Innovative Projects Realized

Effect of neuromuscular electrical stimulation using the StimaWELL 120MTRS system on multifidus muscle morphology and function in patients with chronic low back pain

Chronic low-back pain (CLBP) is a significant health issue in North America. CLBP patients demonstrate structural changes to and reduced function of the multifidus muscle, which is important to low-back health. Neuromuscular electrical stimulation (NMES) is a treatment used to improve pain and function in low-back pain patients, but is frequently painful. Our partner has developed a device (the StimaWELL 120MTRS Stimulation Mat) that is less painful for patients, but its effects on multifidus structure and function are unknown. Therefore, our research goal is to determine whether this device can improve the structure and function of the multifidus in CLBP patients. Our partner will benefit by having access to the ultrasound and MRI measurements that we will provide, and by having the efficacy of his device evaluated in a clinical trial.

Retail Supply Chain Predictive Analytics

The project aims predict the demand of customers for small and medium size businesses. Forecasting models will be developed analyze historical data to understand patterns and correlations. Machine learning will be applied to determine how the accuracy can be improved over existing statistical methods, such as Fourier Regression Analysis which is commonly used in retail demand chain management. The demand forecasting model will examine customer behavior and the context surrounding that behavior, including upcoming holidays, the weather, or a recent event such as COVID-19. The key benefit of the project is to help business better navigate many challenges due to demand uncertainty. In particular, it will support businesses to develop effective strategies to improve the management of resources.

Enhancing Reach and Penetration of Global Potato Industry Using Multi Channel Data Driven Strategies

Food Innovation Online Corp (FIO) is a New Brunswick(NB) based company that operates a website targeting the global potato industry. This website – PotatoPro.com – is the #1 website in the potato sector worldwide. As a result, the company has access to a large amount of data on the interests and information needs of this sector.

To bring this to the next level, the company has decided to move to a more formal data-driven decision model. To do so, data from various sources will be brought together in a single database and be analyzed using Business Intelligence software and used for strategic decisions as well as for personalization and marketing automation.

PotatoPro has always supported various in person events in the potato industry. The current COVID-19 crisis has brought these in person contacts to an abrupt halt, causing a sudden increase in the need for online marketing of products in the potato sector. FIO tries to use its strong position in the sector to fill this demand. To benefit from the current situation, the company has launched its new website early and is accelerating development of analytical and marketing capabilities.

Testing the biofiltration capacity of a hydroponic green wall system for airborne VOCs

Air pollution is linked to 7 million deaths worldwide, making air quality one of the top ten global causes of death. Eighty percent of Canadians live in cities, and growing evidence suggests that indoor air within the built environment has significant health impacts. Our research will evaluate the contaminant removal capabilities of an active ‘living wall’ biofilter designed by a Canadian company for residential use. We will perform a sequence of experiments including short- and longer- term experiments to determine the differences between inflow and outflow air that has passed through the hydroponic living wall after being exposed to representative levels of volatile organic contaminants in indoor air. New Earth Solutions will receive results on the baseline performance of their unit as well as the relative contributions of the unit’s component parts (e.g., contaminant removal capacity of the porous synthetic substrate; the fertilized hydroponic irrigation water, plants, etc.).

An AI-based climate impact assessment framework for infrastructure in northern Canada

The thawing of permafrost, due to climate change or alterations in the ground surface energy balance, poses significant threats to infrastructure and communities in northern Canada. A main step towards the design of resilient infrastructure is to assess the climate threat (exposure) and predict the response of the infrastructure (vulnerability). The stability of permafrost is correlated to the changes in the ground surface temperatures. However, the high-quality projections of surface temperature — an important entity in engineering simulations — are not often available.
In this project, an AI-based geomechanical framework will be developed to predict and assess the integrity of northern infrastructure affected by permafrost degradation. Moreover, life cycle analyses will be performed to compare the advantages and disadvantages of popular mitigation solutions against permafrost thaw.
The results will be processed into several forms, such as standard datasets, geospatial information layers, interactive maps, and an on-demand web service API, to be used by a wide range of stakeholders in various applications, including engineering design and maintenance, agriculture, hydrology, policymaking, and risk management.

