Physics Department Honours Projects

PHYS 4907, 4908, and 4909

in conjunction with

University of Ottawa courses PHY 4006 and 4906

Sangeeta Murugkar

Herzberg Room 2414

(613) 520-2600 ext. 4016

smurugkar@physics.carleton.ca

Projects

Introduction

Within a student’s degree requirements, the Honours project is a course, like any other, in the context of a full course load. However, the project provides a unique opportunity, compared to typical courses, to gain experience carrying out supervised research and developing the associated skills and techniques required. Students will gain a detailed understanding the specific project material, but should alsso develop a broader view and knowledge of the broader area of physics in which the project is situated. Typically, students find that the honours project is one of the most interesting and rewarding elements of their undergraduate studies.

Honours projects are a required component for all students in Physics B.Sc. Honours programs. Students in the combined programs with Biology, Chemistry, or Mathematics can choose to complete their project either with the Physics Department or in the alternate discipline. While most students choose to do the [1.0] credit PHYS 4909, there is also the option of one term projects: [0.5] credit PHYS 4907 (Fall) or PHYS 4908 (Winter). The remaining [0.5] credit can be satisfied with a course in PHYS at the 4000-level.

project with an available supervisor. The supervisor should then notify the Course Coordinator and Academic Administrator, Joanne Martin, of the project and student to be supervised.

Course Requirements

It is expected that each student will spend approximately 6 hours per week on project-related activities. The student and supervisor should agree on the timing of regularly scheduled meetings to discuss the details and progress of the project.

A written mid-course report will be due to the student’s supervisor on December 6, 2019 for the PHYS 4909 (or midway through the relevant term for PHYS 4907 or PHYS 4908). It is very useful to use the mid-course report to get an understanding of the expectations for your final report. Take the opportunity of the mid-course report to articulate your plans for the rest of the project. The final report will be due on April 7, 2020 (Dec 6, 2019 for PHYS 4907).

An oral presentation will be scheduled for each student’s Honours project. You will be asked to submit a title and abstract in advance of your presentation. An abstract should be a concise summary of the contents of the paper to be presented. It should put the talk into perspective by giving some motivation and also state the major results. The abstract should fit onto a single page and should contain the title, authors' names, and institutions. Some guidelines for your seminar are given at the end of this document.

The grading for the honours projects is as follows:

  • Mid-course written progress report10%
  • Final written report70%
  • Oral presentation20%

The student’s supervisor is responsible for grading the reports. Oral presentations are graded by faculty members available to attend the presentations.

Suggested Honours Project Topics

Listed below are suggested projects with supervisors in the Carleton Physics Department. Students should contact supervisors offering projects of interest and select a project with an available supervisor. The supervisor should then notify the Course Coordinator and Academic Administrator Joanne Martin of the project and student to be supervised. Projects not listed below can be pursued if a suitable supervisor is available.

In addition, projects can be carried out at the University of Ottawa.In this case, students should provide a summary of the proposed project to the course coordinator before registration. Please consult the project information at U of O pages.

Projects in Medical Physics

Elsayed Ali

Radiation is an effective treatment to cure cancer or to reduce its symptoms. Most of the radiation treatments for cancer are delivered using medical linear accelerators (linacs). In a linac, electrons are accelerated to megavoltage energies in a waveguide using radiofrequency (RF) power. The energetic electrons are steered to hit a metal target at a focal spot to produce a photon beam. The photon beam is controlled to take specific shapes that maximize the radiation delivered to the cancer while minimizing the harm to other normal structures in the patient body. SIMAC (short for simulate linac) is a novel state-of-the-art virtual simulation tool that mimics the operation of a linac. The tool is meant to enable the understanding of the interplay among the various parameters involved in producing clinically useful radiation beams. SIMAC is free for education and research purposes.

The proposed project is to develop one or more modules for the next generation of this educational tool. In this unique opportunity, the student will learn the basics of linac operation, and will focus on modelling specific aspects of the process. The modules that are currently of particular interest to include in SIMAC are: (1) a model of an ionization chamber to dynamically control (or servo) the linac dose rate; (2) a model of a multi-segment ionization chamber to servo the steering of the electron beam; and, (3) a stand-alone model of a magnetron as an RF power source. Other modules can be considered, depending on the student’s interests and strengths.

