相关实验视频

Updated: Jul 3, 2025

Combining Reflectance Confocal Microscopy with Optical Coherence Tomography for Noninvasive Diagnosis of Skin Cancers via Image Acquisition

Combining Reflectance Confocal Microscopy with Optical Coherence Tomography for Noninvasive Diagnosis of Skin Cancers via Image Acquisition

Published on: August 18, 2022

基于网格的结构和维度皮肤癌分类与自我特色优化可解释的深度卷积神经网络

Kavita Behara¹, Ernest Bhero², John Terhile Agee²

¹Department of Electrical Engineering, Mangosuthu University of Technology, Durban 4031, South Africa.

International journal of molecular sciences

|February 10, 2024

概括

一个新的基于网格的结构和维度可解释的深层卷积神经网络改善了皮肤癌分类的人工智能 (AI). 这种人工智能模型提高了早期检测的准确性和可解释性,优于现有的方法.

科学领域:

皮肤病学和人工智能研究
医学图像分析医学图像分析
医疗保健中的机器学习

背景情况:

传统的皮肤癌诊断是昂贵的,耗时的,需要专业的医疗专业人员.
人工智能 (AI) 提供了自动化皮肤癌诊断的潜力,但面临着复杂性,可重现性和可解释性方面的挑战.
早期发现皮肤癌对于有效治疗和患者的治疗结果至关重要.

研究的目的:

开发一种用于皮肤癌分类的新,准确和可解释的AI模型.
解决现有的AI诊断工具的局限性,特别是复杂性,可复制性和可解释性.
提高人工智能辅助早期皮肤癌检测的准确性和稳定性.

主要方法:

提出了一个基于网格的结构和维度可解释的深层卷积神经网络 (GSD-EDCNN).
使用适应性值来提取感兴趣区域 (ROI) 和VGG-16用于层次特征提取.
采用了自适应智能Coney优化 (AICO) 算法来进行超参数调整和自动功能选择.
在ISIC (10,015张图像) 和MNIST (2,357张图像) 数据集上训练并验证了模型.

主要成果:

在ISIC数据集上实现了高精度 (0.96) 和CSI (0.97),显著超过了各种已建立的CNN模型.
通过AICO优化模型,证明了0.03的最小假阳性率 (FPR) 和0.02的假阴性率 (FNR).
实现了低模型损失值 (ISIC为0.09,MNIST为0.18),表明性能优越.
与现有技术相比,该模型显示了改进的准确性,可解释性和稳定性.

结论:

拟议的GSD-EDCNN模型在人工智能驱动的皮肤癌分类方面取得了重大进展.
该模型的提高准确性和可解释性可以帮助临床医生在更早,更可靠的诊断.
这项研究有助于为皮肤学应用开发更有效的AI工具.

关键词:

在VGG-16中.适应性智能康伊优化算法可以解释的卷积神经网络.基于网格的结构模式.皮肤癌是皮肤癌.

更多相关视频

A 3D Organotypic Melanoma Spheroid Skin Model

A 3D Organotypic Melanoma Spheroid Skin Model

Published on: May 18, 2018

Quantitative Visualization and Detection of Skin Cancer Using Dynamic Thermal Imaging

Quantitative Visualization and Detection of Skin Cancer Using Dynamic Thermal Imaging

Published on: May 5, 2011

相关概念视频

Skin Cancer

Skin Cancer

Skin cancer is a type of cancer that occurs when there is an abnormal growth of skin cells, usually triggered by damage to the DNA within the skin cells. It is primarily caused by exposure to ultraviolet (UV) radiation from the sun or artificial sources like tanning beds. Skin cancer is the most common type of cancer worldwide, and its incidence continues to rise.
Basal Cell Carcinoma (BCC): BCC is the most common type of skin cancer, accounting for about 80% of cases. It typically develops in...

Renewal of Skin Epidermal Stem Cells

Renewal of Skin Epidermal Stem Cells

The skin is divided into epidermis, dermis, and hypodermis, the skin's outermost, middle, and inner layers. The human epidermal layer regularly undergoes renewal, where old, dead cells are replaced by new cells. Epidermal stem cells or EpiSCs divide and differentiate to restore the lost cells. For the renewal process, some EpiSCs continuously self-renew. In contrast, few others differentiate into transit-amplifying cells, which later form prickle or spinous cells, followed by granular...

Clinical Applications of Epidermal Stem Cells

Clinical Applications of Epidermal Stem Cells

Epidermal stem cells (EpiSCs) are mainly located at the basal layer of the epidermis. These cells repair minor injuries of the skin and replace dead skin cells. However, EpiSCs’ cannot heal severe wounds such as major burns or those from diabetes or hereditary disorders. In such cases, culturing the epidermal stem cells from the patient is possible and has yielded successful treatment options, such as laboratory-grown skin grafts. These grafts are synthesized using a patient’s own...