Automatic Annotation of Subsea Pipelines using Deep Learning.

Regulatory requirements for sub-sea oil and gas operators mandates the frequent inspection of pipeline assets to ensure that their degradation and damage are maintained at acceptable levels. The inspection process is usually sub-contracted to surveyors who utilize sub-sea Remotely Operated Vehicles (ROVs), launched from a surface vessel and piloted over the pipeline. ROVs capture data from various sensors/instruments which are subsequently reviewed and interpreted by human operators, creating a log of event annotations; a slow, labor-intensive and costly process. The paper presents an automatic image annotation framework that identifies/classifies key events of interest in the video footage viz. exposure, burial, field joints, anodes, and free spans. The reported methodology utilizes transfer learning with a Deep Convolutional Neural Network (ResNet-50), fine-tuned on real-life, representative data from challenging sub-sea environments with low lighting conditions, sand agitation, sea-life and vegetation. The network outputs are configured to perform multi-label image classifications for critical events. The annotation performance varies between 95.1% and 99.7% in terms of accuracy and 90.4% and 99.4% in terms of F1-Score depending on event type. The performance results are on a per-frame basis and corroborate the potential of the algorithm to be the foundation for an intelligent decision support framework that automates the annotation process. The solution can execute annotations in real-time and is significantly more cost-effective than human-only approaches.

Using process algebra to develop predator-prey models of within-host parasite dynamics.

As a first approximation of immune-mediated within-host parasite dynamics we can consider the immune response as a predator, with the parasite as its prey. In the ecological literature of predator-prey interactions there are a number of different functional responses used to describe how a predator reproduces in response to consuming prey. Until recently most of the models of the immune system that have taken a predator-prey approach have used simple mass action dynamics to capture the interaction between the immune response and the parasite. More recently Fenton and Perkins (2010) employed three of the most commonly used prey-dependent functional response terms from the ecological literature. In this paper we make use of a technique from computing science, process algebra, to develop mathematical models. The novelty of the process algebra approach is to allow stochastic models of the population (parasite and immune cells) to be developed from rules of individual cell behaviour. By using this approach in which individual cellular behaviour is captured we have derived a ratio-dependent response similar to that seen in the previous models of immune-mediated parasite dynamics, confirming that, whilst this type of term is controversial in ecological predator-prey models, it is appropriate for models of the immune system.

A symbolic investigation of superspreaders.

Superspreaders are an important phenomenon in the spread of infectious disease, accounting for a higher than average number of new infections in the population. We use mathematical models to compare the impact of supershedders and supercontacters on population dynamics. The stochastic, individual based models are investigated by conversion to deterministic, population level Mean Field Equations, using process algebra. The mean emergent population dynamics of the models are shown to be equivalent with and without superspreaders; however, simulations confirm expectations of differences in variability, having implications for individual epidemics.

Side effects of the Yuzpe regimen of emergency contraception and two modifications.

We investigated side effects after the standard Yuzpe regimen or two modifications: substituting norethindrone as the progestin or eliminating the second dose. We also examined the impact of taking either dose with food. Nearly two thirds of women reported at least one side effect, the majority of which were mild or moderate. Women in our study experienced more side effects after the second dose than after the first. Taking the first dose within 1 h of a meal or snack was associated with increased nausea and vomiting; taking the second dose within 1 h of a meal or snack was associated with decreased nausea and vomiting. A targeted approach to prophylactic antiemetic use could reduce the number of women given these drugs, and the number who experience unnecessary side effects. The impact of counseling on side effects should be further evaluated.

A rigorous approach to investigating common assumptions about disease transmission: Process algebra as an emerging modelling methodology for epidemiology.

Changing scale, for example, the ability to move seamlessly from an individual-based model to a population-based model, is an important problem in many fields. In this paper, we introduce process algebra as a novel solution to this problem in the context of models of infectious disease spread. Process algebra allows us to describe a system in terms of the stochastic behaviour of individuals, and is a technique from computer science. We review the use of process algebra in biological systems, and the variety of quantitative and qualitative analysis techniques available. The analysis illustrated here solves the changing scale problem: from the individual behaviour we can rigorously derive equations to describe the mean behaviour of the system at the level of the population. The biological problem investigated is the transmission of infection, and how this relates to individual interactions.