The project is done in collaboration with the development team of SIMAC at The Princess Margaret Hospital in Toronto. This collaboration should enrich the learning experience of the student. Interested students are encouraged to watch this short video that describes the first version of SIMAC here, as well visit the SIMAC webpage. SIMAC has recently been converted from Matlab into Python, and its upcoming release will be web-based. The project is particularly suitable for a student with a working knowledge of Python, good general programming skills, and a passion for turning the equations or tabulated data that govern physics processes into lines of code that simulate such processes. If you would like to learn more about this opportunity, please contact Dr. Elsayed Ali at elali@toh.ca.

Emily Heath

Patient-specific motion models using cone-beam CT

Respiratory motion is a challenge to the accurate delivery of radiation therapy. Modern radiotherapy equipment comes with on-board x-ray imaging technology which allows 3D and 4D cone-beam (CBCT) imaging of the patient at the time of treatment. The aim of this project is to use previously acquired 4D CBCT datasets of a novel deformable phantom to test the accuracy of the motion models that can be extracted from 4DCBCT compared to the gold standard 4DCT. The student will work with deformable image registration software to create motion models from the image data. If time permits, the student will also investigate accuracy of 4D Monte Carlo simulations of dose delivery to the phantom using these 4DCBCT-based motion models.

Paul Johns

Project 1: (joint supervision with Tong Xu)

In dual-energy radiography, the object is imaged with beams of two different energies. The dependences of the photoelectric and Compton cross sections on atomic number and photon energy allow these data to be sufficient to characterize the object's x-ray transmission properties at all energies used in diagnostic radiology. Applications include beam-hardening artefact suppression in CT and removal of contrast from objects not of diagnostic interest in projection imaging. In this project a dual-energy imaging protocol will be implemented using an x-ray tube and a Perkin-Elmer flat panel image receptor in our lab. Dual-energy performance will be measured by imaging test objects composed of aluminum and different plastics. The project will entail experiment design and operation, and the implementation of a nonlinear software correction for the polyenergetic nature of the x-ray beams, based on the established dual-material basis decomposition approach from the literature.

Project 2: Mapping x-ray tube source distribution with a miniature detector

In x-ray imaging, the x-ray tube source geometry is a key determinant of an imaging system's resolution. The radiation source region ranges in size from 100 microns square to about 2 mm square. Pinhole radiography of the source is the standard measurement method. Traditionally x-ray film was used to record the distribution. Our goal is to configure a modern digital alternative.

In this project the student will work with a digital detector originally designed for dental intra-oral imaging. The detector has 20 micron pixels - perfect for recording the details of the x-ray tube emission pattern. The student will (i) characterize the detector for sensitivity, resolution, and artefacts, and (ii) configure a mounting system to align the detector, pinhole and x-ray tube to make a radiation source inspection camera. Then (iii) the camera will be tested on two general radiology tubes.

Project 3: Display algorithms for x-ray scatter imaging (suitable for 4907 or 4908).

Our lab's focus is to develop an x-ray imaging system that uses low-angle scattered radiation, below 10 degrees, in addition to the conventional primary radiation. It is not obvious how best to display the results. An ensemble of scatter images can be generated in one acquisition, corresponding to scatter at different angles. We also have the conventional primary image. There are several options besides displaying simple stacks of grey-scale images. Previous work in our lab looked at peak location, centroid, encoding the scatter radial profile as an RBG spectrum, and a mosaic display approach. Our moasaic work needs to be extended and other schemes investigated. For example, peak scatter angle could be encoded in colour and total scatter signal as intensity, or two non-overlapping colour scales could be used for primary and scatter simultaneously. This computational project will be a good way to learn the HSV vs RGB colour systems and will be done in Matlab.

Malcolm McEwen

Novel detectors for primary standards dosimetry

Air-filled ionization chambers have been used for more than 50 years as reference detectors in primary and secondary standard dosimetry laboratories. They have high sensitivity and very good stability but are not ideal detectors for all measurement situations due to there relatively large size and reliance on measuring ionization of air. In recent years a number of novel detectors based on different measuring processes have become available, and these offer potential advantages over ion chambers. These include liquid ion chambers, plastic scintillating fibres and diamond detectors. However, they have typically been developed for clinical medical physics, where accuracy requirements are not as stringent as for primary labs. The aim of this project is to investigate these detectors and establish their limits for precision and accuracy in a range of radiation facilities - from low energy (keV) x-rays to high energy (MeV) electron beams - maintained at the Ionizing Radiation Standards group of the NRC. Suitable detectors will then be used to address a number of measurement problems that are a current focus of the IRS group's research.

