LoRKD: Low-Rank Knowledge Decomposition for Medical Foundation Models

• URL: http://arxiv.org/abs/2409.19540v1
• Date: Sun, 29 Sep 2024 03:56:21 GMT
• Title: LoRKD: Low-Rank Knowledge Decomposition for Medical Foundation Models
• Authors: Haolin Li, Yuhang Zhou, Ziheng Zhao, Siyuan Du, Jiangchao Yao, Weidi Xie, Ya Zhang, Yanfeng Wang,
• Abstract summary: "Knowledge Decomposition" aims to improve the performance on specific medical tasks. We propose a novel framework named Low-Rank Knowledge Decomposition (LoRKD) LoRKD explicitly separates gradients from different tasks by incorporating low-rank expert modules and efficient knowledge separation convolution.
• Score: 59.961172635689664
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The widespread adoption of large-scale pre-training techniques has significantly advanced the development of medical foundation models, enabling them to serve as versatile tools across a broad range of medical tasks. However, despite their strong generalization capabilities, medical foundation models pre-trained on large-scale datasets tend to suffer from domain gaps between heterogeneous data, leading to suboptimal performance on specific tasks compared to specialist models, as evidenced by previous studies. In this paper, we explore a new perspective called "Knowledge Decomposition" to improve the performance on specific medical tasks, which deconstructs the foundation model into multiple lightweight expert models, each dedicated to a particular anatomical region, with the aim of enhancing specialization and simultaneously reducing resource consumption. To accomplish the above objective, we propose a novel framework named Low-Rank Knowledge Decomposition (LoRKD), which explicitly separates gradients from different tasks by incorporating low-rank expert modules and efficient knowledge separation convolution. The low-rank expert modules resolve gradient conflicts between heterogeneous data from different anatomical regions, providing strong specialization at lower costs. The efficient knowledge separation convolution significantly improves algorithm efficiency by achieving knowledge separation within a single forward propagation. Extensive experimental results on segmentation and classification tasks demonstrate that our decomposed models not only achieve state-of-the-art performance but also exhibit superior transferability on downstream tasks, even surpassing the original foundation models in task-specific evaluations. The code is available at here.

Related papers

• KA$^2$ER: Knowledge Adaptive Amalgamation of ExpeRts for Medical Images Segmentation [5.807887214293438]
  We propose an adaptive amalgamation knowledge framework that aims to train a versatile foundation model to handle the joint goals of multiple expert models. In particular, we first train an nnUNet-based expert model for each task, and reuse the pre-trained SwinUNTER as the target foundation model. Within the hidden layer, the hierarchical attention mechanisms are designed to achieve adaptive merging of the target model to the hidden layer feature knowledge of all experts.
  arXiv  Detail & Related papers  (2024-10-28T14:49:17Z)
• FEDKIM: Adaptive Federated Knowledge Injection into Medical Foundation Models [54.09244105445476]
  This study introduces a novel knowledge injection approach, FedKIM, to scale the medical foundation model within a federated learning framework. FedKIM leverages lightweight local models to extract healthcare knowledge from private data and integrates this knowledge into a centralized foundation model. Our experiments across twelve tasks in seven modalities demonstrate the effectiveness of FedKIM in various settings.
  arXiv  Detail & Related papers  (2024-08-17T15:42:29Z)
• Low-Rank Knowledge Decomposition for Medical Foundation Models [37.52464627899668]
  We develop a new perspective called Knowledge Decomposition'' to improve the performance on specific medical tasks. Low-Rank Knowledge Decomposition (LoRKD) incorporates low-rank expert modules and the efficient knowledge separation convolution. Experiments show that decomposed models perform well in terms of performance and transferability, even surpassing the original foundation models.
  arXiv  Detail & Related papers  (2024-04-26T06:30:47Z)
• Rethinking Model Prototyping through the MedMNIST+ Dataset Collection [0.11999555634662634]
  This work presents a benchmark for the MedMNIST+ database to diversify the evaluation landscape. We conduct a thorough analysis of common convolutional neural networks (CNNs) and Transformer-based architectures, for medical image classification. Our findings suggest that computationally efficient training schemes and modern foundation models hold promise in bridging the gap between expensive end-to-end training and more resource-refined approaches.
  arXiv  Detail & Related papers  (2024-04-24T10:19:25Z)
• On the Out of Distribution Robustness of Foundation Models in Medical Image Segmentation [47.95611203419802]
  Foundations for vision and language, pre-trained on extensive sets of natural image and text data, have emerged as a promising approach. We compare the generalization performance to unseen domains of various pre-trained models after being fine-tuned on the same in-distribution dataset. We further developed a new Bayesian uncertainty estimation for frozen models and used them as an indicator to characterize the model's performance on out-of-distribution data.
  arXiv  Detail & Related papers  (2023-11-18T14:52:10Z)
• Learning Through Guidance: Knowledge Distillation for Endoscopic Image Classification [40.366659911178964]
  Endoscopy plays a major role in identifying any underlying abnormalities within the gastrointestinal (GI) tract. Deep learning, specifically Convolution Neural Networks (CNNs) which are designed to perform automatic feature learning without any prior feature engineering, has recently reported great benefits for GI endoscopy image analysis. We investigate three KD-based learning frameworks, response-based, feature-based, and relation-based mechanisms, and introduce a novel multi-head attention-based feature fusion mechanism to support relation-based learning.
  arXiv  Detail & Related papers  (2023-08-17T02:02:11Z)
• Learnable Weight Initialization for Volumetric Medical Image Segmentation [66.3030435676252]
  We propose a learnable weight-based hybrid medical image segmentation approach. Our approach is easy to integrate into any hybrid model and requires no external training data. Experiments on multi-organ and lung cancer segmentation tasks demonstrate the effectiveness of our approach.
  arXiv  Detail & Related papers  (2023-06-15T17:55:05Z)
• Ambiguous Medical Image Segmentation using Diffusion Models [60.378180265885945]
  We introduce a single diffusion model-based approach that produces multiple plausible outputs by learning a distribution over group insights. Our proposed model generates a distribution of segmentation masks by leveraging the inherent sampling process of diffusion. Comprehensive results show that our proposed approach outperforms existing state-of-the-art ambiguous segmentation networks.
  arXiv  Detail & Related papers  (2023-04-10T17:58:22Z)
• Few-shot Medical Image Segmentation using a Global Correlation Network with Discriminative Embedding [60.89561661441736]
  We propose a novel method for few-shot medical image segmentation. We construct our few-shot image segmentor using a deep convolutional network trained episodically. We enhance discriminability of deep embedding to encourage clustering of the feature domains of the same class.
  arXiv  Detail & Related papers  (2020-12-10T04:01:07Z)
• Siloed Federated Learning for Multi-Centric Histopathology Datasets [0.17842332554022694]
  This paper proposes a novel federated learning approach for deep learning architectures in the medical domain. Local-statistic batch normalization (BN) layers are introduced, resulting in collaboratively-trained, yet center-specific models. We benchmark the proposed method on the classification of tumorous histopathology image patches extracted from the Camelyon16 and Camelyon17 datasets.
  arXiv  Detail & Related papers  (2020-08-17T15:49:30Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.