Can Foundation Models Really Segment Tumors? A Benchmarking Odyssey in Lung CT Imaging

• URL: http://arxiv.org/abs/2505.01239v1
• Date: Fri, 02 May 2025 13:04:01 GMT
• Title: Can Foundation Models Really Segment Tumors? A Benchmarking Odyssey in Lung CT Imaging
• Abstract summary: The complexity of tumor morphology, size, and location poses significant challenges for automated segmentation.<n>We compare traditional architectures such as U-Net and DeepLabV3, self-configuring models like nnUNet, and foundation models like MedSAM, and MedSAM2.<n>The results reveal that while traditional models struggle with tumor delineation, foundation models, particularly MedSAM2, outperform them in both accuracy and computational efficiency.
• Score: 25.093744722130594
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Accurate lung tumor segmentation is crucial for improving diagnosis, treatment planning, and patient outcomes in oncology. However, the complexity of tumor morphology, size, and location poses significant challenges for automated segmentation. This study presents a comprehensive benchmarking analysis of deep learning-based segmentation models, comparing traditional architectures such as U-Net and DeepLabV3, self-configuring models like nnUNet, and foundation models like MedSAM, and MedSAM~2. Evaluating performance across two lung tumor segmentation datasets, we assess segmentation accuracy and computational efficiency under various learning paradigms, including few-shot learning and fine-tuning. The results reveal that while traditional models struggle with tumor delineation, foundation models, particularly MedSAM~2, outperform them in both accuracy and computational efficiency. These findings underscore the potential of foundation models for lung tumor segmentation, highlighting their applicability in improving clinical workflows and patient outcomes.

