Identification of Critical Dermoscopic Features

描述(由申请人提供)： 据估计，恶性黑色素瘤的发病率几十年来每年增长约6%，造成相当大的生命损失。然而，如果及早发现，黑色素瘤很容易治愈。数字皮肤镜显示出了更准确的检测前景，特别是在早期阶段。最近的会议强调了对皮肤镜特征定义的普遍一致，以及对最有用的结构特征的适度一致。自动检测这些对早期诊断至关重要并用于各种皮肤镜诊断算法的特定结构将是可取的。然而，在自动检测任何特定的皮肤镜结构方面发表的工作很少。此外，在相当多的病例中，由于系统未能正确地将病变与背景分开，数字皮肤镜系统的诊断准确性受到限制。虽然特定的颜色在最关键的皮肤镜结构的定义中占据了突出的位置，但在寻找黑色素瘤颜色所在的颜色空间中的特定区域方面做的工作很少，特别是参考周围的皮肤。 该建议旨在通过以下方式提高数字皮肤镜系统的性能：1)以更高的精度找到边界2)开发一种算法，该算法使用概率密度函数的三维表示通过聚类法和模糊逻辑技术来指定黑色素瘤的颜色3)高精度地识别关键结构特征，包括棕色小球、边界突然切断、粒度、回归和色素不对称4)开发用于临床图像采集的临床界面5)开发用于图像交互分析的网络工具。这项研究的主要特点包括皮肤病理学对特定结构的确认和相对颜色分析的使用。如果成功，具体的算法将与越来越多使用数字皮肤镜的皮肤科医生分享。在第二阶段，将对算法进行进一步测试，并开发快速接口。一个结合了软件组件的商业套装将被提供给流行的数码相机-光头组合。

DESCRIPTION (provided by applicant): Malignant melanoma, with an estimated growth in incidence of about 6% per year for decades, causes considerable loss of life. Yet melanoma can be easily cured if detected early. Digital dermoscopy has shown promise for more accurate detection, particularly at an early stage. Recent conferences have highlighted a general agreement on definition of dermoscopic features and moderate agreement on the most useful structural features. Automatic detection of these specific structures that are critical for early diagnosis and are used in various dermoscopic diagnostic algorithms would be desirable. Yet little work has been published on automatic detection of any specific dermoscopic structures. In addition, diagnostic accuracy of digital dermoscopic systems is limited by the failure of systems to properly separate the lesion from the background in a significant number of cases. Although specific colors figure prominently in the definition of the most critical dermoscopic structures, little work has been done on finding the specific regions in the color space where melanoma colors are located, particularly with reference to the surrounding skin. This proposal seeks to improve performance of digital dermoscopy systems by 1) finding borders with greater accuracy 2) developing an algorithm that uses a three-dimensional representation of a probability density function to specify melanoma colors via cluster methods and fuzzy logic techniques 3) identifying critical structural features including brown globules, abrupt border cutoff, granularity, regression, and pigment asymmetry with high accuracy 4) developing a clinical interface for acquisition of images within the clinic 5) developing a web-tool for interactive analysis of images. Key features of the research include dermatopathology confirmation of specific structures and the use of relative color analysis. If successful, specific algorithms would be shared with the growing number of dermatologists using digital dermoscopy. In Phase II, further testing of the algorithms and development of a fast interface would be undertaken. A commercial package combining the software components would be made available for a popular combination digital camera-light head.