Building capacity through open approaches: Lessons from developing undergraduate electrophysiology practicals

Background: Electrophysiology has a wide range of biomedical research and clinical applications. As such, education in the theoretical basis and hands-on practice of electrophysiological techniques is essential for biomedical students, including at the undergraduate level. However, offering hands-on learning experiences is particularly difficult in environments with limited resources and infrastructure. Methods: In 2017, we began a project to design and incorporate electrophysiology laboratory practicals into our Biomedical Physics undergraduate curriculum at the Universidad Nacional Autónoma de México. We describe some of the challenges we faced, how we maximized resources to overcome some of these challenges, and in particular, how we used open scholarship approaches to build both educational and research capacity. Results: We succeeded in developing a number of experimental and data analysis practicals in electrophysiology, including electrocardiogram, electromyogram, and electrooculogram techniques. The use of open tools, open platforms, and open licenses was key to the success and broader impact of our project. We share examples of our practicals and explain how we use these activities to strengthen interdisciplinary learning, namely the application of concepts in physics to understanding functions of the human body. Conclusions: Open scholarship provides multiple opportunities for universities to build capacity. Our goal is to provide ideas, materials, and strategies for educators working in similar resource-limited environments.


Introduction
Electrophysiological techniques, like electromyogram (EMG), electrocardiogram (ECG), and electroencephalogram (EEG) recording, are commonly used in both clinical settings and biomedical research. For example, EMG recordings are used to study neuromuscular disorders 1 and spinal cord injury 2,3 ; ECG recordings are used to detect cardiac conduction disorders 4 and heart attack 5 ; and EEG recordings are used to study epileptic seizures 6,7 and sleep disorders 8,9 . Considering the importance of these techniques, it is vital that biomedical students receive training in their physiological basis, how to perform recordings, and how to analyze electrophysiological data, starting preferably at the undergraduate level.
As recently as a decade ago, several factors made doing electrophysiology with groups of students difficult if not prohibitive. Recording equipment was large, not portable, costly, and required expertise to operate. However, in recent years, companies have emerged dedicated to the production of lowcost but high-quality electrophysiology equipment, ideal for use in educational settings. For example, Backyard Brains (BYB; backyardbrains.com) is a company that designs and sells equipment to record action potentials (APs) in insects and plants, EMG and ECG in human subjects, and a variety of other electrophysiology products and accessories, most at prices below $300 U.S. dollars (USD). Many of these devices fit in the palm of your hand and connect to any smartphone, making them highly portable and easy to use. We have entered a new era when electrophysiology can now be easily brought into the classroom. However, many lesson plans and degree programs have yet to catch up.
In 2014, the Faculty of Science at the Universidad Nacional Autónoma de México (UNAM) -Latin America's largest public university -launched its first undergraduate degree program in biomedical physics 10,11 . The overall goal of the program is to provide students with integrative theoretical and practical training in the areas of physics, mathematics, and biomedical sciences, to produce inter-disciplinary professionals that can work in diverse clinical and research environments. Specifically, the objectives of the program include, but are not limited to, educating students in: (1) physics applied to the study of the human body; (2) physics applied to medical diagnosis and therapy; and (3) physical principles underlying the instrumentation and function of the latest biomedical devices 11 . We believe electrophysiology training is an important part of meeting these educational objectives. However, due to limited resources and infrastructure, none of our core courses previously included laboratory practicals in electrophysiology. The same limitations were also affecting our ability to develop electrophysiology research projects with our students.
In 2017 and 2019, we received funds through UNAM's educational innovation grants (PAPIME) program to develop electrophysiology practicals for our biomedical physics students. With a total of nearly $17,000 USD over the last three years, we were able to buy recording equipment, microscopes, computers, instrumentation accessories, and more, and successfully developed electrophysiology practicals which we have released (electrophys.wordpress.com and github.com/emckiernan/electrophys) as Open Educational Resources (OERs) 12,13 . Here we share examples of some of these practicals, their use in biomedical physics education, and how we integrated them into our curriculum. Furthermore, we describe the techniques and tools we used to make the most of the grant funds in a limited-resource environment, and specifically how open scholarship practices (open data, open education, open hardware, open protocols, open source) helped us broaden our impact and build not just educational but also research capacity. We hope sharing our experience will help other academics working in similar environments.

