相关实验视频

Updated: Jun 8, 2026

Combining Intravital Fluorescent Microscopy IVFM with Genetic Models to Study Engraftment Dynamics of Hematopoietic Cells to Bone Marrow Niches

Combining Intravital Fluorescent Microscopy IVFM with Genetic Models to Study Engraftment Dynamics of Hematopoietic Cells to Bone Marrow Niches

Published on: March 21, 2017

多模式融合和功能增强U-Net与干细胞的合,对voxel-wise GBM复发预测的近距离估计

Changzhe Jiao¹, Yi Lao², Wenwen Zhang¹

¹Department of Radiation Oncology, UC San Francisco, San Francisco, CA 94143, United States of America.

Physics in medicine and biology

|July 17, 2024

概括

一个新的多模融合和特征增强U-Net (MFFE U-Net) 通过整合成像数据和利基的近距离,显著改善了质母细胞瘤 (GBM) 复发预测. 这种AI模型为GBM患者的早期检测和干预提供了一个有希望的工具.

科学领域:

医疗成像中的人工智能
神经瘤学神经瘤学
机器学习用于癌症预测

背景情况:

准确预测质母细胞瘤 (GBM) 复发对于及时调整治疗至关重要.
现有的方法经常与GBM成像数据的复杂性和多模式性质作斗争.
沃克塞尔智能预测需要能够处理高维图像特征的复杂模型.

研究的目的:

开发和验证多模式融合和特征增强U-Net (MFFE U-Net) 以改进质母细胞瘤 (GBM) 复发预测.
将干细胞利基近距离估计与深度学习相结合,以提高预测准确度.
为了比较MFFE U-Net与已建立的机器学习和深度学习模型的性能.

主要方法:

在四个数据库中对57名质母细胞瘤患者手术前和后的多模态MRI扫描进行了回顾性分析.
开发一种稀疏的多模特功能融合U-Net,结合干细胞利基近距离估计.
模型训练和验证使用70/10/20分割,在7名患者使用转移学习进行外部验证;通过精度,回忆,F1得分和豪斯多夫距离 (HD95) 评估性能.

主要成果:

在MFFE U-Net实现了高性能指标:精度0.79±0.08,召回0.85±0.11,和F1得分0.82±0.09.
与SVM_PE,mU-Net和Deeplabv3模型相比,观察到显著的改善.
低豪斯多夫距离 (HD95) 2.75 ± 0.44 毫米 (内部) 和 3.91 ± 0.83 毫米 (外部) 表示预测复发的精确定位.

结论:

多元金融基金U-Net代表了一种先进的深度学习框架,用于有效的voxel-wise GBM复发预测.
该模型在最先进的方法中表现出卓越的性能,突出显示了多模式融合和功能增强的好处.
这种方法有可能指导复发性质母细胞瘤的早期放射治疗干预.

关键词:

预测GBM的复发情况功能增强功能增强功能增强多模式融合融合多模式融合干细胞的位置是干细胞.

更多相关视频

Co-culture of Glioblastoma Stem-like Cells on Patterned Neurons to Study Migration and Cellular Interactions

Co-culture of Glioblastoma Stem-like Cells on Patterned Neurons to Study Migration and Cellular Interactions

Published on: February 24, 2021

A Rapid Screening Workflow to Identify Potential Combination Therapy for GBM using Patient-Derived Glioma Stem Cells

A Rapid Screening Workflow to Identify Potential Combination Therapy for GBM using Patient-Derived Glioma Stem Cells

Published on: March 28, 2021

相关概念视频

Stem Cell Niche

Stem Cell Niche

The stem cell niche is the dynamic microenvironment where stem cells reside. Inside these niches, the cells may remain undifferentiated, undergo high self-renewal, or become lineage-specific progenitors. Stem cells coexist with other niche cells, such as stromal cells. They also interact closely with the ECM. Cell-cell and cell-matrix communication occur via adhesion molecules or soluble factors that signal the stem cells and determine their fate. Stromal cells also provide survival signals to...

Multipotency and Niche of Bulge Stem Cell

Multipotency and Niche of Bulge Stem Cell

A hair follicle or HF is a small part of the skin that produces the hair shaft. Paul Gerson Unna was the first to observe a bulge in the human hair follicle's outer root sheath (ORS). The bulge is present between the sebaceous gland and the arrector pili muscle and is the niche for hair follicle stem cells (HFSCs). The bulge is also a niche for melanocyte stem cells, and their loss results in graying of hair. The HFSCs express Sox9 and Lhx2, which help them maintain stemness and prevent...

Neural Circuits

Neural Circuits

Neural circuits and neuronal pools are two of the main structures found in the nervous system. Neural circuits are networks of neurons that work together to carry out a specific task or process. They consist of interconnected neurons and glial cells, which provide structural and metabolic support.
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...