Search for: All records

We propose a prognostic machine learning (ML) framework to support the behavioural outcome prediction for cancer survivors. Specifically, our contributions are four-fold: (1) devise a data-driven, clinical domain guided pipeline to select the best set of predictors among cancer treatments, chronic health conditions, and socio-environmental factors to perform behavioural outcome predictions; (2) use the state-of-the-art two-tier ensemble-based technique to select the best set of predictors for the downstream ML regressor constructions; (3) develop a StackNet Regressor Architecture (SRA) algorithm, i.e., an intelligent meta-modeling algorithm, to dynamically and automatically build an optimized multilayer ensemble-based RA from a given set of ML regressors to predict long-term behavioural outcomes; and (4) conduct a preliminarily experimental case study on our existing study data (i.e., 207 cancer survivors who suffered from either Osteogenic Sarcoma, Soft Tissue Sarcomas, or Acute Lymphoblastic Leukemia before the age of 18) collected by our investigators in a public hospital in Hong Kong. In this pilot study, we demonstrate that our approach outperforms the traditional statistical and computation methods, including Linear and non-Linear ML regressors. 

Abstract. Eastern boundary upwelling systems (EBUS) contribute a disproportionatefraction of the global fish catch relative to their size and are especiallysusceptible to global environmental change. Here we present the evolution ofcommunities over 50 d in an in situ mesocosm 6 km offshore of Callao, Peru, andin the nearby unenclosed coastal Pacific Ocean. The communities weremonitored using multi-marker environmental DNA (eDNA) metabarcoding and flowcytometry. DNA extracted from weekly water samples were subjected toamplicon sequencing for four genetic loci: (1) the V1–V2 region of the 16SrRNA gene for photosynthetic eukaryotes (via their chloroplasts) andbacteria; (2) the V9 region of the 18S rRNA gene for exploration ofeukaryotes but targeting phytoplankton; (3) cytochrome oxidase I (COI) forexploration of eukaryotic taxa but targeting invertebrates; and (4) the 12SrRNA gene, targeting vertebrates. The multi-marker approach showed adivergence of communities (from microbes to fish) between the mesocosm andthe unenclosed ocean. Together with the environmental information, thegenetic data furthered our mechanistic understanding of the processes thatare shaping EBUS communities in a changing ocean. The unenclosed oceanexperienced significant variability over the course of the 50 d experiment,with temporal shifts in community composition, but remained dominated byorganisms that are characteristic of high-nutrient upwelling conditions(e.g., diatoms, copepods, anchovies). A large directional change was found inthe mesocosm community. The mesocosm community that developed wascharacteristic of upwelling regions when upwelling relaxes and watersstratify (e.g., dinoflagellates, nanoflagellates). The selection ofdinoflagellates under the salinity-driven experimentally stratifiedconditions in the mesocosm, as well as the warm conditions brought about bythe coastal El Niño, may be an indication of how EBUS will respond underthe global environmental changes (i.e., increases in surface temperature andfreshwater input, leading to increased stratification) forecast by the IPCC. 

Dense, cold waters formed on Antarctic continental shelves descend along the Antarctic continental margin, where they mix with other Southern Ocean waters to form Antarctic Bottom Water (AABW). AABW then spreads into the deepest parts of all major ocean basins, isolating heat and carbon from the atmosphere for centuries. Despite AABW’s key role in regulating Earth’s climate on long time scales and in recording Southern Ocean conditions, AABW remains poorly observed. This lack of observational data is mostly due to two factors. First, AABW originates on the Antarctic continental shelf and slope wherein situmeasurements are limited and ocean observations by satellites are hampered by persistent sea ice cover and long periods of darkness in winter. Second, north of the Antarctic continental slope, AABW is found below approximately 2 km depth, wherein situobservations are also scarce and satellites cannot provide direct measurements. Here, we review progress made during the past decades in observing AABW. We describe 1) long-term monitoring obtained by moorings, by ship-based surveys, and beneath ice shelves through bore holes; 2) the recent development of autonomous observing tools in coastal Antarctic and deep ocean systems; and 3) alternative approaches including data assimilation models and satellite-derived proxies. The variety of approaches is beginning to transform our understanding of AABW, including its formation processes, temporal variability, and contribution to the lower limb of the global ocean meridional overturning circulation. In particular, these observations highlight the key role played by winds, sea ice, and the Antarctic Ice Sheet in AABW-related processes. We conclude by discussing future avenues for observing and understanding AABW, impressing the need for a sustained and coordinated observing system. 

Osteoarthritis (OA) is a degenerative joint disease and a leading cause of disability globally. In OA, the articulating surface of cartilage is compromised by fissures and cracks, and sometimes even worn away completely. Due to its avascular nature, articular cartilage has a poor self-healing ability, and therefore, understanding the mechanisms underlying degradation is key for OA prevention and for optimal design of replacements. In this work, the articulating surface of bovine cartilage was investigated in an environment with enhanced calcium concentration -as often found in cartilage in relation to OA- by combining atomic force microscopy, spectroscopy and an extended surface forces apparatus for the first time. The experimental results reveal that increased calcium concentration irreversibly weakens the cartilage's surface layer, and promotes stiction and high friction. The synergistic effect of calcium on altering the cartilage surface's structural, mechanical and frictional properties is proposed to compromise cartilage integrity at the onset of OA. Furthermore, two mechanisms at the molecular level based on the influence of calcium on lubricin and on the aggregation of the cartilage's matrix, respectively, are identified. The results of this work might not only help prevent OA but also help design better cartilage replacements. 

Continental slopes – steep regions between the shelf break and abyssal ocean – play key roles in the climatology and ecology of the Arctic Ocean. Here, through review and synthesis, we find that the narrow slope regions contribute to ecosystem functioning disproportionately to the size of the habitat area (∼6% of total Arctic Ocean area). Driven by inflows of sub-Arctic waters and steered by topography, boundary currents transport boreal properties and particle loads from the Atlantic and Pacific Oceans along-slope, thus creating both along and cross-slope connectivity gradients in water mass properties and biomass. Drainage of dense, saline shelf water and material within these, and contributions of river and meltwater also shape the characteristics of the slope domain. These and other properties led us to distinguish upper and lower slope domains; the upper slope (shelf break to ∼800 m) is characterized by stronger currents, warmer sub-surface temperatures, and higher biomass across several trophic levels (especially near inflow areas). In contrast, the lower slope has slower-moving currents, is cooler, and exhibits lower vertical carbon flux and biomass. Distinct zonation of zooplankton, benthic and fish communities result from these differences. Slopes display varying levels of system connectivity: (1) along-slope through property and material transport in boundary currents, (2) cross-slope through upwelling of warm and nutrient rich water and down-welling of dense water and organic rich matter, and (3) vertically through shear and mixing. Slope dynamics also generate separating functions through (1) along-slope and across-slope fronts concentrating biological activity, and (2) vertical gradients in the water column and at the seafloor that maintain distinct physical structure and community turnover. At the upper slope, climatic change is manifested in sea-ice retreat, increased heat and mass transport by sub-Arctic inflows, surface warming, and altered vertical stratification, while the lower slope has yet to display evidence of change. Model projections suggest that ongoing physical changes will enhance primary production at the upper slope, with suspected enhancing effects for consumers. We recommend Pan-Arctic monitoring efforts of slopes given that many signals of climate change appear there first and are then transmitted along the slope domain.