相关实验视频

Updated: Jul 19, 2025

Reconstruct Human Retinoblastoma In Vitro

Reconstruct Human Retinoblastoma In Vitro

Published on: October 11, 2022

视网膜母细胞瘤瘤在 fundus 图像中的半监督细分

Amir Rahdar¹, Mohamad Javad Ahmadi¹, Masood Naseripour²

¹Chashmyar Company, Tehran, Iran.

Scientific reports

|August 10, 2023

概括

本研究引入了一种半监督机器学习模型,用于早期检测视网膜母细胞瘤. 该模型在从眼底图像中对眼睛异常进行细分时达到93%的准确性,有助于早期癌症诊断.

科学领域:

眼科医生眼科眼科
医疗成像医学成像
人工智能的人工智能

背景情况:

视网母细胞瘤是幼儿的主要眼内恶性瘤.
视网母细胞瘤的早期检测提供了>90%的治愈率.
准确的瘤细分对于有效的治疗计划至关重要.

研究的目的:

开发一个半监督的机器学习模型用于视网膜母细胞瘤细分在 fundus 图像.
为了达到与专家眼科医生可比的细分精度.
创建一个具有成本效益的自动检测系统.

主要方法:

利用高斯混合模型在4200个 fundus图像中检测初始异常.
使用初始检测结果训练了一种具有成本效益的模型,以提高效率.
采用半监督机器学习用于图像分割.

主要成果:

拟议的模型使用Sørensen-Dice系数实现了93%的平均准确性.
在从 fundus 图像中提取详细的瘤边界方面表现出高精度.
该模型提供了与医疗专家可比的结果.

结论:

半监督机器学习为准确的视网膜母细胞瘤细分提供了一种可行的方法.
开发的模型可以帮助早期诊断和治疗计划的视网膜母细胞瘤.
自动细分可以提高视网膜母细胞瘤检测的效率和一致性.

更多相关视频

Subretinal Transplantation of Human Embryonic Stem Cell Derived-retinal Pigment Epithelial Cells into a Large-eyed Model of Geographic Atrophy

Subretinal Transplantation of Human Embryonic Stem Cell Derived-retinal Pigment Epithelial Cells into a Large-eyed Model of Geographic Atrophy

Published on: January 22, 2018

Establishment and Propagation of Human Retinoblastoma Tumors in Immune Deficient Mice

Establishment and Propagation of Human Retinoblastoma Tumors in Immune Deficient Mice

Published on: August 4, 2011

相关概念视频

The Retinoblastoma Gene

The Retinoblastoma Gene

Tumor suppressor genes are normal genes that can slow down cell division, repair DNA mistakes, or program the cells for apoptosis in case of irreparable damage. Hence, they play an essential role in preventing the proliferation of damaged cells.
The first-ever tumor suppressor gene called Rb was identified in retinoblastoma - a rare eye tumor in children. In inherited forms of the disease, a child inherits one defective copy of the Rb gene, which predisposes them to retinoblastoma. However,...