Innovation Day 2024

Capstone Projects 2024

Welcome the BMEG 457 Capstone Design Project suite for 2024. Using the knowledge and skills they have gained during their studies, undergraduate students are tasked with solving real world problems that require immediate solutions. See below for this year’s project presentations.

Quantitative Analysis of Goblet Cells

Project Description:

Our project aims to increase the efficiency of the analysis of goblet cells in histology slides, contributing to the understanding and treatment of diseases such as Inflammatory Bowel Disease (IBD). Goblet cells are usually found in the intestine and their primary function being to secrete mucin; they play a crucial role in gut health.

Accurately identifying and quantifying them from histology slides is a crucial goal in intestinal research. Since, there is currently no dedicated software to identify and isolate goblet cells from histology slides, our team came together to design a novel multi-faceted approach to address this problem and make an efficient way for researchers to count and identify these cells. By replacing manual counting methods, our algorithm streamlines research efforts, saving time and improving accuracy.

Our algorithm is designed to optimize the image segmentation pipeline for goblet cell analysis. The pipeline includes preprocessing steps like normalization and deconvolution, followed by advanced processing through our Watershed algorithm to generate prompts. Subsequently, the processed image is inputted into Meta’s Segment Anything Model (SAM) to provide researchers with the required metrics. We also designed a simple yet functional UI that can help researchers easily navigate the interface and get a cell count whenever they need. Being the first of its kind, this project lays the groundwork for a more robust way to study goblet cells and help advance biomedical research.

Quantitative Analysis of Goblet Cells Team:

Alvin Hendricks, Ankang (Steve) Luo, Rhea Kaul, Sashreek Datta


Dr. Carolina Tropini, Tropini Lab

Design and UI Optimization of M.R. VEIN

Project Description

Our project, “Design and UI Optimization of M.R. VEIN,” addresses the global challenge of difficult intravenous access. Intravenous access is required in over 80% of hospital admissions and up to 40% of patients encounter difficult intravenous access. Patients with hard-to-access vems can face delays in medical procedures, extended hospital stays, higher healthcare cost expenditure, and clinical supply wastage. Our goal is to revolutionize vein detection with a wearable, immersive vein finder for healthcare providers.

Utilizing cutting-edge ML technology, our solution quickly and accurately locates veins, superimposing them in real-time 3D on the skin’s surface. The goal is to create a solution that is cost-effective, user friendly and portable to facilitate mass adoption of vein finding technology in healthcare facilities.

Our team’s role is to create an intuitive UI/UX design that enhances vem visualization, streamlines intravenous procedures, and addresses existing vein detection challenges. Our UI design can be implemented with any AR glasses equipped with the necessary sensors and cameras that the ML model requires.

Design and UI Optimization of M.R. VEIN Team

Gaurab Acharya, Tolu Adegbehingbe , Jasraj Brar, Sebastian Dzikowski, Jivitesh Ganjoo


Rohith Krishnamurthy, Co-founder, VanTech Medical

Developing an Artificial Intelligence Program for Patient-Centric Clinical Trial Matching

Project Description:

The clinical trial industry, essential for medical advancements, is a multi-billion-dollar global market. Yet, it faces a critical challenge: pharmaceutical and tech companies struggle to find and recruit suitable trial participants. Meanwhile, there’s a vast, untapped potential in consumers eager to engage in research for their unmet healthcare needs. However, they often encounter barriers to accessing and understanding these opportunities. Our team has developed the critical infrastructure required to produce a patient-facing tool that makes the world of biomedical research accessible and understandable to everyone. Using novel tools of artificial intelligence, coupled with insights and training provided by world-leading clinicians and scientists, we created MyTrials. AI that matches patients to appropriate clinical trials based on their disease state, and prioritizes matches based on patient preferences and values.

