Towards the analysis of coral skeletal density-banding using deep learning

X-ray micro–computed tomography (µCT) is increasingly used to record the skeletal growth banding of corals. However, the wealth of data generated is time consuming to analyse for growth rates and colony age. Here we test an artificial intelligence (AI) approach to assist the expert identification of annual density boundaries in small colonies of massive Porites spanning decades. A convolutional neural network (CNN) was trained with µCT images combined with manually labelled ground truths to learn banding-related features. The CNN successfully predicted the position of density boundaries in independent images not used in training. Linear extension rates derived from CNN-based outputs and the traditional method were consistent. In the future, well-resolved 2D density boundaries from AI can be used to reconstruct density surfaces and enable studies focused on variations in rugosity and growth gradients across colony 3D space. We recommend the development of a community platform to share annotated images for AI.

Skeletal Growth Rates in Porites lutea Corals from Pulau Tinggi, Malaysia

Skeletal records of massive Porites lutea corals sampled from reefs around Malaysia have previously shown average decadal declines in growth rates associated with sea warming. However, there was a variability in growth declines between sites that warrant the need for investigations into more site-specific variations. This study analyzed decade-long (December 2004–November 2014) annual growth records (annual linear extension rate, skeletal bulk density, calcification rate) reconstructed from five massive P. lutea colonies from Pulau Tinggi, Malaysia. Significant non-linear changes in inter-annual trends of linear extension and calcification rates were found, with notable decreases that corresponded to the 2010 El Niño thermal stress episode and a pan-tropical mass coral bleaching event. Coral linear extension and calcification were observed to return to pre-2010 rates by 2012, suggesting the post-stress recovery of P. lutea corals at the study site within 2 years. Although no long-term declines in linear extension and calcification rates were detected, a linear decrease in annual skeletal bulk density by ≈9.5% over the 10-year study period was found. This suggests that although coral calcification rates are retained, the skeletal integrity of P. lutea corals may be compromised with potential implications for the strength of the overall reef carbonate framework. The correlation of coral calcification rates with sea surface temperature also demonstrated site-specific thermal threshold at 29 °C, which is comparable to the regional thermal threshold previously found for the Thai-Malay Peninsula.

Promotion of Kreweras words

Kreweras words are words consisting of n$$\mathrm {A}$$ A ’s, n$$\mathrm {B}$$ B ’s, and n$$\mathrm {C}$$ C ’s in which every prefix has at least as many $$\mathrm {A}$$ A ’s as $$\mathrm {B}$$ B ’s and at least as many $$\mathrm {A}$$ A ’s as $$\mathrm {C}$$ C ’s. Equivalently, a Kreweras word is a linear extension of the poset $$\mathsf{V}\times [n]$$ V × [ n ] . Kreweras words were introduced in 1965 by Kreweras, who gave a remarkable product formula for their enumeration. Subsequently they became a fundamental example in the theory of lattice walks in the quarter plane. We study Schützenberger’s promotion operator on the set of Kreweras words. In particular, we show that 3n applications of promotion on a Kreweras word merely swaps the $$\mathrm {B}$$ B ’s and $$\mathrm {C}$$ C ’s. Doing so, we provide the first answer to a question of Stanley from 2009, asking for posets with ‘good’ behavior under promotion, other than the four families of shapes classified by Haiman in 1992. We also uncover a strikingly simple description of Kreweras words in terms of Kuperberg’s $$\mathfrak {sl}_3$$ sl 3 -webs, and Postnikov’s trip permutation associated with any plabic graph. In this description, Schützenberger’s promotion corresponds to rotation of the web.

Hourly Bridge Freeway Traffic Crash Prediction Using Log Count Data Poisson Time Series INGARCH Model and Its Comparisons With Regression-based Methods

Traffic crash prediction is vital for relevant agencies to take precautionary measures to minimize the economic and social losses from traffic accidents. Currently, the popularity of machine learning, deep learning, and traditional regression-based models in crash predictions eclipsed the use of count data time series models. Count data model has many intrinsic advantages over machine learning based methods in crash analysis. It is an extension of conventional time series regression by extending normal distribution to Poisson and Negative binomial. Meanwhile, covariate variables can get properly incorporated and their influence on dependent variable is well interpreted. This study attempts to compare and examine the performances of the count data time series model with the regression-based models in hourly crash prediction, utilizing traffic crash data from the Sutong Yangtze River Bridge in China. Log linear extension of Poisson distribution integer valued generalized autoregressive conditional heteroscedasticity models (INGARCH), as a type of count data model, is adopted and compared with the zero-inflated Poisson model (ZIP), as well as the cumulative link model for ordinal regression (CLM). The performances of ZIP and log linear extension of INGARCH count data model are similar and superior to the performances of CLM. Results showed that previous traffic accidents influence the crash occurrence in the near future and the employment of count data time series model in hourly crash prediction can appropriately capture this influence, with an average model sensitivity rate of 77.5%.

Linear extension, skeletal density, and calcification rates of the blue coral Heliopora coerulea

The brooding reef-building octocoral Heliopora is widespread on Indo-West Pacific reefs and appears to be relatively resistant to thermal stress, which may enable it to persist locally while scleractinians diminish under Anthropocene conditions. However, basic physiological measurements of “blue corals” are lacking and prevent their inclusion in trait-based studies. We address this by quantifying rates (mean ± SE) of linear extension (0.86 ± 0.05 cm yr−1) and skeletal density (2.01 ± 0.06 g cm−3) to estimate calcification rates (0.87 ± 0.08 g cm−2 yr−1) for the small branching/columnar morphology of Heliopora coerulea. We postulate that H. coerulea may become an increasingly important reef-builder under ocean warming due to its relative resistance to thermal stress and high skeletal density that make colonies less vulnerable to storm damage under ocean acidification. Moreover, Heliopora corals are likely dispersal limited suggesting they may be an underappreciated genus for restoration of stress-tolerant reef-building capacity on degraded reefs.

