Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
Neighborhood Self-Dissimilarity Attention for Medical Image Segmentation
Authors: Junren Chen, Rui Chen, Wei Wang, Junlong Cheng, Gang Liang, zhanglei-scu, Liangyin Chen
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | The experimental results demonstrate the effectiveness and generalization of our method. This study presents a parameter-free attention paradigm, designed with clinical prior knowledge, to improve neural network performance for medical image analysis and contribute to digital health equity in low-resource settings. |
| Researcher Affiliation | Academia | Junren Chen1, Rui Chen2, Wei Wang3, Junlong Cheng1, Gang Liang4 , Lei Zhang1 , Liangyin Chen1 1 College of Computer Science, Sichuan University, Chengdu, China 2 Department of Electronic Engineering, Tsinghua University, Beijing, China 3 School of Automation, Chengdu University of Information Technology, Chengdu, China 4 School of Cyber Science and Engineering, Sichuan University, Chengdu, China EMAIL |
| Pseudocode | No | The paper describes the methodology using narrative text and mathematical equations (Equations 1-6) and provides figures (e.g., Figure 1 for an overview of NSDA) but does not include any explicitly labeled pseudocode or algorithm blocks. |
| Open Source Code | Yes | The code is available at https://github.com/Chen Junren-Lab/Neighborhood-Self-Dissimilarity-Attention. |
| Open Datasets | Yes | Datasets. We evaluate the effectiveness of NSDA on three prominent medical image segmentation benchmarks with diverse modalities and scales: Synapse (multi-organ abdominal CT) [45], ACDC (cardiac MRI) [5], and BUSI (breast ultrasound) [1]. To thoroughly assess generalization capabilities for the proposed NSDA, we extend validation to COVID-19 pneumonia lesion detection (CPLDet) [51], endoscopic bladder tissue classification (EBTCls) [38], and natural image segmentation (VOCSeg, combining PASCAL VOC07 [17] and VOC12 [65]). |
| Dataset Splits | Yes | We adopt different splits for each benchmark: a 6:2:2 ratio (train/validation/test) for BUSI, an 8:1:1 ratio for NCPDet, and a 9:1 train-test split for VOCSeg. |
| Hardware Specification | Yes | We conduct experiments on an NVIDIA Ge Force RTX 4090 GPU using the Py Torch framework. |
| Software Dependencies | No | The paper mentions the use of the Py Torch framework and the Adam optimizer but does not specify their version numbers, which is necessary for a reproducible description of software dependencies. |
| Experiment Setup | Yes | The models are trained for 300 epochs using the Adam optimizer [37], with a composite loss function that combines cross-entropy and Dice loss [12]. The initial learning rate is set to 1 10 4 and decayed using a cosine annealing scheduler, with a minimum value of 1 10 6. |