Developing an Artificial Intelligence Program for Patient-Centric Clinical Trial Matching Team:

Kristian Isa, Omar Asaker, Tyler Connelly, Aminul Islam


Zachary Laksman, MD, MSc, FRCPC, Center for Heart Lung Innovation

Miniaturized Hypoxic Cell Chamber

Project Description:

Tumor hypoxia is a phenomenon where tumor cells are deprived of oxygen due to excessive tumor growth and an insufficient access to blood supply. In order to effectively study the tumor environment, it is important to replicate the hypoxic conditions seen in the human body during tumor growth. Researchers often use hypoxia chambers to replicate these conditions; however, modern hypoxia chambers occupy a lot of space, require high volumes of gas, and are unable to monitor all factors that affect and are affected by tumor growth (such as oxygen levels and pH of the environment). These limitations impact our ability to achieve a more comprehensive understanding of the tumor environment and can lead to slower research. Our project partners with Apricell Biotechnologies to create a Miniaturized Hypoxic Cell Chamber that allows for controlled growth and monitoring of tumor cell culture conditions. By developing this system, cancer research can be expedited, facilitating discovery of drugs and development of novel therapies. Long-term, this project can lead to reduced healthcare costs and improved patient outcomes.

Miniaturized Hypoxic Cell Chamber Team:

Charlie Lake, Leon Li, Kaiwen Liu, Emily Flaschner, and Parsa Moheban


Amir Seyfoori, Apricell Biotechnologies

Design of a Robust EMG Pattern-Recognition (EMG-PR) Prosthesis Control Scheme for Everyday Use

Project Description:

Transradial amputees, defined as upper limb amputees from below the elbow, have previously been able to train and use machine learning (ML) algorithms to classify gestures for their bionic arms. This has allowed patients with amputees to contract their forearm muscles and have the prosthetic perform a desired gesture, providing independence and returned function to the user. Further investigations found that many ML model accuracies worsen in performance when used in real world applications when compared to lab results.

This drop in performance is due to the simplification of training data when simulating realistic gestures and object interactions. In most models, training data was collected with the user’s arm placed on a table performing contractions enacting gestures, providing a relatively simple signal. However, in practical applications, the arm can be in various stationary and dynamic positions with varied speeds, which complicates and changes the EMG signal for the ML model, resulting in poor classification accuracy.

Our project aim was to overcome this limitation by training our ML model with realistic data in order to achieve a robust model. Gesture data was collected to closely resemble practical use performance, as in the simulated labs. However, robustness was accounted for by the inclusion of both dynamic and stationary table data. Through these data collection procedures, we acquired our own robust EMG data and achieved a k-NN model that is able to predict and identify 40 unique gestures while accounting for robustness.

Design of a Robust EMG Pattern-Recognition (EMG-PR) Prosthesis Control Scheme for Everyday Use Team:

Sean Liang, Dima Asmaro, Red Petzen, Josh Goguen, Uday Madaan


Fraser Douglas & Calvin Kuo, Human Motion Biomechanics Lab (HuMBL)

Closing the Control Loop – Design of a Tactile Feedback System for Transradial Prosthesis Users

Project Description:

Prosthetic devices are pivotal in enhancing the lives of individuals with limb loss, offering them increased independence and functionality. Building on this foundation, our project introduces a sensory feedback glove tailored for users with transradial prostheses. Equipped with pressure sensors on the fingertips and flex sensors along the fingers, it captures detailed information on hand positioning and grip dynamics. The data collected by these sensors is then translated into vibrotactile feedback, communicated to the user through a vibration motor-equipped armband. This feedback mechanism allows users to adjust their grip with precision, enhancing their ability to interact with their environment.

A key feature of this system is its wireless connectivity to a computer program, enabling visualization of collected data. This allows users to monitor their hand’s movements and the pressure applied in real-time, offering insights into their prosthetic hand’s functionality. Additionally, the program facilitates the customization of the feedback according to individual preferences, ensuring a personalized user experience. This approach not only improves the usability of prosthetic hands by providing a synthetic sense of touch and grip force but also empowers users to tailor the feedback mechanism to optimize their interaction with various objects, improving their daily life quality.

