Widely used for medical analysis, the texture of the human scar tissue is characterized by irregular and extensive types. The quantitative detection and analysis of the scar texture as enabled by image analysis technology is of great significance to clinical practice. However, the existing methods remain disadvantaged by various shortcomings, such as the inability to fully extract the features of texture. Hence, the integration of second harmonic generation (SHG) imaging and deep learning algorithm is proposed in this study. Through combination with Tamura texture features, a regression model of the scar texture can be constructed to develop a novel method of computer-aided diagnosis, which can assist clinical diagnosis. Based on wavelet packet transform (WPT) and generative adversarial network (GAN), the model is trained with scar texture images of different ages. Generalized Boosted Regression Trees (GBRT) is also adopted to perform regression analysis. Then, the extracted features are further used to predict the age of scar. The experimental results obtained by our proposed model are better compared to the previously published methods. It thus contributes to the better understanding of the mechanism behind scar development and possibly the further development of SHG for skin analysis and clinic practice.

This work developed a phase congruency algorithm combined with texture analysis to quantitatively characterize collagen morphology in second-harmonic generation (SHG) images from human scars. The extracted phase and texture parameters of the SHG images quantified collagen directionality, homogeneity, and coarseness in scars and varied with scar duration. Phase parameters showed an increasing tendency of the mean of phase congruency with scar duration, indicating that collagen fibers are better oriented over time. Texture parameters calculated from local difference local binary pattern (LD-LBP) and Haar wavelet transform, demonstrated that the LD-LBP variance decreased and the energy of all subimages increased with scar duration. It implied that collagen has a more regular pattern and becomes coarser with scar duration. In addition, the random forest regression was used to predict scar duration, demonstrating reliable performance of the extracted phase and texture parameters in characterizing collagen morphology in scar SHG images. Results indicate that the extracted parameters using the proposed method can be used as quantitative indicators to monitor scar progression with time and can help understand the mechanism of scar progression.

Ischemic stroke is a leading cause of death and permanent disability worldwide. Middle cerebral artery occlusion (MCAO) of variable duration times could be anticipated to result in varying degrees of injury that evolve spatially over time. Therefore, investigations following strokes require information concerning the spatiotemporal dimensions of the ischemic core as well as of perilesional areas. In the present study, multiphoton microscopy (MPM) based on two-photon excited fluorescence (TPEF) and second harmonic generation (SHG) was applied to image such pathophysiological events. The ischemic time-points for evaluation were set at 6, 24, 48, and 72 hours after MCAO. Our results demonstrated that MPM has the ability to not only identify the normal and ischemic brain regions, but also reveal morphological changes of the cortex and striatum at various times following permanent MCAO. These findings corresponded well with the hematoxylin and eosin (H&E) stained tissue images. With the technologic progression of miniaturized imaging devices, MPM can be developed into an effective diagnostic and monitoring tool for ischemic stroke.