Institutional context
To explain some of the motivation behind this project and its potential impact, it is important to first understand the environment in which we work, both within UNAM and the Faculty of Science.
UNAM UNAM is the largest public university in Latin America 14 . As of 2019-2020, over 360,000 students were enrolled at UNAM, including more than 217,000 undergraduates and 30,000 graduate students 15 . The university has 129 undergraduate and 41 graduate degree programs. Also, in 2019 UNAM served over one million students through its continuing education, including online, programs 15 .
As a public insitution, education at UNAM is nearly free, subsidized by federal funds. Students pay an annual registration fee of just 20 Mexican cents (equivalent to ∼ 0.01 USD). This is combination with UNAM's prestige and reputation for quality education results in a high demand for entry. Each year, less than 10% of applicants are accepted at the undergraduate level through UNAM's admissions testing 16 . In other words, a huge percentage of the eligible student population in Mexico is unable to study at this university that receives the largest share of public funds -an annual budget equivalent to approximately 2 billion USD 17,18 . One could argue that, more than any other public institution in Mexico, UNAM has a responsibility to give back to the community. On the other hand, while UNAM receives more funds than other public universities in Mexico, it still operates on a relatively limited budget considering its size and the number of services offered by the institution. For comparison, consider the University of California, which has a similar though smaller population of over 285,000 students 19 but almost 20 times the budget of UNAM 20 . So, how can institutions like UNAM maximize the use of public funds, both for their benefit and that of the larger Mexican population?
Faculty of Science UNAM comprises 15 faculties, 34 institutes, and various other centers, schools, and units 15 . There are fundamental differences for academics working in faculties versus institutes, which are important for understanding our work as professors in the Faculty of Science.
In institutes, the primary focus is research. Laboratory space is assigned to many faculty at the time of hiring and their teaching load is low. According to the UNAM Statute of Academic Personnel, researchers in institutes must teach a minimum of 3 contact hours per week each semester 21 , equivalent to a 1-1 teaching load at Canadian or U.S. institutions. In contrast, faculties are focused on teaching. Entry-level professors are required to teach a minimum of 9 contact hours per week each semester 21 , equivalent to a 3-3 teaching load. Unlike at many institutions in North America, there are no standard mechanisms for 'buying out' of teaching if a professor receives a grant. In addition, professors are expected to contribute significantly to 'formation of human resources' by directing student social service projects and theses, serving on committees, and tutoring. Many professors work almost exclusively with undergraduates, especially for the first few years when their professoriate level does not allow advising graduate students in many degree programs. Despite the heavy teaching and service load, there is still a research expectation. However, professors are not necessarily assigned laboratory space and receive no start-up funds. Availability of laboratory space in the Faculty of Science has become especially problematic as student and academic population growth puts increasing demands on an already overloaded infrastructure.
These conditions raise a number of questions for professors working in faculties like UNAM's Faculty of Science: With limited resources and infrastructure, how do I provide high-quality, hands-on educational experiences for my classes?; how do I develop research projects for social service and thesis students?; and how do I build up my own research program and start producing?

PAPIME educational grants
A partial answer to some of the above questions comes in the form of internal grants offered by the General Directorate for Academic Personnel Affairs (Dirección General Asuntos del Personal Académico; DGAPA) at UNAM. One of these grant programs -the Support Program for Projects to Innovate and Improve Education (Programa de Apoyo a Proyectos para Innovar y Mejorar la Educación; PAPIME) -focuses on education 22 . This program has been key for us in building capacity and is a funding mechanism we believe more universities should emulate. The goal of PAPIME is to, "Promote the improvement and development of academic staff by supporting projects that lead to innovation and improvement of the teaching-learning process and benefit students...Teaching innovation projects should revolve around themes that allow creative teaching, with new ways of thinking, to motivate the interest and imagination of students" 22 .
The 2020 call for applications 23 shows that these grants fund a wide range of projects and diverse products, including but not limited to: (1) teaching materials, like exercises or practicals, case design, tutorials, digital applications, software, and websites; (2) publications, like books or articles in areas such as educational research; (3) innovative educational evaluation systems, strategies, and instruments; (4) organization and participation in academic events, like colloquia and seminars; and (5) training activities, like in-person or online courses and workshops, or fieldwork.
These grants are typically 1 year in duration, and as of 2020 can be awarded up to $250,000 MXN annually 23 , or ∼12,500 USD. Interestingly, while not explicitly using the language, PAPIME grants can function to a certain extent as OER grants. Products resulting from PAPIME projects are required to be uploaded to UNAM's Repository of Educational Innovation (Repositorio de Innovación Educativa, RIE: innovacioneducativa.unam.mx). Digital materials in particular must be uploaded to UNAM's University Learning Network (Red Universitaria de Aprendizaje, RUA: rua.unam.mx). The stated objective of sharing these materials is to "disseminate and extend coverage for the benefit of the university community and thus optimize the resources invested by UNAM in development of the project" 23 . In line with this, projects are evaluated on several characteristics related to broadness of impact, including: (1) number of students that will benefit from the project; (2) where students come from, whether inside the academic entity, university, or beyond; (3) number and names of classes that will benefit from the materials; and (4) number of professors that will use the products. Using an open approach can help academics argue broader impact, i.e. a larger population of both students and educators are reached, within and beyond the institution, and materials can be reused, revised, remixed, and redistributed 12 .