Driver motion prediction using behaviour classifications of vehicles

A significant portion of decision making, path planning and navigation algorithms for Autonomous Vehicles (AV) rely heavily on accurate estimation of the current location as well as future trajectories of the surrounding road users. There are different kinds of drivers in urban environments, and an expert human driver will identify dangerous drivers and avoid them accordingly. However, existing autonomous driving systems often treat all neighboring vehicles the same and do not take actions to avoid the dangerous drivers.
For active safety and reduced reaction times, Gatik’s AVs need to accurately predict the behaviours of surrounding agents to be able to make safe & reliable complex decisions such as merging, unprotected left turns, lane change,
etc
The goal of this research project is to develop new techniques for enabling accurate & reliable driver behaviour
prediction to ensure safer reactions in avoiding dangerous neighboring drivers, pedestrians and cyclists, and
efficient navigation around careful drivers.

Automatic Annotation of Vertebral Heart Score and Tibial Plateau Angle in X-ray Images

The proposed work combines artificial intelligence and diagnostic medicine. Using X-ray images, radiologists and veterinarians can perform an array of measurements to assess patient health. In a veterinary setting, standard measurements and annotations are performed on X-ray images to assess heart and knee health in canines, namely vertebral heart score (VHS) and tibial plateau angle (TPA). Provided a database of related radiographic images, the chosen intern can develop a method to automatically place annotations and perform these measurements through applications of machine learning. The methods developed in this research will be immediately applicable to the partner organization; these tools will be integrated within iMi’s x-ray imaging system for use in veterinary clinics. Additionally, the methods developed will allow iMi to quickly implement new auto-annotation modalities in human and veterinary clinic settings.

AI Powered Adaptive Assessment

This project focuses on the problem of time and the human resources involved in conducting a secure offline/online pre-recruitment or training assessment. We are proposing an online secured platform powered by Artificial Intelligence and Machine Learning for safe and smooth online assessments, boasting of an advanced reporting and automated process to save time and ensure the right decisions are made regarding employee performance.

Next Generation candidate screening and assessment platform featuring psychological profiling though gamification

This project is the first step to providing Thinking North’s Purple Squirrel recruitment platform to go beyond traditional matching with a novel, data-backed holistic candidate matching process. To provide a robust system, Thinking North is collaborating with Seneca’s School of Software Design and Data Science to use advanced artificial intelligence and gamification techniques to combine “psychology” and “gamification,” known as “psychification,” to enhance the screening aspect for a recruitment process. Psychification builds on the data at Thinking North to create a gamified, interactive way to assess how motivated candidates are for open positions within select technology industries. This psychification and screening platform will benefit companies that are challenged by the need for speed and accuracy in recruiting. Specifically, we are addressing the emerging gig economy where staffing for projects and shorter commitments calls for an even more effective process than we have seen before.

Nature Based Outdoor Recreation ROI Framework

The project will develop a conceptual framework to quantify the value of nature based recreation. The tool will focus on improvements in physical health, improvements in mental health and health benefits associated with improved ecosystem protection. The tool will provide provincial, municipal and community organizations a mechanism to support informed program, policy and planning decisions and will help users better understand and communicate the value of investments in nature based outdoor recreation.

Identification of high-frequency periodic acoustic fish tags with deep learning

Innovasea produces fish tags and receivers to track the presence and motion of fish and marine mammals while underwater. Fish tracking (acoustic telemetry) is used by researchers worldwide to determine the abundance and habits of marine life, make decisions about fishing seasons and allowed catches, and help protect marine mammals. Innovasea has developed a novel high-frequency tag technology that is suitable for very small fish and generates more precise trajectories. However, the new smaller fish tags send no explicit identification information so signals from a specific fish tag are isolated from background noise and other fish tags based on the period and/or pattern of the signals. To obtain useful fish tracking trajectories, Innovasea currently applies manual processing which requires expert knowledge.
In this project we will apply advanced deep learning techniques to large manually processed training sets provided by Innovasea to eliminate the manual preprocessing steps. The project is scoped with an initial phase to test feasibility of the concept and subsequent phases for development with an eventual aim of transitioning the best performing prototype system to a fully realized system for filtering Innovasea data.

Automated diagnosis of liver fibrosis and steatosis using deep-learning algorithms applied to conventional liver ultrasound

Non-alcoholic fatty liver disease (NAFLD) is one of the most common liver disorders worldwide. NAFLD could lead to end-stage liver disease and considered as one of the most common causes of liver transplantation. Moreover, a large number of liver transplant recipients develop NAFLD after transplantation due to side effects of the medications that they have to use to keep their new liver healthy. Although liver biopsy has been used for a long time to evaluate the liver condition it is associated with a risk of bleeding and other side effects. Consequently, we aim to improve simple liver ultrasound method by adding some artificial intelligence algorithms to the images obtained from this method to diagnose NAFLD and other liver disorders after transplantation. By doing so, clinicians will be able to perform real-time liver ultrasound without the need to refer the patients to a radiologist for the procedure.