Sangeeta Murugkar

Raman Spectroscopy and Multimodal Raman Imaging

Raman spectroscopy is a non-invasive optical technique that is based on the inelastic scattering of light by vibrating molecules. Various projects involving applications of Raman spectroscopy and multimodal nonlinear optical imaging are available in the areas of skin cancer detection and treatment; understanding cellular response to ionizing radiation; and for developing applications in radiation microdosimetry.

Rowan Thomson

Radiation transport simulations are broadly used to study many aspects of physics related to radiotherapy treatments for cancer. A variety of projects involving computational and theoretical studies of the interactions of radiation with matter are possible and may be tailored to the student's strengths. Some projects involve use of egs_brachy, a fast Monte Carlo dose calculation for brachytherapy (developed in the Carleton Laboratory for Radiotherapy Physics), to investigate questions in brachytherapy physics (brachytherapy is a type of radiation treatment in which radioactive sources are placed next to or inside a tumour). One project involves an investigation of approaches to model the biological effects of radiation. Other research directions include studies of cell dosimetry using Monte Carlo simulations, with applications to skeletal dosimetry.

Ruth Wilkins

Biological dosimetry is a method of measuring the amount of ionizing radiation received by an individual using biological material. This type of dosimeter is essential when an individual is accidentally exposed and no physical dosimetry is available. Currently, the accepted method of biological dosimetry is the dicentric chromosome assay which involves examining chromosomes for characteristic damage caused by ionizing radiation. This is a very tedious and time consuming assay, requiring weeks to process a sample from one individual. For large scale radiological events where thousands of individuals might be exposed to ionizing radiation, a biological dosimeter is desirable that could analyse many samples in a timely manner. Our laboratory is currently exploring new techniques for biological dosimetry. This project would involve testing an image analysis system for detecting biological damage in DNA.

Tong Xu

Accurate geometric calibration of a linac-mounted kilovoltage x-ray system (joint supervision with Elsayed Ali and Rolf Clackdoyle)

Cone Beam CT (CBCT) scanners is becoming a popular imaging tools in health care. Equipped with big area imaging detectors, CBCT can acquire a full 3D scan within a single rotation. It is widely used in image guided radiation therapy and surgeries. However, in these applications, as the heavy detector and x-ray source rotate around the patient, the gravity will cause the supporting structures of the scanner to deform. This deformation, if unaccounted for, will cause image blurring and artifacts. This study will investigate how a new 3D geometry calibration can improve image quality of CBCT. Furthermore, we will also study the feasibility of quantify the actual deformation using proposed calibration method.

The measurements are carried out at The Ottawa Hospital Cancer Center and the student will get exposure to state-of-the-art clinical system and the field of medical physics. It is a well balanced project that includes experimental design, code development/refinement, computer simulation, image processing and reconstruction.

Glenn Wells

Heart disease is one of the leading causes of death in Canada and the world. Myocardial perfusion imaging (MPI) is an essential tool in evaluating patients with known or suspected heart disease. SPECT is the clinical workhorse for performing MPI. Recent innovations in the design of cardiac SPECT cameras allow dynamic imaging and thus absolute measures of blood flow in the heart. This technique potentially allows more accurate identification of patients with multi-vessel disease, patients whose disease is underestimated in 50% of cases with current methods. Projects include work on the optimization and evaluation of protocols for SPECT blood flow measurement and/or algorithms for dynamic SPECT reconstruction. The student will work with anonymized patient datasets using a combination of commercial and in-house software. Interested students should contact Dr. Glenn Wells at the University of Ottawa Heart Institute for more information.

Projects in Theoretical Particle Physics

Particle physicists at Carleton are concerned with the discovery of the most fundamental laws that govern the behaviour of subatomic matter and its elementary constituents.

Bruce Campbell

Topics in theoretical particle physics: in particular the possibility of physics beyond the Standard Model of the fundamental interactions, and the possibility of new symmetries of nature at higher energy scales.

