenerate on target-absent pairs; disregards the spatial distribution of false positives.
Recall. The recovery of reference FG voxels is ( Taha and Hanbury, 2015 ) $\displaystyle\operatorname{Recall}(M_{r},\widehat{M}_{r})=\frac{TP}{TP+FN}.$ (14) It is equivalently the sensitivity or true-positive rate.	Proportion of reference FG voxels recovered by the prediction; quantifies under-segmentation, the error type with the greatest clinical cost in surgical and radiotherapy planning.	Attains its maximum for any model labeling the entire brain as FG; entirely insensitive to false positives; degenerate on target-absent pairs.
Specificity. The correct rejection of BG voxels is ( Taha and Hanbury, 2015 ) $\displaystyle\operatorname{Specificity}(M_{r},\widehat{M}_{r})=\frac{TN}{TN+FP}.$ (15) It is equivalently the true-negative rate.	Proportion of BG voxels correctly rejected as BG; describes the negative class, which the overlap metrics exclude; provides a volume-wide check against indiscriminate FG activation.	Saturates near $100\%$ because BG voxels dominate the volume, hence offers little resolution among models; a high value may still correspond to a clinically significant number of false-positive voxels.

Table 2: Definitions, main advantages, and main limitations of the volume-level evaluation metrics.

Figure 4: Training and validation dynamics of the compared models on BraTS20.

	WT		TC		ET
Model	Dice	HD95	Dice	HD95	Dice	HD95
ALPNet ( Ouyang et al., 2020 )	$76.38\pm 6.39$	$41.48\pm 3.86$	$56.74\pm 10.96$	$47.80\pm 18.76$	$47.88\pm 4.35$	$36.36\pm 23.24$
PANet ( Wang et al., 2019 )	$80.44\pm 5.38$	$32.48\pm 2.46$	$75.33\pm 2.06$	$26.14\pm 7.86$	$62.65\pm 2.78$	$15.65\pm 2.64$
Q-Net ( Shen et al., 2023 )	$82.74\pm 0.58$	$37.23\pm 1.53$	$66.53\pm 10.30$	$45.36\pm 4.22$	$55.06\pm 3.23$	$40.43\pm 3.15$
CATNet ( Lin et al., 2023b )	$82.48\pm 0.10$	$37.70\pm 0.98$	$68.89\pm 2.60$	$39.01\pm 14.16$	$69.34\pm 1.17$	$18.08\pm 5.25$
GMRD ( Cheng et al., 2024 )	$85.98\pm 1.05$	$26.13\pm 3.64$	$79.09\pm 1.20$	$24.16\pm 5.95$	$68.95\pm 7.75$	$15.57\pm 9.13$

Table 3: BraTS20 few-shot results reported as mean

\pm

sample standard deviation over five deterministic support draws. Best point estimates are bold.

	WT		TC		ET
Model	Dice	HD95	Dice	HD95	Dice	HD95
U-Net ( Ronneberger et al., 2015 )	86.20	28.01	80.20	25.78	75.72	17.65
UNet++ ( Zhou et al., 2019 )	86.35	27.61	80.61	27.68	74.77	23.74
TransUNet ( Chen et al., 2021 )	82.72	40.91	75.24	39.16	70.43	16.79
SwinUNet ( Cao et al., 2022 )	84.29	24.24	74.01	23.28	68.23	19.07
VMUNet ( Ruan et al., 2025 )	88.32	18.41	80.77	20.97	75.09	13.44

Table 4: BraTS20 volume-level Dice and HD95 for non-FSS baselines and FSS-UBrain. The best point estimate in each column is shown in bold.

	BraTS-Africa		BraTS23
Model	Mean Dice	Mean HD95	Mean Dice	Mean HD95
ALPNet ( Ouyang et al., 2020 )	$66.83\pm 0.85$	$40.40\pm 2.05$	$69.15\pm 4.32$	$40.50\pm 4.10$
PANet ( Wang et al., 2019 )	$74.37\pm 1.08$	$16.29\pm 1.01$	$76.77\pm 0.28$	$21.94\pm 1.17$
Q-Net ( Shen et al., 2023 )	$68.83\pm 6.04$	$28.94\pm 0.43$	$71.02\pm 0.76$	$42.39\pm 0.02$
CATNet ( Lin et al., 2023b )	$70.65\pm 2.01$	$26.56\pm 1.95$	$71.34\pm 0.43$	$36.21\pm 2.61$
GMRD ( Cheng et al., 2024 )	$77.65\pm 0.86$	$18.46\pm 1.14$	$81.84\pm 0.98$	$13.44\pm 1.01$

Table 5: Target-supported cross-cohort mean results. BraTS-Africa values are mean

\pm

sample standard deviation over five deterministic support draws; BraTS23 values are mean

\pm

sample standard deviation over two deterministic support draws (seeds 1337 and 1338). Best point estimates are bold.