Electrophysiology grant
We were awarded our first PAPIME grant in 2017 and our second in 2019. The idea for the overall project came from what we perceived to be a gap in the education of our Biomedical Physics undergraduates, namely a lack of hands-on training in electrophysiology and related skills. We set out to design electrophysiology laboratory practicals that could be incorporated into our plan of study. Not all these practicals were intended to be 100% novel; resources exist on the basics of EMG 24,25 and ECG 26,27 recording, for example. BYB has already developed over 60 experiments that can be performed using their equipment and released these on their website (backyardbrains.com/ experiments) as OERs under an open license. However, there are a few ways we wanted to expand and extend existing work.
First, we wanted all our practicals to be accompanied by more in-depth lesson plans. The BYB tutorials are excellent starting points, but are too simple for our fourth-semester human physiology undergraduates (e.g., their EMG tutorial 24 is marked as 'beginner' for elementary school students 5th grade and up). On the other hand, many of the resources we found in the scientific literature were too complicated, aimed more at graduate students or working professionals (e.g., 25). In addition, many of these latter resources have a clinical rather than biophysics focus. We saw a need for electrophysiology OERs designed for a more intermediate, undergraduate level that would reinforce material seen in our core courses, including physics as applied to the human body.
Second, we aimed to develop novel practicals that would combine electrophysiology with other physiological measurements like spirometry, helping our students see how different systems in the body work together. Currently, as in many universities, our human physiology course is taught as a sequence of system-based modules (e.g., nervous system, cardiovascular system, etc.). As Conford 28 writes, "One assumption that many modular courses presently reflect is that effective learning proceeds via self-contained chunks of information...Modules, however, by their very structure, tend to fragment knowledge rather than to integrate it" (pg. 243). We see these practicals as a way to recover this integration and connect concepts across modules.
Third, we sought to develop bilingual materials. We have struggled to find quality Spanish-language OERs, especially in biophysics. Language can be a significant barrier to OER reuse and remixing [29][30][31][32] . From an UNESCO report 33 , "Not only does the English language dominate OER provision, but English-language content tends to be based on Western learning theory. This limits the relevance and accessibility of OER materials in non-English, non-Western settings. There is a risk that language barriers and cultural differences could consign less developed countries to the role of OER consumers rather than contributors to the expansion of knowledge" (pg. 12).
Finally, we wanted to develop a suite of products around each practical and release not just written OERs but also accompanying code, data, images, and videos, all under open licenses. We reasoned this was one important way to increase the impact of the project. For example, educators without the resources to buy recording equipment could at least reuse our data and code to graph and analyze electrophysiology recordings with their students.

Building capacity
With our first grant in 2017, we were awarded the equivalent of ∼$10,500 USD. We used the bulk of the funds to purchase electrophysiology recording equipment (Table 1), microscopes, and related accessories like electrodes and dissection tools ( Table 2). The remaining funds were used to finance scholarships for two undergraduates to work on the project. With our second grant in 2019, we were awarded ∼$6,300 USD that we used to purchase computer equipment, surface electrodes for recording, instrumentation accessories like Raspberry Pi 3 Model B and Module V2 cameras, Arduino sensor kits, and food and bedding for experimental animals. (Many of these are standard products available from multiple providers, so we did not itemize these in table form). We also gave scholarships to two more undergraduates. While the amounts awarded us may sound sizeable -similar OER grants in the U.S. and Canada often cap at $5,000 USD 34-36 -this was still a limited budget considering we were starting from zero in terms of equipment and materials. We had to maximize use of these funds to build capacity.