Closing the Control Loop – Design of a Tactile Feedback System for Transradial Prosthesis Users Team:

Alexandra Hoffman, Ben Clay, Daniella Cameron, Mohamed Khaled, Rochelle Chui


Fraser Douglas & Calvin Kuo, Human Motion Biomechanics Lab (HuMBL)

The development of a Full-Field Optical Coherence Tomography Dynamic Microscope

Project Description:

This project aims to address the critical challenges associated with diagnosing and monitoring ocular diseases through the development of a dynamic full field optical coherence tomography (FF-OCT) microscope. Leveraging advanced temporal dynamics optical coherence microscopy technologies, this innovative approach seeks to enhance optical microscope systems, enabling researchers to explore cellular environments with unprecedented detail. By focusing on the intricate processes and molecular interactions within the eye, the project endeavors to improve the early detection, precise diagnosis, and effective management of conditions such as glaucoma, macular degeneration, and diabetic retinopathy. These ocular diseases represent significant clinical challenges, especially for the elderly and individuals with underlying health conditions, leading to vision impairment and substantial human and financial costs. The project’s objective is to mitigate these burdens by introducing a novel imaging device capable of multimodal imaging, combining the structural contrasts of optical coherence tomography (OCT) and optical coherence microscopy (OCM) and is set to revolutionize the diagnosis and treatment of eye-related health issues, contributing to enhanced quality of life and reduced healthcare expenditures.

The development of a Full-Field Optical Coherence Tomography Dynamic Microscope Team:

Umar Ali, Marvin Wu, Anjali Menon, Madhini Vigeswaran, Johnny Zhao


Dr. Myeong Jin Ju, Computational, Opthalmologic Imaging Laboratory (COIL)

Automated Cell and Hydrogel Dispensing Machine for 3D Tumor Model Development

Project Description:

In recent years, the use of 3D culture systems has gained popularity for mimicking the environment found within cancer tumors in the body. This has made growing 3D tumors a key step in discovering anti-cancer drugs. However, current 3D culture systems are labour-intensive, low-throughput, and require accurate dispensing of highly viscous fluids such as hydrogels. Our client Apricell Biotechnology has requested an automated, high-throughput cell and hydrogel dispenser that is compatible with the company’s current 3D culture plates. The team started with a modified 3D printer prototype provided by Apricell Biotechnology which was upgraded to meet the client’s needs. The device is capable of carrying out automated dispensing protocols at a higher throughput and accuracy than manual operation with standard micropipettes. It features 4 nozzles for high-throughput dispensing, a custom dispensing system to accommodate high-viscosity fluids, an enclosure suitable for sterilization, and a user interface for inputting custom procedures. This device improves the workflow for anti-cancer drug screening, leading to higher chances of success in the pursuit of anti-cancer discoveries. The device’s automation means users can spend less time on mindless 3D tumor growing and more time on thoughtful analysis.

Automated Cell and Hydrogel Dispensing Machine for 3D Tumor Model Development Team:

Max Arsenault, Amine Benhammou, Aditya Goshalia, Victor Oyebolu, Angelica Phelan


Dr. Amir Seyfoori, Apricell Biotechnology

Scalable Low-Cost Bioreactor for High Throughput Biology

Project Description:

Cells are highly reactive to environmental conditions, such as temperature, oxygen level, and nutrient availability. Yeast cells grow fast and are widely used in laboratories to study various biological processes. However, their short doubling time of 90 minutes poses an inconvenience on researchers to manually change the media and dilute the cell culture to ensure that there are enough nutrients for cells to remain at a constant and healthy growth rate. These manual steps can unfortunately also introduce noise and hinder the accuracy of experimental results.

Therefore, we designed a bioreactor, a vessel made to culture cells under tightly controlled conditions, to autonomously run over one week and regulate the environment of yeast culture. We implemented an optical density (cell density) feedback loop to feed the cells additional media when the cell density is too high, while cell products are harvested for further analysis. Another pumping system is used for circulation inside the bioreactor vessel so that the nutrients and oxygen dissolved in media can be evenly distributed to yeast cells.