Related papers

• A Unified Benchmark of Deep Learning Models for Multi-task 3D Brain Tumor Segmentation from Magnetic Resonance Imaging [0.0]
  This work presents a unified benchmark for comparing representative convolutional neural networks (CNNs) and Transformer-based models.<n>Five state-of-the-art three-dimensional segmentation models, including 3D U-Net, SegResNet, Swin UNETR, SegMamba, and SegMambaV2, are evaluated.<n>Performance is assessed using segmentation accuracy metrics together with computational cost indicators, including inference time and the size of each model.
  arXiv  Detail & Related papers  (2026-07-30T21:56:03Z)
• Investigating the Impact of Histopathological Foundation Models on Regressive Prediction of Homologous Recombination Deficiency [52.50039435394964]
  We systematically evaluate foundation models for regression-based tasks.<n>We extract patch-level features from whole slide images (WSI) using five state-of-the-art foundation models.<n>Models are trained to predict continuous HRD scores based on these extracted features across breast, endometrial, and lung cancer cohorts.
  arXiv  Detail & Related papers  (2026-01-29T14:06:50Z)
• Demystifying Deep Learning-based Brain Tumor Segmentation with 3D UNets and Explainable AI (XAI): A Comparative Analysis [1.5958130875154202]
  The study focused on applying UNet models for brain tumor segmentation.<n>Three deep learning models were evaluated to identify the best-performing model.<n>ResUNet was found to be the best-performing model.
  arXiv  Detail & Related papers  (2025-10-09T05:03:31Z)
• Fine-grained Multi-class Nuclei Segmentation with Molecular-empowered All-in-SAM Model [15.622506758735476]
  We propose the molecular-empowered All-in-SAM Model to advance computational pathology.<n>This model incorporates a full-stack approach, focusing on: (1) annotation-engaging lay annotators through molecular-empowered learning to reduce the need for detailed pixel-level annotations, and (2) learning-adapting the SAM model to emphasize specific semantics.
  arXiv  Detail & Related papers  (2025-08-21T17:49:21Z)
• Multi-encoder nnU-Net outperforms Transformer models with self-supervised pretraining [0.0]
  This study addresses the essential task of medical image segmentation, which involves the automatic identification and delineation of anatomical structures and pathological regions in medical images.<n>We propose a novel self-supervised learning Multi-encoder nnU-Net architecture designed to process multiple MRI modalities independently through separate encoders.<n>Our Multi-encoder nnU-Net demonstrates exceptional performance, achieving a Dice Similarity Coefficient (DSC) of 93.72%, which surpasses that of other models such as vanilla nnU-Net, SegResNet, and Swin UNETR.
  arXiv  Detail & Related papers  (2025-04-04T14:31:06Z)
• MAST-Pro: Dynamic Mixture-of-Experts for Adaptive Segmentation of Pan-Tumors with Knowledge-Driven Prompts [54.915060471994686]
  We propose MAST-Pro, a novel framework that integrates dynamic Mixture-of-Experts (D-MoE) and knowledge-driven prompts for pan-tumor segmentation.<n>Specifically, text and anatomical prompts provide domain-specific priors guiding tumor representation learning, while D-MoE dynamically selects experts to balance generic and tumor-specific feature learning.<n>Experiments on multi-anatomical tumor datasets demonstrate that MAST-Pro outperforms state-of-the-art approaches, achieving up to a 5.20% improvement in average improvement while reducing trainable parameters by 91.04%, without compromising accuracy.
  arXiv  Detail & Related papers  (2025-03-18T15:39:44Z)
• GS-TransUNet: Integrated 2D Gaussian Splatting and Transformer UNet for Accurate Skin Lesion Analysis [44.99833362998488]
  We present a novel approach that combines 2D Gaussian splatting with the Transformer UNet architecture for automated skin cancer diagnosis.<n>Our findings illustrate significant advancements in the precision of segmentation and classification.<n>This integration sets new benchmarks in the field and highlights the potential for further research into multi-task medical image analysis methodologies.
  arXiv  Detail & Related papers  (2025-02-23T23:28:47Z)
• Detection-Guided Deep Learning-Based Model with Spatial Regularization for Lung Nodule Segmentation [2.4044422838107438]
  Lung cancer ranks as one of the leading causes of cancer diagnosis and is the foremost cause of cancer-related mortality worldwide. The segmentation of lung nodules plays a critical role in aiding physicians in distinguishing between malignant and benign lesions. We introduce a novel model for segmenting lung nodules in computed tomography (CT) images, leveraging a deep learning framework that integrates segmentation and classification processes.
  arXiv  Detail & Related papers  (2024-10-26T11:58:12Z)
• Segmentation Strategies in Deep Learning for Prostate Cancer Diagnosis: A Comparative Study of Mamba, SAM, and YOLO [0.6116681488656472]
  This study presents a comparative analysis of three deep learning-based methods, Mamba, SAM, and YOLO, for segmenting prostate cancer histopathology images. We evaluated the performance of these models on two comprehensive datasets, Gleason 2019 and SICAPv2, using Dice score, precision, and recall metrics. The H-vmunet model's advanced architecture, which integrates high-order visual state spaces and 2D-selective-scan operations, enables efficient and sensitive lesion detection.
  arXiv  Detail & Related papers  (2024-09-24T16:04:29Z)
• Transformer-Based Self-Supervised Learning for Histopathological Classification of Ischemic Stroke Clot Origin [0.0]
  Identifying the thromboembolism source in ischemic stroke is crucial for treatment and secondary prevention. This study describes a self-supervised deep learning approach in digital pathology of emboli for classifying ischemic stroke clot origin.
  arXiv  Detail & Related papers  (2024-05-01T23:40:12Z)
• Computational Pathology at Health System Scale -- Self-Supervised Foundation Models from Three Billion Images [30.618749295623363]
  This project aims to train the largest academic foundation model and benchmark the most prominent self-supervised learning algorithms by pre-training. We collected the largest pathology dataset to date, consisting of over 3 billion images from over 423 thousand microscopy slides. Our results demonstrate that pre-training on pathology data is beneficial for downstream performance compared to pre-training on natural images.
  arXiv  Detail & Related papers  (2023-10-10T21:40:19Z)
• Competence-based Multimodal Curriculum Learning for Medical Report Generation [98.10763792453925]
  We propose a Competence-based Multimodal Curriculum Learning framework ( CMCL) to alleviate the data bias and make best use of available data. Specifically, CMCL simulates the learning process of radiologists and optimize the model in a step by step manner. Experiments on the public IU-Xray and MIMIC-CXR datasets show that CMCL can be incorporated into existing models to improve their performance.
  arXiv  Detail & Related papers  (2022-06-24T08:16:01Z)
• Improving Classification Model Performance on Chest X-Rays through Lung Segmentation [63.45024974079371]
  We propose a deep learning approach to enhance abnormal chest x-ray (CXR) identification performance through segmentations. Our approach is designed in a cascaded manner and incorporates two modules: a deep neural network with criss-cross attention modules (XLSor) for localizing lung region in CXR images and a CXR classification model with a backbone of a self-supervised momentum contrast (MoCo) model pre-trained on large-scale CXR data sets.
  arXiv  Detail & Related papers  (2022-02-22T15:24:06Z)
• A multi-stage machine learning model on diagnosis of esophageal manometry [50.591267188664666]
  The framework includes deep-learning models at the swallow-level stage and feature-based machine learning models at the study-level stage. This is the first artificial-intelligence-style model to automatically predict CC diagnosis of HRM study from raw multi-swallow data.
  arXiv  Detail & Related papers  (2021-06-25T20:09:23Z)
• CoRSAI: A System for Robust Interpretation of CT Scans of COVID-19 Patients Using Deep Learning [133.87426554801252]
  We adopted an approach based on using an ensemble of deep convolutionalneural networks for segmentation of lung CT scans. Using our models we are able to segment the lesions, evaluatepatients dynamics, estimate relative volume of lungs affected by lesions and evaluate the lung damage stage.
  arXiv  Detail & Related papers  (2021-05-25T12:06:55Z)

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.