Electrophysiology equipment
All electrophysiological recording equipment was obtained from BYB (backyardbrains.com), including devices to record APs in insects (Neuron SpikerBox), and EMGs (Muscle SpikerBox) or ECGs (Heart and Brain SpikerBox) in human subjects (Table 1). We purchased these as bundles, which included the recording device, cables, surface electrodes, conductive gel, and other accessories. The low cost of these bundles (<$250 USD each), compared to conventional electrophysiology equipment, allowed us to purchase multiple devices. With 3 or 4 devices, we could work in groups of 5-10 and pilot practicals with classes of 20-30 students.
We also purchased DIY kits from BYB to build additional recording devices. This served two purposes. First, students will assemble the devices, learning valuable instrumentation skills in the process, which is one of the core objectives of our Biomedical Physics plan of study. Students will fully document the assembly process with step-by-step protocols, photos, and videos, which will be shared online as OERs. Second, at around half the price of the fully assembled device bundles, DIY kits allowed us to buy more equipment without exceeding our budget. Once assembled, our recording capacity will double, meaning we can work with more students.
Affordability was not the only advantage of the BYB equipment. The small size and portability of the devices meant we did not need a dedicated laboratory space, solving one of our infrastructure issues. We could take these devices into any classroom and record with students using their smartphones. We also allow students to borrow these devices and take them home to work on individual research projects. A few years ago, having students do electrophysiology at home would have been impossible. Now we can offer them this unique experience, which can be a huge motivating factor in their academic development. Since 2017, students and professors in our program have used the equipment in core and elective coursework, social service projects, and thesis research. In other words, purchasing a small amount of equipment has greatly increased our capacity to provide high-quality educational and research opportunities for our undergraduates.

Electrophysiology accessories
We purchased several accessories to improve both recording experiences for students and potentially research capacity ( Table 2). For example, while the BYB electrodes that come with the Neuron SpikerBox are sufficient for basic AP recording in large insects, they are stainless steel sewing needles with a relatively large tip diameter (0.25-0.6 mm) 37 , non-insulated, and not ideal for finer recordings in smaller preparations or cells. So, we purchased insulated Tungsten electrodes with a 2-3 µm tip diameter. At ∼$19 USD each, these electrodes are only $9 more than BYB's, but should provide a substantial improvement in recording capabilities and quality. For less than $200 USD we can buy a packet of 10 Tungstens and upgrade 10 SpikerBoxes.
We also purchased manipulators to improve control and precision of electrode placement. Conventional 3-axis manual micromanipulators, like those made by Narishige (usa.narishige-group.com), cost ∼1,000 USD and were out of our price range. However, BYB provides a 3-D printed plastic manipulator

Dissection tools and microscopes
Electrophysiology often involves dissection to prepare tissues or cells for recording. Dissection tools, especially those for fine dissection, are costly. Fortunately, companies like VWR provide economic solutions in the form of classroom dissection sets, which include scissors, forceps, scalpels, pins, and more. With tools for up to 20 students and priced at ∼ $200 USD or less, the cost comes out to only ∼ $10 USD per student. With the remaining funds we had for tools, we bought just two fine dissection kits for use in more advanced, individual student projects.
Dissection and fine detail instrumentation also requires visualization and magnification. With this in mind, we purchased several microscopes with different characteristics. The Fisherbrand Illuminated Pocket Microscope weighs just 85 grams, measures 140L × 38W × 22H mm, and has 60-100x magnification. Similarly, the BYB High Power RoachScope weighs 400 grams, measures 142L × 94W × 74H mm, and can be used in combination with any smartphone camera. With digital zoom, it has 5-100x magnification. Both these microscopes are designed for maximum portability, so they can be taken into the classroom. In addition, both cost less than $100 USD each, so we could buy several to work with groups of students. We also purchased three Fisher Science Education Advanced Stereomicroscopes for just under $300 USD each. These microscopes are not very portable, but should give us better optics.
To increase the utility of these microscopes, we are planning on 3-D printing a low-cost adapter that will attach to the eyepiece and allow us to mount any smartphone to take high-quality pictures or video. Open plans for such an adapter are available via the NIH 3D Print Exchange 42 and pictured in 39.
Finally, we purchased an advanced trinocular microscope (National Optical via Fisher Scientific) with 4x, 10x, 40x, and 100x objectives and a built-in digital camera (Moticam 1080 HDMI & USB) for high-resolution viewing of tissues and cells. The higher cost of this microscope meant we could only buy one. However, connecting the camera to a large computer monitor allows us to carry out demonstrations and have groups of students view samples simultaneously. We have also hosted "open house" events for new students using this microscope.