	Source support		Target support
Model	Mean Dice	Mean HD95	Mean Dice	Mean HD95
ALPNet ( Ouyang et al., 2020 )	$61.32\pm 2.82$	$38.07\pm 11.75$	$66.83\pm 0.85$	$40.40\pm 2.05$
PANet ( Wang et al., 2019 )	$73.05\pm 1.71$	$16.90\pm 0.77$	$74.37\pm 1.08$	$16.29\pm 1.01$
Q-Net ( Shen et al., 2023 )	$70.60\pm 2.49$	$29.45\pm 1.29$	$68.83\pm 6.04$	$28.94\pm 0.43$
CATNet ( Lin et al., 2023b )	$72.59\pm 1.21$	$25.82\pm 3.46$	$70.65\pm 2.01$	$26.56\pm 1.95$
GMRD ( Cheng et al., 2024 )	$\mathbf{76.62\pm 2.18}$	$18.33\pm 2.43$	$77.65\pm 0.86$	$18.46\pm 1.14$

Table 6: BraTS-Africa mean results under source and target support (mean

\pm

sample standard deviation over five draws; same 85 query cases). Best point estimates within each support condition are bold.

Metric	Full model	w/o DMP	w/o SGB	w/o PC	w/o BR
Mean Dice	83.13	82.33	82.42	82.75	82.98
Mean HD95	8.20	9.27	9.96	9.86	9.59

Table 7: BraTS20 mean component-ablation results across WT, TC, and ET. Best point estimates are bold.

Panel of Figure 5 — Figure 5: Qualitative few-shot comparisons using four-modality MRI input; only FLAIR is displayed. SP denotes support, GT ground truth, and Pred prediction.

Cohort	Train Hard Neg	Train Full Neg	Val Hard Neg	Val Full Neg	Mean Dice	Mean HD95	Mean IoU	Precision	Recall	Specificity	FP-neg-rate
BraTS20	0	0	0	0	75.46	41.97	63.93	72.82	83.33	99.89	0.35
	20	0	25	0	82.72	10.68	73.91	82.11	86.99	99.93	0.16
	0	10	0	15	81.52	12.81	72.41	80.68	86.57	99.92	0.23
	10	5	15	10	83.21	11.42	74.27	82.09	87.06	99.93	0.15
	20	10	25	15	83.13	8.20	74.34	82.59	86.81	99.93	0.10
	30	15	35	20	84.19	7.61	75.90	83.75	87.45	99.94	0.12

Table 8: BraTS20 sensitivity to hard- and fully negative-query sampling. Mean Dice and mean HD95 are unweighted averages across WT, TC, and ET; ratios denote training hard/full and validation hard/full sampling. Best point estimates are bold.

Model	Params (M)	FLOPs (G)	MACs (G)
Without DMP	19.83	200.45	100.22
Without SGB	41.82	351.05	175.53
Without PC	36.88	328.39	164.19
Without BR	42.61	347.00	173.50
FSS-UBrain (Full model)	42.65	352.05	176.03

Table 9: Model-side complexity of FSS-UBrain ablations per

4\times 256\times 256

one-shot support–query episode at batch size one. Test-time augmentation and volume post-processing are excluded. Lowest values are bold.

Model	Params (M)	FLOPs (G)	MACs (G)
SwinUNet ( Cao et al., 2022 )	6.94	6.90	3.45
U-Net ( Ronneberger et al., 2015 )	7.70	73.91	36.96
UNet++ ( Zhou et al., 2019 )	36.62	276.83	138.42
TransUNet ( Chen et al., 2021 )	105.32	77.09	38.55
VMUNet ( Ruan et al., 2025 )	44.28	19.66	9.82
SAM ( Kirillov et al., 2023 )	93.74	976.50	488.25

Table 10: Model-side complexity at batch size one. Support-conditioned methods are profiled per support–query episode and non-FSS methods per query slice; test-time augmentation and volume post-processing are excluded. Lowest values are bold separately for the non-FSS group (upper block, including SAM) and the support-conditioned group (lower block, including FSS-UBrain).

