Fast Two-photon Microscopy by Neuroimaging with Oblong Random Acquisition (NORA)

Advances in neural imaging have enabled neuroscience to study how the joint activity of large neural populations conspire to produce perception, behavior and cognition. Despite many advances in optical methods, there exists a fundamental tradeoff between imaging speed, field of view, and resolution that limits the scope of neural imaging, especially for the raster-scanning multi-photon imaging needed for imaging deeper into the brain. One approach to overcoming this trade-off is in computational imaging: the co-development of optics and algorithms where the optics are designed to encode the target images into fewer measurements that are faster to acquire, and the algorithms compensate by inverting the optical image coding process to recover a larger or higher resolution image. We present here one such approach for raster-scanning two-photon imaging: Neuroimaging with Oblong Random Acquisition (NORA). NORA quickly acquires each frame in a microscopic video by subsampling only a fraction of the fast scanning lines, ignoring large portions of each frame. NORA mitigates the information loss by extending the point-spread function in the slow-scan direction to integrate the fluorescence of neighboring lines into a single set of measurements. By imaging different, randomly selected, lines at each frame, NORA diversifies the information collected across frames and enables video-level reconstruction. Rather than reconstruct the video frame-by-frame using image-level recovery, NORA recovers full video sequences through a nuclear-norm minimization (i.e., matrix completion) on the pixels-by-time matrix. We simulated NORA imaging using the Neural Anatomy and Optical Microscopy (NAOMi) biophysical simulation suite. Using these simulations we demonstrate that NORA imaging can accurately recover 400 um X 400 um fields of view at subsampling rates up to 20X, despite realistic noise and motion conditions.

Light-weight Retinal Layer Segmentation with Global Reasoning

Automatic retinal layer segmentation with medical images, such as optical coherence tomography (OCT) images, serves as an important tool for diagnosing ophthalmic diseases. However, it is challenging to achieve accurate segmentation due to low contrast and blood flow noises presented in the images. In addition, the algorithm should be light-weight to be deployed for practical clinical applications. Therefore, it is desired to design a light-weight network with high performance for retinal layer segmentation. In this paper, we propose LightReSeg for retinal layer segmentation which can be applied to OCT images. Specifically, our approach follows an encoder-decoder structure, where the encoder part employs multi-scale feature extraction and a Transformer block for fully exploiting the semantic information of feature maps at all scales and making the features have better global reasoning capabilities, while the decoder part, we design a multi-scale asymmetric attention (MAA) module for preserving the semantic information at each encoder scale. The experiments show that our approach achieves a better segmentation performance compared to the current state-of-the-art method TransUnet with 105.7M parameters on both our collected dataset and two other public datasets, with only 3.3M parameters.

Robust semi-automatic vessel tracing in the human retinal image by an instance segmentation neural network

The morphology and hierarchy of the vascular systems are essential for perfusion in supporting metabolism. In human retina, one of the most energy-demanding organs, retinal circulation nourishes the entire inner retina by an intricate vasculature emerging and remerging at the optic nerve head (ONH). Thus, tracing the vascular branching from ONH through the vascular tree can illustrate vascular hierarchy and allow detailed morphological quantification, and yet remains a challenging task. Here, we presented a novel approach for a robust semi-automatic vessel tracing algorithm on human fundus images by an instance segmentation neural network (InSegNN). Distinct from semantic segmentation, InSegNN separates and labels different vascular trees individually and therefore enable tracing each tree throughout its branching. We have built-in three strategies to improve robustness and accuracy with temporal learning, spatial multi-sampling, and dynamic probability map. We achieved 83% specificity, and 50% improvement in Symmetric Best Dice (SBD) compared to literature, and outperformed baseline U-net. We have demonstrated tracing individual vessel trees from fundus images, and simultaneously retain the vessel hierarchy information. InSegNN paves a way for any subsequent morphological analysis of vascular morphology in relation to retinal diseases.