Data acquisition and analysis
Commercial software used to record, process, and analyze electrophysiology data is often a significant expense for many laboratories. We did not have the budget to pay for software licenses, but also felt to do so would Python code was developed inside Jupyter notebooks, which provide an interactive way to document and share code 50,51 . Our notebooks walk students through the process of opening and graphing recordings, applying filters, and quantifying aspects of electrical activity. The notebooks include exercises for students to perform in or outside of class as data analysis practicals. In other words, we create OERs out of this shared code [52][53][54] . As Downes 53 writes, use of Jupyter notebooks in this way "changes the conception of an educational resource from something static to something that's interactive, to something that can be used to create, as well as to consume" (pg. 9). This also helps us meet another core learning objective of the Biomedical Physics plan of study, namely programming skills. In fourth semester, when students start with our electrophysiology practicals in their human physiology course, they also take a programming course which primarily teaches Python. Data analysis practicals are a good way for them to apply new programming skills to biomedically relevant data analysis, and integrate knowledge from these two core courses.

Workflow and related tools
The development of most of our practicals began as a free-form process. We had a general idea of the type of recording we wanted to perform, and piloted these ideas first with classes of 20-30 undergraduates. Students were encouraged to experiment, for example by trying out different electrode placements and exercises. Subsequently, students wrote individual reports with background information, protocols, results, and conclusions. Students shared their photographs, videos, data, and reports with us via Google Drive. From the resources provided by students, we collected the best examples and used this information to build our master documents for each practical. In addition, four students were given scholarships with PAPIME funding to help us run pilots, gather materials, analyze data, and draft protocols. In this way, students played an active role in OER development, reinforcing Buckland's idea of "students as content creators" 55 .
Our master documents were written in LaTeX using the Overleaf platform (overleaf.com). LaTeX presents a variety of advantages over word processing software, including control over document layout and figure placement, excellent equation handling, and automatic reference formatting 56 . The Overleaf platform in particular provides several benefits. First, basic accounts are free, so there were no additional costs, as would be incurred by using commercial packages like Microsoft Office. Second, we could easily share and collaborate on a master file with integrated commenting functions and version control. Finally, Overleaf provides a rich text viewing option, which is more user-friendly, especially for undergraduates just starting out with LaTeX.
Once we had final versions of the master documents, these were uploaded to a public repository on GitHub (github.com/ emckiernan/electrophys), along with images, data, and code associated with each practical. GitHub provides Git version control 57,58 , which means OERs can continue to evolve as necessary while preserving the history of resource development 59 .
GitHub also provides collaboration features, which we hope students and educators will use to improve and customize these materials. However, we recognize not everyone uses GitHub, and that only hosting our materials there could represent a barrier to reuse. So, we built a Wordpress website to share materials in a more user-friendly way. This was done by converting our LaTeX documents to html using Pandoc (pandoc.org), and then copying the html to a free Wordpress template (wordpress.org). Minor formatting to improve visual presentation was done by hand. Jupyter notebooks were uploaded by creating public gists (gist.github.com) and then copying these links to the Wordpress site for embedding. Using the free Wordpress services meant we did not incur any costs for website creation or hosting.
Our workflow is visualized in Figure 1. Moving forward, there are ways we could improve this workflow. For example, a more efficient way to set up our website would be to use GitHub Pages (pages.github.com). This would allow for automatic syncing of the website when the repository materials are updated, but requires more in-depth html knowledge to properly format and maintain the site. We would also like to explore open source alternatives to several of the tools we used. Bosman and Kramer outline a potential open science workflow (as well as other workflows ranging from traditional to experimental) that could be useful for researchers and educators 60 . More information on tools, including some open source alternatives, is available in Table 3.