Stephen Godfrey

Projects in particle physics phenomenology such as hadron spectroscopy and Beyond the Standard Model phenomenology at the LHC and the International Linear e+e- Collider.

Thomas Gregoire

Projects on the phenomenology of beyond the Standard Model physics (Supersymmetry, little Higgs, dark matter, etc).

Heather Logan

Projects on the phenomenology of extended Higgs sectors

Daniel Stolarski

Projects in phenomenology beyond the Standard Model including topics in collider physics, dark matter, and grand unification.

Projects in Experimental Particle Physics

John Armitage

Use of Muons in an automated particle detector system.

The project would have two parts: first, a static detector calibration system will be set up. This will use plastic scintillator read out by wavelength shifting optical fiber, into Silicon photomultipliers (SiPM). Short scintillators will be used to confirm the passage of a muon and the longer detector under test, will then be interrogated to see if it registered the muon. In this way the efficiency of the detector can be determined.

While this is taking place, preparatory work will be going on in the workshop, to make a moveable stand that will translate the short detectors along the length of the detector under test. The second part of the project will be to set up this moveable stand. The motion of the stand will be controlled by a microcontroller (an ARDUINO). A set of instructions in G-Code will be uploaded to the ARDUINO to translate the short scintillators by a given amount, then to pause and take a datapoint. This to be repeated several times until the whole length of the long detector is scanned.

The ATLAS Experiment

Thomas Koffas and Dag Gillberg

The ATLAS group at Carleton is playing a leading role in developing a new generation of detectors that will be installed and used at the ATLAS experiment at the CERN Large Hadron Collider in Geneva, Switzerland. In 2019-2022, the ATLAS Carleton group together with seven other Canadian universities will take part in the construction of a new Inner Tracking detector (ITk) based on state-of-the-art silicon sensor technology. Carleton's role in this project will be to perform careful electrical characterization and quality control of the thin silicon sensors prior to the assembly of the ITk detector modules. The group is also developing a small data acquisition

As part of this project the students will work on the development of a DAQ test system based on field-programmable gate arrays (FPGA), attached to custom boards that hold single ABCstar chips, distribute power and provide communication interfaces. This includes the programming of a Nexys-Video board featuring the latest Artix-7 FPGA from Xilinx. To complete the DAQ chain the students will also have to install the ITSDAQ dedicated control and communications software on a Linux machine, compile and operate it. The successful completion of the above DAQ chain will then allow the programming and transfer of analog and digital test vectors developed on a computer to the ABCstar ASIC and the collection of the response data for further analysis. To this end the students will eventually have to develop C++/ROOT-based analysis software for both rapid online pass/fail decisions and for further offline high level analysis and chip performance studies.

The full work will be carried out with the help of experts from the Physics Department under the supervision of Prof. Thomas Koffas and Dag Gillberg.

Through this work the students will acquire valuable experience with FPGA operations and with basic DAQ techniques typical for state-of-the-art ASICs. They will develop important skills in handling and operating advanced custom electronics. The students will also be exposed to basic programming and communication of scientific instrumentation, in diagnosing and fixing operational problems and in understanding the basic operating properties of a CMOS ASIC. Finally the students will interact and work together with skilled professionals that have long experience in designing and operating laboratory and detector systems and with other members of the larger Carleton ATLAS group.

The EXO Experiment

Razvan Gornea

Noble liquid time projection chambers (TPC) are excellent detectors for rare event searches in neutrino physics. For example, the EXO-200 detector employs a 200 kg liquid xenon (LXe) with a TPC configuration which allows the collection of the electric charge produced by ionizing particles. When energy is deposited in a noble liquid, pairs of electron/hole are formed and can be collected by applying an external electric field. The amount of electrons collected is proportional to the initial energy of the ionizing particle.

During their drift in the TPC, electrons may encounter impurities with a large electron affinity and, therefore, part of the initial electric charge could be lost. When the noble liquid in a TPC is of low purity a large amount of charge is lost and then accurate event energy reconstruction becomes challenging. The average time for the electron to get captured is called the "electron lifetime". Experimentally, it often occurs that a LXe TPC can easily suffer from a short electron lifetime or, equivalently, a large concentration of impurities. On the other hand, the accurate measurement of a very long electron lifetime is challenging when using a compact device.