Explore similar work

RUFNet: Query-Guided Support Mask Refinement and Uncertainty Fusion based on Hybrid Mamba for Few-Shot Brain Tumor Segmentation
Jul 6, 2026Dongyi He, Xiangkai Wang, Binbing Xu +5Few-Shot Medical Image Segmentation Segmentation Uncertainty Estimation
Department of Language Science and Technology, The Hong Kong Polytechnic University, Hong Kong SAR, China · School of Artificial Intelligence, Chongqing University of Technology, Chongqing, China · Affiliated Lianyungang Hospital of Xuzhou Medical University, Lianyungang, China +1
SegGuidedNet: Sub-Region-Aware Attention Supervision for Interpretable Brain Tumor Segmentation
May 21, 2026Hasaan Maqsood, Saif Ur Rehman Khan, Sebastian Vollmer +2Image Segmentation BraTS
National University of Sciences and Technology (NUST), Islamabad, Pakistan. · Department of Computer Science, Rhineland-Palatinate Technical University of Kaiserslautern-Landau, Kaiserslautern, 67663, Germany. · German Research Center for Artificial Intelligence, Kaiserslautern, 67663, Germany. +2
A Multimodal Feature Distillation with Mamba-Transformer Network for Brain Tumor Segmentation with Incomplete Modalities
Apr 22, 2024Ming Kang, Fung Fung Ting, Shier Nee Saw +3Cross-Modal Knowledge Distillation Cross-Modality Medical Image Segmentation
School of Information Technology, Monash University, Malaysia Campus, Subang Jaya 47500, Malaysia · Augmented Intelligence and Multimodal analytics (AIM) for Health Lab, Faculty of Information Technology, Monash University, Clayton VIC 3800, Australia · Monash Medical AI Group, Monash University, Clayton VIC 3800, Australia +1

Jul 6, 2026J/K move · Enter open · S save

RUFNet: Query-Guided Support Mask Refinement and Uncertainty Fusion based on Hybrid Mamba for Few-Shot Brain Tumor Segmentation

Dongyi He, Xiangkai Wang, Binbing Xu, Bin Jiang+4

Department of Language Science and Technology, The Hong Kong Polytechnic University, Hong Kong SAR, China · School of Artificial Intelligence, Chongqing University of Technology, Chongqing, China · Affiliated Lianyungang Hospital of Xuzhou Medical University, Lianyungang, China+1

Few-Shot Medical Image Segmentation Segmentation Uncertainty Estimation Brain Tumor Segmentation

Open paper AskPDF arXiv

Few-shot brain tumor segmentation remains challenging due to noisy support masks, inter-patient variations between support and query images, and the lack of pixel-wise confidence estimation. This study proposes RUFNet, a Hybrid Mamba-based few-shot framework that combines support mask refinement with uncertainty-aware posterior fusion. To preserve support-query dependencies with manageable cost, RUFNet adopts a Hybrid Mamba interaction backbone with linear complexity. To reduce support-mask noise, an Attention-Guided Mask Refinement module (AGMR) uses query features to recalibrate support masks and improve prototype consistency. To handle ambiguous predictions, an Uncertainty-Aware Posterior Fusion module (UAPF) estimates pixel-wise variance and adaptively balances few-shot predictions with query-aligned priors. On the Brain Tumor Segmentation Challenge (BraTS) 2020 dataset, RUFNet achieves Dice coefficients of 84.3% and 86.1% in the 1-way 1-shot and 1-way 5-shot settings, respectively, outperforming the compared state-of-the-art methods. These results suggest that Hybrid Mamba interaction, mask refinement and uncertainty modelling can improve the robustness of few-shot medical image segmentation. The official implementation code is available at https://github.com/hdy6438/RUFNet.

Abstract

Explore similar work

RUFNet: Query-Guided Support Mask Refinement and Uncertainty Fusion based on Hybrid Mamba for Few-Shot Brain Tumor Segmentation

SegGuidedNet: Sub-Region-Aware Attention Supervision for Interpretable Brain Tumor Segmentation

A Multimodal Feature Distillation with Mamba-Transformer Network for Brain Tumor Segmentation with Incomplete Modalities