Scholarlink

<p>Combining traditional imaging techniques with artificial intelligence (AI) is fundamentally changing dermatologic diagnostics. Digital dermoscopy allows temporal monitoring of skin lesions and detects minimal changes using convolutional neural networks. AI calculates scores such as the DEXI (dermoscopy explainable intelligence) score based on color, shape, and size. Three-dimensional total-body scanners digitally capture the entire skin surface and enable objective monitoring in skin cancer screening and inflammatory dermatoses. Confocal laser scanning microscopy provides high-resolution optical sections both in vivo and ex vivo, with lateral resolution up to 1 μm. The first deep learning models are able to reliably detect tumor patterns such as basal cell or squamous cell carcinoma and enable automated segmentation for intraoperative margin control. Optical coherence tomography (OCT) provides layered visualization of superficial skin alterations and supports lesion differentiation, e.g., in basal cell carcinoma, actinic keratosis, or psoriasis. Line-field OCT (LC-OCT) combines high resolution with depth penetration and enables cellular-level imaging. AI automatically segments skin layers and keratinocyte nuclei, supports assessment of atypia, and visualizes the likelihood of tumor nests. By combining human expertise with AI assistance, diagnostic accuracy is increased while simultaneously improving efficiency in treatment planning and follow-up monitoring.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Dermatologische Diagnostik im Wandel

  • 期刊论文

  • 发表日期：

    2025年6月13日

查看全文
  • Maximilian Deußing

摘要

Combining traditional imaging techniques with artificial intelligence (AI) is fundamentally changing dermatologic diagnostics. Digital dermoscopy allows temporal monitoring of skin lesions and detects minimal changes using convolutional neural networks. AI calculates scores such as the DEXI (dermoscopy explainable intelligence) score based on color, shape, and size. Three-dimensional total-body scanners digitally capture the entire skin surface and enable objective monitoring in skin cancer screening and inflammatory dermatoses. Confocal laser scanning microscopy provides high-resolution optical sections both in vivo and ex vivo, with lateral resolution up to 1 μm. The first deep learning models are able to reliably detect tumor patterns such as basal cell or squamous cell carcinoma and enable automated segmentation for intraoperative margin control. Optical coherence tomography (OCT) provides layered visualization of superficial skin alterations and supports lesion differentiation, e.g., in basal cell carcinoma, actinic keratosis, or psoriasis. Line-field OCT (LC-OCT) combines high resolution with depth penetration and enables cellular-level imaging. AI automatically segments skin layers and keratinocyte nuclei, supports assessment of atypia, and visualizes the likelihood of tumor nests. By combining human expertise with AI assistance, diagnostic accuracy is increased while simultaneously improving efficiency in treatment planning and follow-up monitoring.