Deep-learning-based 3D blood flow reconstruction in transmissive laser speckle imaging

Breaking through scattering: The H-Net CNN model for image retrieval

BACKGROUND

In scattering media, traditional optical imaging techniques often find it significantly challenging to accurately reconstruct images owing to rapid light scattering. Thus, to address this problem, we propose a convolutional neural network architecture called H-Net, which is specifically designed to recover structural information from images distorted by scattering media.

METHOD

Our approach involves the use of dilated convolutions to capture local and global features of the distorted images, allowing for the effective reconstruction of the underlying structures. First, we developed a diffuse image dataset by projecting handwritten numbers through diffusers with different thicknesses, capturing the resulting distorted images. Second, we generated a synthetic speckle images dataset, composed of simulated speckle patterns. These datasets were designed to train the model to recover structures within scattering media. To evaluate the model's performance, we calculated the Structural Similarity Measure Index between the model's predictions and the original images on unseen data.

RESULT

This proposed architecture achieves reconstructions with an average structural similarity index measure of 0.8 while maintaining low computational costs.

CONCLUSION

The results of this study indicate that H-Net offers an alternative to more complex and computationally expensive models, providing efficient and reliable image reconstruction in scattering media.

Reconstructing 3D De-Blurred Structures from Limited Angles of View through Turbid Media Using Deep Learning

Recent studies in transillumination imaging for developing an optical computed tomography device for small animal and human body parts have used deep learning networks to suppress the scattering effect, estimate depth information of light-absorbing structures, and reconstruct three-dimensional images of de-blurred structures. However, they still have limitations, such as knowing the information of the structure in advance, only processing simple structures, limited effectiveness for structures with a depth of about 15 mm, and the need to use separated deep learning networks for de-blurring and estimating information. Furthermore, the current technique cannot handle multiple structures distributed at different depths next to each other in the same image. To overcome the mentioned limitations in transillumination imaging, this study proposed a pixel-by-pixel scanning technique in combination with deep learning networks (Attention Res-UNet for scattering suppression and DenseNet-169 for depth estimation) to estimate the existence of each pixel and the relative structural depth information. The efficacy of the proposed method was evaluated through experiments that involved a complex model within a tissue-equivalent phantom and a mouse, achieving a reconstruction error of 2.18% compared to the dimensions of the ground truth when using the fully convolutional network. Furthermore, we could use the depth matrix obtained from the convolutional neural network (DenseNet-169) to reconstruct the absorbing structures using a binary thresholding method, which produced a reconstruction error of 6.82%. Therefore, only one convolutional neural network (DenseNet-169) must be used for depth estimation and explicit image reconstruction. Therefore, it reduces time and computational resources. With depth information at each pixel, reconstruction of 3D image of the de-blurred structures could be performed even from a single blurred image. These results confirm the feasibility and robustness of the proposed pixel-by-pixel scanning technique to restore the internal structure of the body, including intricate networks such as blood vessels or abnormal tissues.