Scalable Low-Cost Bioreactor for High Throughput Biology team:

Alex Chen, Matthew Kim, Kelvin Lee, Charu Pokhrel, Yi Xing


Omar Tariq and Will Cheney, de Boer Lab

Developing Computational Tool for Predicting Gene Expression Activity

Project Description:

Even though the world of gene editing has advanced greatly, there are so many factors in our DNA that can affect gene expression. Our DNA is so sensitive, even one base pair change has the ability to turn off an entire gene? As a result, scientists researching genes are forced into an endless loop of “trial and error” before they get the sequence they want.
What if there was a way to test and visualize what genes would be expressed before stepping foot into a lab? This was the question proposed to our group for our capstone project. Using machine learning models, different research groups were able to predict gene expression given a DNA sequence with high levels of accuracy. One lacking aspect was the interpretability of results and the actual usability of these models. Our goal is to utilize these machine learning models by creating a user friendly application that allows users to input sequences, compare their predicted expression, and make a decision about which sequence works best for their purpose, all while saving them money and time. With our tool, we aim to make these technological advancements more accessible and useful to researchers, saving them time and money, and ensuring that their efforts are spent not on grunt work, but on their actual research.

Developing Computational Tool for Predicting Gene Expression Activity Team:

Ammar Sallam, Janet He, Ryan Choi, Flora Deng


Najmeh Nakpour, PhD, De Boer Lab

Glow-in-the-Dark Surgical Tracker

Project Description:

This project aimed to develop highly visible sutures for surgical purposes. We’ve developed glow-in-the-dark modules infused with strontium aluminate powder that ensure visibility with the naked eye in hopes to allow better outcomes for surgery.

Glow-in-the-Dark Surgical Tracker Team:

Anmol Anoop, Zaid Badawi, Yu-Rui Chou, Ruini Xiong, Eric Zhang


Cameron Van Oort, Occupational Therapist, HealthOne Physiotherapy

Microfluidic Chip for Size-based Sorting of Multicellular Tumor Spheroids (MCTS)

Project Description:

Multicellular tumor spheroids (MCTS) serve as 3-dimensional models for cancer research due to their similarity to tumor architecture and limited drug penetration properties. They are widely used for studying tumor organization and therapeutic responses. However, generating uniformly-sized batches for preclinical drug screening poses challenges. This complexity arises from diverse genetic, epigenetic, and phenotypic cell characteristics, affecting MCTS physiology. Effective technology is needed to separate, collect, and maintain MCTS while sorting them, without relying on specialized growth substances. Microfluidic devices provide many avenues in small-scale research such as providing predictable control of spheroid translocation and maintenance. Computer simulations were used to design and predict particle flow trajectories and devices were fabricated using 3D printing techniques. We present a hybrid channel design that utilizes inertial and dean flow to separate and focus particles of various sizes into defined fluid streams for sorting and downstream applications.

Microfluidic Chip for Size-based Sorting of Multicellular Tumor Spheroids (MCTS) Team:

Kisa Naqvi, Elly Kim, Prateeksha Aggarwal, Kira Li, Stephanie Nguyen


Dr. Govind Kaigala & Dr. Aditya Kashyap, Laboratory of Microtechnologies for Quantitative Biomedicine

A non-inavasive pH and temperature measurement module for monitoring long-term cell culture

Project Description:

Commonly, when taking pH and temperature measurements for cell cultures, you have to directly interface with the liquid in order to capture specific analytes relating to the culture. However, this can often alter the very data you are trying to collect, which can be frustrating to manage, especially in long term cell cultures. As such, at the request of our client, we took it upon ourselves to create a system which could run constantly, and precisely measure both pH and temperature in a non-invasive manner. After trial and error through sensor selection, we finalized our two choices, displayed through the image on the left, with the pH sensor on the left, and the temperature being measured on the right. In the middle image, you can see the green LED used for pH measurements, and the slot which perfectly fits a petri dish for cell cultures. Finally, the image displayed on the right shows the filter we use to remove ambient light measurements from the pH sensor, which allows us to remove the lid from the enclosure, permitting regular gas flow into and out of the petri dish.