In this project, the student will design, optimize and build a compact purity monitoring device able to measure both short and long electron lifetimes using a custom pulsed VUV-LED-based electron source previously developed at Carleton. SIMION simulations will be employed to design and optimize a compact drift volume able to recycle the probing electron cloud. Hardware development, construction and testing will also be part of the project. Elements of front-end electronics, computer control and system programing using LabVIEW complete the program. A full year commitment is better suited for this project.

Silicon Photomultipliers (SiPM) Development

Simon Viel

DEAP-3600 is a dark matter detector with a target of liquid argon at SNOLAB, located 2 km underground in Sudbury, Canada. Dark matter would interact with the liquid argon, creating a flash of scintillation light that is detected by photomultiplier tubes.

The goal of this project is to improve the background event model of DEAP-3600. Students will write programs in Python and C++ to analyze real data from DEAP-3600 and detector simulations to quantify the rates of background events, which is essential in order to determine whether we see an excess of scintillation events that could be due to dark matter. These studies will also inform the design of future-generation liquid argon detectors looking for dark matter.

SiPM are a photodetector technology, very useful wherever the detection of individual particles of light is required. In addition to nuclear and particle physics, their use is rapidly growing for a wide variety of applications: medical imaging, security, natural resource exploration, photonics, geomatics, meteorology, seismology, forestry, the automotive industry, and many more. Photomultiplier tubes, currently in use for many of these applications, are being progressively replaced by SiPM, which are generally easier to operate while featuring similar performance.

At Carleton, the optical and electronic properties of SiPM prototypes will be characterized towards applications in astroparticle physics (nEXO and future-generation liquid argon dark matter detection, see above). Students will learn how to operate measurement systems such as vacuum and noble gas systems, electronic readout systems, and control systems, as well as developing data analysis skills using software in C++ and Python.

Advice on Giving Seminars

The Honours project seminar provides a valuable opportunity to develop skill in communicating your scientific work to an audience. The following notes will attempt to define an acceptable seminar and to point out some pitfalls to avoid.

When presenting a paper at a conference, one may have only ten minutes to give the paper and then five minutes to answer questions. Communicating weeks or months of research effectively under these constraints requires careful preparation. A strategy has to be worked out, crucial aspects of the research must be identified and presented in a logically structured way and every word must be made to count. Cohesive organization of your content is most important for the way in which your talk comes across.

Practice your talk in advance with a friend or your supervisor, and get the timing right. In your practice sessions get a friend to describe any distracting mannerisms in your style so that you can consciously attempt to avoid them. Some classic examples are the jingling of keys in pockets and random gesticulations with pointers as well as "ums", "uhs", etc.

Pitfalls

The main object, in giving a seminar, is to communicate the research in an intelligible way. It is not to justify how you spent the last N months of your life or to show how crafty you are. Because we are subjectively involved in our work, it is all too easy to misplace emphasis. You may well have spent the last six months building a technically difficult and elegant (you think) transport system for your fragile samples and are dying to talk about it. Limiting yourself to a five word offhand reference is difficult, but usually, no more is justified. Neither is a seminar a party atmosphere. You are not expected to be entertaining; humorous asides are more often distracting than funny except when used by a master. Assuming that your research project was worth doing, a clear, concise and lucid presentation will keep your audience interested.

Finally, although you will normally be speaking to other physicists, avoid jargon. You cannot assume that they are familiar with your narrow specialization. Restrict yourself to expressions and vocabulary that any physicist understands. Define specialist terms; define symbols; take the time to describe graphical and tabular material.

Organization

Avoid the historical narrative. The order in which you executed the various tasks is usually not appropriate for an intelligible description of your work. It is very disconcerting to hear a speaker describe some component of an apparatus while we are still completely ignorant about its purpose in the research. Like a written article, a talk must be structured along the logical connections inherent to the topic. First, a base must be established and then strata of information can be built, layer by layer, developing into a coherent whole. Make clear what your contribution has been since your talk is given in the context of existing work.

Suggested structure for the presentation

The chair of the session will introduce you and give the title of your presentation. So there is no need to repeat this. Just say “thank you” and proceed with your talk.

Outline

  • Extremely brief--about one minute. Show the main headings of your talk. Try to do a little more than just tell your audience that your talk consists of an Introduction, Theory,... It’s a good idea to reproduce the title of your presentation at the top of this page (Powerpoint slide, Adobe screen, etc.) in case the audience missed it when you were introduced.

