Glazer Alex M, Farberg Aaron S, Winkelmann Richard R, Tucker Natalie, Rigel Darrell S
Department of Medicine, Division of Dermatology, University of Arizona College of Medicine, Tucson, Arizona.
Dermatol Online J. 2017 Aug 15;23(8):13030/qt9xx2907s.
Background/Study Aim: A multispectral digital skin lesion analysis (MSDSLA) device has proven to be sensitive and specific for malignant melanoma (MM) detection by dermatologists and may have other useful applications. This study aimed to develop and test objective quantitative Asymmetry, Border irregularity, Color, and Diameter (qABCD) parameters for MSDSLA and correlate them with the presence of clinical ABCD features to aid the decision to biopsy a suspicious pigmented skin lesions (PSLs).
1632 benign and malignant [175 MM/High Grade Dysplastic Nevi (HGDN)] were evaluated for their qABCD parameters. Quantitative characteristics were correlated with the presence of clinical ABCD features identified by independent dermatologists.
qA, qB, qC, and qD had correlations of 78%, 73%, 76%, and 86%, respectively, for non-MM/HGDN lesions. The correlations for qA, qB, qC, and qD for MM/HGDN lesions was 86.3%, 83%, 89%, and 89%, respectively. All repeatability parameters were statistically significant.
This study demonstrates qABCD values are repeatable and strongly correlate with the clinical ABCD features. qABCD characteristics provide an additional objective and reliable means of identifying PSLs that need further evaluation to rule out MM and, in combination with the clinical ABCDs, may allow for improved assessment when evaluating the malignant potential of PSLs.
背景/研究目的：多光谱数字皮肤病变分析（MSDSLA）设备已被证明对皮肤科医生检测恶性黑色素瘤（MM）具有敏感性和特异性，并且可能还有其他有用的应用。本研究旨在开发并测试MSDSLA的客观定量不对称性、边界不规则性、颜色和直径（qABCD）参数，并将它们与临床ABCD特征的存在情况相关联，以辅助决定对可疑色素沉着性皮肤病变（PSL）进行活检。
对1632例良性和恶性[175例MM/高级别发育异常痣（HGDN）]进行qABCD参数评估。将定量特征与独立皮肤科医生确定的临床ABCD特征的存在情况相关联。
对于非MM/HGDN病变，qA、qB、qC和qD的相关性分别为78%、73%、76%和86%。对于MM/HGDN病变，qA、qB、qC和qD的相关性分别为86.3%、83%、89%和89%。所有重复性参数均具有统计学意义。
本研究表明qABCD值具有可重复性，并且与临床ABCD特征密切相关。qABCD特征提供了一种额外的客观且可靠的方法来识别需要进一步评估以排除MM的PSL，并且与临床ABCD特征相结合，在评估PSL的恶性潜能时可能会改进评估。