A non-inavasive pH and temperature measurement module for monitoring long-term cell culture team:

Kalen Lacroix, Vivian Osiek, Pranav Tendon, Rafaela Zamataro


Dr. Govind Kaigala, Laboratory of Microtechnologies for Quantitative Biomedicine

Mitigating intimate partner violence – development of a neck cuff for animal modeling for non fatal strangulation

Project Description:

Our project is a cuff that will be placed around the neck of a lab rat and inflated to apply non- fatal strangulation force. This design is meant to be a tool used by Dr. Cripton, Dr. Wellington, and the researchers at ICORD to observe the changes in animal behavior after being exposed to non-fatal strangulation. This research will be used to gain insights into brain damage that happens during domestic violence cases as well as for any other situation where non-fatal strangulation occurs (i.e. martial arts). Our device works very similarly to a blood pressure cuff, where the cuff is secured around the neck and a hand pump is squeezed to apply pressure. We have included a quick release valve so that the researchers can quickly release all the pressure in the cuff quickly during experimentation so as to not expose the animal for longer than necessary. We have also included a pressure gauge on a mount so that the applied pressure can be easily read during experiments. This project is being done in conjunction with Dr. Peter Cripton, Dr. Wellington, and their team at ICORD and we are very grateful for the opportunity and their support throughout this project.

Mitigating intimate partner violence – development of a neck cuff for animal modeling for non fatal strangulation team:

Christophe Charbonneau, Angela Li, Aditya Govind Menon, Emilio Spagnolo, Brianna Zhao


Dr. Peter Cripton & Dr. Cheryl Wellington

Design and Development of a Test Rig for Smart Swim Goggle Validation

Project Description:

Swimming is a pass time and means of physical activity enjoyed by many worldwide. Many swimmers are interested in improving their technique and while the sport has not been historically data driven, researchers and companies are attempting to characterize a swimmer’s motion using biomechanical principles. The movement of the head while swimming has a significant effect on performance as well as risk of injury. For this reason, FORM swim developed a novel algorithm for their smart swim goggle that allows the user to track their head movements while swimming in front crawl.

Our task was to develop a test rig that allows FORM to repeatably and accurately test this algorithm. The rig replicates the rotation of the head in the pitch and roll directions to mimic breathing. The design features two servo motors, each responsible for movement in one of the aforementioned directions. The motors are controlled by an Arduino Mega via a user interface where the user can specify a desired angle range, speed, and test duration. This rig is functional within ±135 degrees in the roll direction and from 0 to 90 degrees in the pitch direction, all within 1 degree of accuracy.

Design and Development of a Test Rig for Smart Swim Goggle Validation Team:

Tochukwu Ayadiuno, Josie Berry, Robyn Beyleveldt, Charlene Harasym, Naomi Jung


Erica Buckeridge, Data Scientist, James Fox, Mechanical Designer, FORM Swim

Lunar Shelter

Project Description:

Our project is to design a lunar shelter capable of protecting the PEEKbot Rover from harmful radiation for the Canadian Space Agency (CSA). The lunar environment is completely unprotected from high energy radiation that is emitted by supernovae across the galaxy. Solar flares also frequent the lunar surface causing massive amounts of radiation to be released at once. This radiation is not only harmful to electronics such as those on the PEEKbot Rover, it is also incredibly damaging to human health. As CSA looks forward to the Artemis project, seeking to put humans back on the moon, strategies must be considered that are lightweight, affordable, and protect from the large amounts of radiation on the lunar surface. In our design we are using polyethylene for its high hydrogen content as well as regolith to decrease payload mass to effectively shield against these radiative particles. The regolith is contained in kevlar bags that can be filled by the assembling astronaut. We are accounting for ergonomics and human dexterity with our design by making it easy to assemble. Additionally, the design prioritizes simplicity, to prevent the abrasive lunar dust from degrading moving components.

Lunar Shelter Team:

Chloe Bolongaro, Maharshi Panchal, Clare Sheedy, Matt Rokosh, Danielle Rowlands


Canadian Space Agency

Contact Information

Contact Us

Want to learn more about our Capstone program?
Reach out!

Danielle Walker
Partnerships Manager
School of Biomedical Engineering