Practicals and course integration
In fourth semester, students in UNAM's Biomedical Physics undergraduate degree program take a human physiology course, which is divided into modules: (1) nervous, (2) musculoskeletal, (3) biofluids (4) cardiovascular, (5) respiratory (6) gastrointestinal, and (7) renal systems. At present, this class is only lecture. One of our goals with this project was to design hands-on electrophysiology activities to be integrated throughout the course. We briefly describe some of these practicals, where they fit into the course, and how they reinforce  EMG basics: recording from the body's lever systems The first practical is designed to teach students the basics of EMG recording, carried out at the end of the musculoskeletal system module. The background written information reinforces physiology concepts seen in class, as well as the application of basic physics concepts seen in other coursework. It begins with a description of how muscle-bone-joint complexes function as lever systems. Students are encouraged to think back to the three types of classical lever system and find corresponding examples of these in the human body. This involves visualizing biomechanics and how the relative position of bones, joints, muscles, and loads will affect movement. The written documentation goes on to reinforce concepts such as how muscle structure affects tension development, length-tension relationships, and the energy requirements for muscle contraction. We then describe the basics of EMG recording, comparing the advantages and disadvantages of invasive versus surface recording, and the basic bipolar differential recording configuration. Study questions prompt students to think about where they will need to place electrodes to record from different muscles and what potential limitations they might encounter.
Students then move on to the experimental phase of the practical where they carry out their own EMG recordings, in groups of 4-6 students depending on class size.
Step-by-step instructions on how to perform the recordings are included in the written documentation. However, these are designed to be informative without being too prescriptive and still allowing for exploratory learning. The only requirement is that students record from at least one muscle from each type of lever system, but we do not tell them which muscles to record from or how they should activate these muscles. Students are encouraged to design their own experiments using everyday items available in the classroom or simple exercise aids, like resistance bands or hand grippers, brought from home. Students have performed EMGs from facial muscles while eating, tricep muscles while doing pushups, bicep muscles while arm wrestling or lifting their backpacks, and forearm muscles while performing martial arts movements. Students are also encouraged to explore different types of contraction, including intermittent versus sustained and increasing versus decreasing force.
We have several other EMG practicals still under development, including experimental ones to measure fatigue in the bicep muscle, dual recordings from antagonistic muscle pairs, simultaneous recording of EMG and force sensor measurements from the forearm, and others. In addition, we are developing EMG data analysis practicals. Students will take the recordings they gathered in the first practical, graph them, and learn about the design and application of band-pass and low-pass filters to process their data. They will learn about different techniques used to smooth data, calculate an envelope, and use thresholds to detect start and stop times of muscle contractions. These practicals will be carried out using Jupyter notebooks running Python, thereby simultaneously strengthening students' programming skills. All practicals are listed in the extended data (doi.org/10.5281/zenodo.4540355 65 ).
ECG basics: recording heart electrical activity before and after exercise In module 3 of their human physiology course, students learn about the cardiovascular system and carry out a practical to record their ECG before and after exercise. The background written information begins with a description of how the heart performs external mechanical work. Students are encouraged to visualize the heart as a single-chamber pump with inflow and outflow valves, and examine the pressure-volume relationships similar to the way one would with an internal combustion engine 66 . Students learn about sequential pressure and volume changes in different chambers of the heart during the cardiac cycle, and how to graph this with a pressure-volume loop. The documentation goes on to describe the electrical activity of specialized populations of cells in the heart, including the ionic basis of APs in these cells. Discussing cardiac muscle activity also encourages students to think back to module 2 of the human physiology course when we discussed contraction mechanisms in this muscle type. Finally, we describe the basics of ECG recording, including how the summation of individual potentials leads to the extracellularly recorded events, different recording configurations, and the importance of electrode placement.
Students then move on to the experimental phase, working in groups of 4-6. Volunteers from each group record their ECGs, while other students help with organizing and exporting the data. Students first record their baseline ECG under resting conditions for at least 1-2 minutes. Then, they disconnect the recording device while leaving the electrodes in place and perform light to moderate exercise for at least 5 minutes. After this, students reconnect the device and record their ECG again for at least 1-2 minutes. Students can choose the type of physical activity they perform. For example, students have done push-ups or burpees, ran laps around the building, or gone up and down stairs outside the classroom. We encourage students to compare how different levels of activity change the ECG signal, and how the signal varies across subjects (e.g., athletes versus non-athletes).
We are developing additional experimental practicals designed to explore the relationship between heart and respiratory activity, using simultaneous ECG and spirometry, for example. We are also working on ECG data analysis practicals. Students will take the recordings they gathered during the first ECG practical, graph them, and learn how to detect the peaks of the QRS complex to calculate heart rate and quantify how it changes after different levels of exercise. They will also examine techniques for detecting the P and T waves, and calculating intervals important in clinical evaluations. A full list of these practicals is available in the extended data (doi.org/10.5281/zenodo.4540355 65 ).
Recording accessory muscles during normal and forced respiration In module 5 of their human physiology course, students learn about the respiratory system, an important part of which is understanding the mechanics of breathing. How do respiratory muscles expand or contract the thoracic cavity and change pressure gradients? How does the participation of different muscles change when respiration is normal versus forced? And to relate back to the musculoskeletal module, how is respiratory muscle contraction related to electrical activity?
In this practical, students record EMGs from the rectus abdominis. This muscle is known as an accessory respiratory muscle because it is not activated during normal exhalation, but is activated during forced exhalation when additional effort is needed to reduce the volume of the thoracic cavity beyond that accomplished by simple elastic recoil 67 . While recording rectus abdominis EMG, students simultaneously use a spirometer (Vernier) to measure the volume of air moved in and out of the lungs. Students are instructed to perform a sequence of normal breaths interspersed with maximal forced inhalations and exhalations.
The dual recordings allow students to see firsthand that, during normal respiration and forced inhalation, little to no electrical activity is recorded on the EMG because the rectus abdominis is not contracting. However, during forced exhalation, the EMG shows an increase in both the amplitude and frequency of the signal with increased effort and increased volume exhaled. The written materials for the practical are designed to reinforce several physical concepts applied to the study of respiration, including: (1) pressure-volume relationships and Boyle's Law as applied to the lungs; (2) importance of pressure gradients and Ohm's Law as applied to airflow; (3) Poiseuille's Law as applied to measuring airflow through a spirometer, and (4) biomechanics of active lung expansion versus passive elastic recoil. This practical also gives students the opportunity to integrate knowledge from two modules to understand how the musculoskeletal and respiratory systems work together. We are working on developing more practicals that combine electrophysiological recordings with other physiological measurements (e.g., from force, displacement, or gas sensors) to provide similar integrative learning experiences.

