相关实验视频

Updated: Jul 21, 2025

Magnetic Levitation Coupled with Portable Imaging and Analysis for Disease Diagnostics

Magnetic Levitation Coupled with Portable Imaging and Analysis for Disease Diagnostics

Published on: February 19, 2017

一个可解释的视觉转换器模型基于白细胞分类和定位的白细胞分类和定位

Oguzhan Katar¹, Ozal Yildirim^1,2

¹Department of Software Engineering, Firat University, Elazig 23119, Turkey.

Diagnostics (Basel, Switzerland)

|July 29, 2023

概括

本研究介绍了一种可解释的视觉转换器 (ViT) 模型,用于自动地从血液图片中分类白细胞 (WBC),实现高精度并帮助病理学家检测疾病.

科学领域:

血液学血液学血液学
医疗成像医学成像
人工智能的人工智能

背景情况:

从血液图片中手动识别白细胞 (WBC) 亚型是费时的,容易出现错误.
像CNN这样的现有深度学习模型在图像分析中与远程依赖性和全球背景作斗争.
准确的WBC分类对于诊断感染,白血病和其他血液恶性瘤至关重要.

研究的目的:

开发一个可解释的视觉变压器 (ViT) 模型,用于自动检测和分类WBC从血液膜图像.
与手动方法和现有的自动化方法相比,提高WBC分析的准确性和效率.
为病理学家提供可靠和可解释的工具,以帮助疾病诊断.

主要方法:

利用视觉转换器 (ViT) 模型,结合自注意机制,从WBC图像中提取特征.
在一个公共数据集上训练并验证了该模型,其中包括五个亚型的16633个WBC样本.
进行了对花细胞和花细胞的二次二进制分类培训,以解决观察到的错误分类问题.
使用Score-CAM算法可视化模型预测并确保可靠性.

主要成果:

ViT模型在分类五种WBC亚型时达到99.40%的准确性.
对花细胞和花细胞的二进制分类ViT模型达到99.70%的准确性,99.54%的回忆,99.32%的精度和99.43%的F-1分数.
对错误分类的分析揭示了与细胞颗粒的存在或不存在的相关性.
评分-CAM可视化证实了该模型对相关图像区域的关注.

结论:

提出的可解释的ViT模型证明了自动化WBC分类的卓越性能和可靠性.
它能够捕捉全球背景并提供可解释的结果的能力使其适合临床应用.
这种人工智能驱动的方法可以显著提高WBC分析的效率和准确性,为病理学家提供更快,更精确的诊断.

关键词:

这就是Score-CAM.深度学习是一种深度学习.可解释的人工智能模型视觉变压器视觉变压器白细胞是白细胞的组成部分.

更多相关视频

Label-Free Identification of Lymphocyte Subtypes Using Three-Dimensional Quantitative Phase Imaging and Machine Learning

Label-Free Identification of Lymphocyte Subtypes Using Three-Dimensional Quantitative Phase Imaging and Machine Learning

Published on: November 19, 2018

Wide-field Fluorescent Microscopy and Fluorescent Imaging Flow Cytometry on a Cell-phone

Wide-field Fluorescent Microscopy and Fluorescent Imaging Flow Cytometry on a Cell-phone

Published on: April 11, 2013

相关概念视频

Vision

Vision

Vision is the result of light being detected and transduced into neural signals by the retina of the eye. This information is then further analyzed and interpreted by the brain. First, light enters the front of the eye and is focused by the cornea and lens onto the retina—a thin sheet of neural tissue lining the back of the eye. Because of refraction through the convex lens of the eye, images are projected onto the retina upside-down and reversed.

Anatomy of the Eyeball

Anatomy of the Eyeball

The eye is a spherical, hollow structure composed of three tissue layers. The outer layer — the fibrous tunic, comprises the sclera — a white structure — and the cornea, which is transparent. The sclera encompasses some of the ocular surface, most of which is not visible. However, the 'white of the eye' is distinctively visible in humans compared to other species. The cornea, a clear covering at the front of the eye, enables light penetration. The eye's middle...