Significance: Two-dimensional (2-D) fully convolutional neural networks have been shown capable of producing maps of sO 2 from 2-D simulated images of simple tissue models. However, their potential to produce accurate estimates in vivo is uncertain as they are limited by the 2-D nature of the training data when the problem is inherently three-dimensional (3-D), and they have not been tested with realistic images. Aim: To demonstrate the capability of deep neural networks to process whole 3-D images and output 3-D maps of vascular sO 2 from realistic tissue models/images. Approach: Two separate fully convolutional neural networks were trained to produce 3-D maps of vascular blood oxygen saturation and vessel positions from multiwavelength simulated images of tissue models. Results: The mean of the absolute difference between the true mean vessel sO 2 and the network output for 40 examples was 4.4% and the standard deviation was 4.5%. Conclusions: 3-D fully convolutional networks were shown capable of producing accurate sO 2 maps using the full extent of spatial information contained within 3-D images generated under conditions mimicking real imaging scenarios. We demonstrate that networks can cope with some of the confounding effects present in real images such as limited-view artifacts and have the potential to produce accurate estimates in vivo.

Collagen type I protein is the most abundant protein in mammals. It is a crucial component of the extracellular-matrix where it robustly self-assembles into fibrils of specific striped architectures that are crucial for the correct collagen function. The molecular features that determine such robust fibril architectures are currently not well understood. Here we develop a minimal coarse-grained model to connect the design of collagen-like molecules to the architecture of the resulting self-assembled fibrils. We find that the pattern of charged residues on the surface of molecules can drive the formation of collagen-like fibrils and fully control their architectures. Our findings can help understand changes in collagen architectures observed in diseases and guide the design of synthetic collagen scaffolds..

The linear unmixing technique is an appealing method for estimating blood oxygen saturation (sO2) from multiwavelength photoacoustic tomography images, as estimates can be acquired with a straightforward matrix inversion. However, the technique can only rarely provide accurate estimates in vivo, as it requires that the light fluence at the voxels of interest is constant with wavelength. One way to extend the set of cases where accurate information related to sO2 can be acquired with the technique is by taking the difference in sO2 estimates between vessels. Assuming images are perfectly reconstructed, the intervascular difference in sO2 estimates is accurate if the error in the estimates due to the wavelength dependence of the fluence is identical for both. An in silico study was performed to uncover what kinds of conditions may give rise to accurate sO2 differences for a vessel pair. Basic criteria were formulated in simple tissue models consisting of a pair of vessels immersed in two-layer skin models. To assess whether these criteria might still be valid in more realistic imaging scenarios, the sO2 difference was estimated for vessels in more complex tissue models.

No abstract