Discussion
Incorporating electrophysiology into undergraduate education Less than a decade ago, providing hands-on electrophysiology learning experiences for undergraduates, especially large classes, was not feasible. However, over the last few years, technological advancements have opened up new possibilities for educators. With the introduction of the BYB Neuron SpikerBox in 2011, an easy-to-use, low-cost bioamplifier brought neurophysiology into the classroom 37 . Since then, the single-channel SpikerBox, and the later two-channel version, have been used to design practicals for undergraduates to record from cricket sensory organs 68 , grasshopper neurons responding to visual stimuli 69 , and to study AP conduction velocity in earthworms 70 . Surveys from these studies indicate that students not only enjoy these hands-on activites, but that they also improve learning outcomes, increasing test scores by as much as 25% on average 70 . The SpikerBox has even been used as part of a larger program to provide undergraduates the opportunity to teach neuroscience to highschool students 71 .
In recent years, BYB has released more complex devices for recording ECG, EMG, and single-channel EEG, which have also been used in undergraduate class settings to improve learning. For example, Catena and Carbonneau (2018) describe using the BYB Muscle SpikerBox Pro to record dual-channel EMG as part of an undergraduate biomechanics course 72 . Their survey results show that students reported "better motivation" and higher "personal responsibility for learning". Test scores for students who had these hands-on learning experiences were also 7% higher compared to students who did not 72 . Similarly, Judge and colleagues (2020) used BYB equipment to develop ECG and EMG exercises for community college anatomy and physiology courses 73 . Students who carried out these exercises showed "significant learning gains" 73 .
Other groups have also developed and shared plans for low-cost electrophysiological recording devices, and in the process created instrumentation exercises for undergraduates. Matsuzaka and colleagues (2012) describe the development of a low-cost (only $85 USD per unit) amplifier for recording EEGs, EMGs, and other electrophysiological signals with students 74 . Importantly, the authors mention that potential problems of reproducibility and quality control when building these devices "could be resolved if the optimized circuit layout is freely available" (pg. A124). Crisp and colleagues (2016) provide step-by-step instructions for students to build a simple EMG device using a breadboard amplifier, with few components and an assembly time of just 30 minutes 75 . Wyttenbach and colleagues (2018) review these and other devices as part of a larger discussion on "reducing the cost of electrophysiology in the teaching laboratory" 76 .
It is not just low cost that is important, but moreso the open approaches taken that have increased the impact of many projects like the ones described above. Sharing hardware schematics and building instructions, openly licensing and publicly documenting code, and growing online support communities -characteristics of projects like Arduino, BYB, and Raspberry Pi -have allowed classrooms and laboratories in limited-resource countries to build capacity 77 . When we thought about code as not just for research but also an educational resource, it changed how we thought about sharing this product. Previously, we might have simply shared our code as a raw Python file in a GitHub repository, and included some in-line comments and a README file as documentation. However, when we envisioned students or educators reusing our code to learn, we realized it needed more in-depth explanations and exercises. Using Jupyter notebooks, we built up tutorials or lesson plans surrounding the code and transformed these into OERs [52][53][54] . Importantly, after completing these resources we not only had quality OERs to use for our classes, but we also had a bank of well-documented analysis tools to use for future research projects. Organizing and documenting our code in this way may help us with lab group onboarding, as any incoming students can go through the tutorials and quickly get up to speed on our data analysis techniques. Furthermore, in the process of elaborating didactic explanations of our analysis, sometimes we realized ways in which our code could be improved. Therefore, the interaction went both ways: building educational capacity through open practices led to building research capacity and vice versa.