Colony-Level 3D Photogrammetry Reveals That Total Linear Extension and Initial Growth Do Not Scale With Complex Morphological Growth in the Branching Coral, Acropora cervicornis

The ability to quantify changes in the structural complexity of reefs and individual coral colonies that build them is vital to understanding, managing, and restoring the function of these ecosystems. However, traditional methods for quantifying coral growth in situ fail to accurately quantify the diversity of morphologies observed both among and within species that contribute to topographical complexity. Three-dimensional (3D) photogrammetry has emerged as a powerful tool for the quantification of reefscape complexity but has yet to be broadly adopted for quantifying the growth and morphology of individual coral colonies. Here we debut a high-throughput method for colony-level 3D photogrammetry and apply this technique to explore the relationship between linear extension and other growth metrics in Acropora cervicornis. We fate-tracked 156 individual coral transplants to test whether initial growth can be used to predict subsequent patterns of growth. We generated photographic series of fragments in a restoration nursery immediately before transplanting to natural reef sites and re-photographed coral at 6 months and 1 year post-transplantation. Photosets were used to build 3D models with Agisoft Metashape, which was automated to run on a high-performance computing system using a custom script to serially process models without the need for additional user input. Coral models were phenotyped in MeshLab to obtain measures of total linear extension (TLE), surface area, volume, and volume of interstitial space (i.e., the space between branches). 3D-model based measures of TLE were highly similar to by-hand measurements made in the field (r = 0.98), demonstrating that this method is compatible with established techniques without additional in water effort. However, we identified an allometric relationship between the change in TLE and the volume of interstitial space, indicating that growth in higher order traits is not necessarily a linear function of growth in branch length. Additionally, relationships among growth measures weakened when comparisons were made across time points, implying that the use of early growth to predict future performance is limited. Taken together, results show that 3D photogrammetry is an information rich method for quantifying colony-level growth and its application can help address contemporary questions in coral biology.

Core-CT: A MATLAB application for the quantitative analysis of sediment and coral cores from X-ray computed tomography (CT)

<div> <p>X-ray computed tomography (CT) is a non-destructive imaging technique that provides three-dimensional (3D) visualisation and high-resolution quantitative data in the form of CT numbers. CT numbers are derived as a function of the X-ray energy, effective atomic number and density of the sample. The sensitivity of the CT number to changes in material density allows it to successfully identify facies changes within sediment cores by detecting downcore shifts in sediment properties, and quantify skeletal linear extension rates and the volume of internal voids from biological erosion of coral cores. Here we present two algorithms to analyse CT scan images specific to geoscience research packaged within an open source MATLAB application (Core-CT). The first algorithm facilitates the computation of representative CT numbers from a user-defined region of interest to identify boundaries of density change (e.g. sedimentary facies, laminations, coral growth bands). The second algorithm enables the segmentation of regions with major density contrast (e.g. internal void space or biogenic material) and the geometric measurements of these irregularities. The versatility of Core-CT for geoscience applications is then demonstrated by utilising CT scans from a range of environmental settings comprising both sediment (Lake Huelde, Chile and Kallang River Basin, Singapore) and coral cores (Thuwal region of Red Sea, Saudi Arabia). Analysis of sediment cores show the capabilities of Core-CT to: 1) locate tsunami deposits from lacustrine sediments, 2) provide rapid and detailed measurement of varved sediments, and 3) identify sedimentary facies from an unsplit shallow marine sediment core. Analysis of coral cores allow us to successfully measure skeletal linear extension from annual growth bands, and provide volumetric quantification and 3D visualisation of internal bioerosion. Core-CT is an accessible, multi-use MATLAB based program that is freely available at GitHub  (https://github.com/yuting-yan/Core-CT).</p> </div><p> </p>

Modeling of a Soft-Rigid Gripper Actuated by a Linear-Extension Soft Pneumatic Actuator

Soft robot has been one significant study in recent decades and soft gripper is one of the popular research directions of soft robot. In a static gripping system, excessive gripping force and large deformation are the main reasons for damage of the object during the gripping process. For achieving low-damage gripping to the object in static gripping system, we proposed a soft-rigid gripper actuated by a linear-extension soft pneumatic actuator in this study. The characteristic of the gripper under a no loading state was measured. When the pressure was >70 kPa, there was an approximately linear relation between the pressure and extension length of the soft actuator. To achieve gripping force and fingertip displacement control of the gripper without sensors integrated on the finger, we presented a non-contact sensing method for gripping state estimation. To analyze the gripping force and fingertip displacement, the relationship between the pressure and extension length of the soft actuator in loading state was compared with the relationship under a no-loading state. The experimental results showed that the relative error between the analytical gripping force and the measured gripping force of the gripper was ≤2.1%. The relative error between analytical fingertip displacement and theoretical fingertip displacement of the gripper was ≤7.4%. Furthermore, the low damage gripping to fragile and soft objects in static and dynamic gripping tests showed good performance of the gripper. Overall, the results indicated the potential application of the gripper in pick-and-place operations.