Computer-Aided Detection (CADe) System with Optical Coherent Tomography for Melanin Morphology Quantification in Melasma Patients

Abstract

Dark skin-type individuals have a greater tendency to have pigmentary disorders, among which melasma is especially refractory to treat and often recurs. Objective measurement of melanin amount helps evaluate the treatment response of pigmentary disorders. However, naked-eye evaluation is subjective to weariness and bias. We used a cellular resolution full-field optical coherence tomography (FF-OCT) to assess melanin features of melasma lesions and perilesional skin on the cheeks of eight Asian patients. A computer-aided detection (CADe) system is proposed to mark and quantify melanin. This system combines spatial compounding-based denoising convolutional neural networks (SC-DnCNN), and through image processing techniques, various types of melanin features, including area, distribution, intensity, and shape, can be extracted. Through evaluations of the image differences between the lesion and perilesional skin, a distribution-based feature of confetti melanin without layering, two distribution-based features of confetti melanin in stratum spinosum, and a distribution-based feature of grain melanin at the dermal-epidermal junction, statistically significant findings were achieved (p-values = 0.0402, 0.0032, 0.0312, and 0.0426, respectively). FF-OCT enables the real-time observation of melanin features, and the CADe system with SC-DnCNN was a precise and objective tool with which to interpret the area, distribution, intensity, and shape of melanin on FF-OCT images.

Keywords: cellular resolution; computer-aided detection; convolutional neural networks; deep learning; full-field OCT; image denoising; melanin; melasma; optical coherence tomography; photoaging.

Figure 1

Block diagram of the proposed computer-aided detection (CADe) system.

Figure 2

The deep learning architecture of the denoising convolutional neural network (DnCNN).

Figure 3

Schematic diagram of how to generate low-speckle ground truth images.

Figure 4

The structure of the spatial compounding-based denoising convolutional neural networks (SC-DnCNN) trained for optical coherence tomography (OCT) denoising.

Figure 5

Illustration of automatic melanin segmentation in CADe system. (a) Clinical image of the imaged melasma lesions (rectangular) on the cheek. (b) The original en face scan (E-scan) image. (c) The image after SC-DnCNN. (d) The image after performing contrast-limited adaptive histogram equalization (CLAHE). (e) The candidate targets segmented by thresholding. (f) The selected grain melanin after image opening. (g) The selected confetti melanin after morphological operations. The field of view is 475 × 476 µm.

Figure 6

The performance comparison of CADe with and without SC-DnCNN on the representative lesion images. (a,b) are the input image and the denoised image, respectively, while (c,d) are the output results when superimposing the detected melanin on those images (grain melanin is represented in red, and confetti melanin is represented in yellow), respectively. The field of view is 475 × 476 µm.

Figure 7

The performance comparison of CADe with and without SC-DnCNN on the perilesional skin images. (a,b) are the input image and the denoised image, respectively, while (c,d) are the output results when superimposing the detected melanin on those images (grain melanin is represented in red, and confetti melanin is represented in yellow), respectively. The field of view is 475 × 476 µm.

Figure 8

Comparison of melanin appearance between the representative lesion and perilesional skin images in different skin layers displayed by the CADe system. (a–f) are the perilesional skin images and the lesion images in the stratum spinosum, dermal–epidermal junction (DEJ), and papillary dermis, respectively. The field of view is 475 × 476 µm.