Connections between open approaches
We have visually mapped out some of the potential connections between different open approaches ( Figure 2). While this is not an exhaustive map -other open approaches could be included and other connections explored -it is a representation of how these connected for us in this project.
Importantly, one open approach did not automatically lead into the next; there were transformations of the materials and certain conditions that needed to be met at each stage to maintain the flow between each. For example, simply sharing our data would not necessarily allow others to develop new analysis code ( Figure 2, upper right arrow). For this to occur, the data need to be well organized, labelled, and documented, with meta-data included. Admittedly, we are still struggling with the best ways to do this to optimize reuse of our data, and believe this is one area where researchers would benefit from more training. Similarly, shared code does not necessarily become an OER (Figure 2, central left-pointing arrow). This requires that the code be well explained, often with a surrounding lesson plan and exercises. Open licensing at each stage was also key, since locking down content at any point would stop the flow. However, licensing is different for code, data, and documents. To select licenses for each product, we used resources like the Creative Commons License Chooser (creativecommons. org/choose) and GitHub's Choose an Open Source License tool (choosealicense.com). We would also like to encourage researchers to expand their ideas of what they consider an OER. When working with students, there is no clear line where education ends and research begins. The research we do with students, especially with undergraduates, is not necessarily to discover new things but rather to teach students how to do research. It is more about the process than the end result, and as such, everything we create during that process -protocols, code, data, notebooks -can potentially be transformed into an OER to train others. We also believe that in thinking of research as a teaching-learning process, with all the documentation and explanations that entails, we may in turn enhance the research itself, improving experimental design and reproducibility.  [85][86][87] . We hope to see even more of this intersectional work in coming years, and expect that much of it will arise in libraries.
Libraries are also increasingly both leading and funding open scholarship projects, including the development and implementation of OERs. In a 2016 survey of U.S. universities, 64% responded that it was the library who had originated affordable course content (ACC) or OER initiatives at their institutions 88 . For those with governing bodies overseeing these initiatives, 89% said that libraries were participating members and half said that libraries led the group. Over half of respondents also indicated that funding for ACC/OER initiatives came from library general operating budgets -more than any other institutional or external funding source.
Despite library support for open initiatives, it seems other institutional policies have not necessarily caught up. Walz and colleagues 88 write, "survey responses indicate that current university-wide tenure and promotion policies do not explicitly encourage faculty adoption, adaptation, or creation of ACC/OER" (pg. 5). We believe it was important for the success of this project, as well as our own professional development, that our department recognizes and values participation in PAPIME projects in annual performance, promotion, and tenure reviews, and gives us space on evaluation forms to report on non-traditional digital products, including OERs. We encourage institutions to rethink and reform their evaluation policies to incentivize open scholarship, including OER development and adoption, and to seek guidance from libraries on how best to do this.

Conclusions
We succeeded in developing a suite of electrophysiology practicals to give biomedical physics undergraduates handson training in this important area of study.