Introduction and Roadmap

  • Context and motivation
    • Why, how the work was done; explain the nature of the problem you are interested in and its importance. Since you are probably not the first person to do work in this field, reference to other scientists whose work you are trying to verify, refute, expand upon etc. is in order and will set the stage for your own contribution.
  • Orient the listener
    • Where is the presentation going?
    • Its scope and limits
  • Preview of subtle points and difficulties

Theory/Background

  • Relation to general theory
  • Key concepts, equations
  • Assumptions
  • Approximations

Method (& apparatus if experimental)

  • Reference to standard method if appropriate
  • Your innovations
  • Main features of the apparatus
  • Integration of experimental parameters into theory

Results and Discussion

  • Graphs and tables
  • Comparison to theory, expectations

Conclusion

  • Summarize your conclusions
  • Significance of research
  • Open questions, possible future work

A truly relaxed presentation comes only with experience. Nevertheless, it is possible to control the outward signs of nervousness. Avoid moving around unnecessarily, it can be distracting for your listeners. Also, speak clearly and slowly! A thousand word essay can easily be read in ten minutes. Remember that, since your talk is well prepared, you are making every word count - so don’t rush them.

Let your audience know when you have finished your talk. Simply say “thank you” and allow the audience to clap. The control of the session is then back in the hands of the chair who will moderate the question period. Do not end your talk by just stating “Any questions?” because the audience is then unsure what to do: should they applaud or ask a question right away?

The questions from staff and students may relate to particulars of your talk, or they may be of a more general nature. Answer them to the best of your knowledge, stating when you are venturing an opinion. There will often be questions to which you do not have the answer or to which there is no answer. When these come up you are not necessarily expected to have an answer and will do best by being honest and careful in your reply. Some of the questions may be surprisingly simple. Treat them at face value as the listener may not be familiar with your area of endeavour. Using audio-visual material.

Using audio-visual material

Packages such as Microsoft Powerpoint have become extremely popular for presentations. One can combine text, graphs, images, and even animations into one very effective presentation which can be archived electronically.

Here is some basic advice:

  • The most common error is to cram too much information on each slide. This should be avoided for two reasons. The first and obvious one is that there is a minimum size of type below which the text cannot be read. A font size of 24 to 28 is reasonable for the main text; use a larger size for titles. The second reason is that unless you stop and give the audience time to read the text, they will not have the opportunity to do so because their attention will be focused on what you are saying. Don’t present something that just looks like paragraphs from a book or paper.
  • Avoid colour combinations that are hard to read. In particular, avoid yellow, pale blue, and green lines on a white background. Usually, the contrast between colours projected on the screen is less than on your computer monitor.
  • In diagrams, make sure the lines are thick enough to be visible.
  • Although electronic presentation allows all kinds of fancy visual and sound effects, many listeners find these an annoyance and prefer them to be minimized.
  • Some presenters like the content of the slide to appear one item at a time, each prompted by a mouse click or key press. Don’t overdo this: the audience can find it annoying, and it slows down your ability to page forward/backward through the talk e.g. during question period.
  • When showing graphs, expand the graph area for readability. If you have important graphs to show, generally it’s best to only show one at a time. Crowding multiple graphs onto the same slide often makes them illegible.
  • Label axes on all graphs. Don’t try to crowd too much information on a table or graph. Use colour or pattern to distinguish curves. If scales on axes are broken make this obvious.
  • Acknowledge the source of all illustrations that you did not draw yourself. If you scan a figure from a book or journal, or copy something from the internet, write down your source on the slide. Formally, one should be requesting permission from the copyright holder to use such material!
  • Pointer: one has a choice of laser pointer, a stick, or using the mouse. In practice, using the mouse can be awkward, especially if it’s not your own computer. It also ties you to standing at the computer.

If you can, practice your talk using a computer with the same operating system and same version of Powerpoint as will be used for the actual presentation. It’s common for different computers to have different fonts installed. If the presentation computer is missing a font that you used, it will substitute something else, often with unintelligible results. When this problem arises, it is usually for special characters such as symbols. Similarly, movie animations that work on one computer might not launch on another. If it does happen to you, relax - these things happen all the time to speakers - just carry on.

This seminar guide has been adapted over the years from from one by Martin Yaffe at the University of Toronto.