Experts in medicine, academia, and industry explored the promises and perils of the applications of artificial intelligence (AI) in health care during panel discussions with the Raw Talk Podcast on May 7. The event was organized by graduate students of U of T’s Institute of Medical Science.

The two panels, collectively named “Medicine Meets Machine: The Emerging Role of AI in Healthcare,” aimed to demystify sensationalism and clarify misconceptions about the growing field of study.

“On one hand, it seems like everyone has heard about [AI],” said Co-executive Producer Grace Jacobs. “But on the other hand, it seems like there’s a lot of misunderstanding and misconceptions that are quite common.”

How AI is used in health care

While discussing the reality of AI, several panelists emphasized that it should be viewed and treated as a tool. “It is statistics where you don’t have to predefine your model exactly,” said Dr. Jason Lerch of the University of Oxford.

Other speakers agreed that AI is an expansion of — or a replacement for — traditional statistics, image processing, and risk scores, as it can provide doctors with more robust and accurate information. However, final health care recommendations and decisions remain in the hands of doctors and patients.

“You always need a pilot,” said Dr. Marzyeh Ghassemi, a U of T assistant professor of computer science and medicine.

But what advantages can this tool provide? Ghassemi thinks it can assimilate clues from a wider range of patients’ conditions to predict treatment outcomes, replacing the experience-based intuition that doctors currently rely on.

Speaking on her time in the Intensive Care Unit as an MIT PhD student, Ghassemi said, “A patient would come in, and I swear they would look to me exactly the same as prior patients, and the… senior doctors would call it. They would say, ‘oh, this one’s not going to make it. They’re going to die.’ And I would say, ‘Okay… why?’ And they said, ‘I’m not sure. I have a sense.’”

“They used different words — gestalt, sense — but they all essentially said the same thing. ‘I just — I have a sense.'”

Doctors develop this sense by seeing many cases during their training, but they can intuit only the cases that they had personally experienced; AI algorithms can potentially understand many more cases using a wider dataset.

Accessing those cases requires access to patient data, and access to data requires conversations about consent and privacy. Ghassemi and Dr. Sunit Das, a neurosurgeon at St. Michael’s Hospital and Scientist at the Keenan Research Centre for Biomedical Science, said that “de-identification” — the removal of information that can be traced back to individual identities — protects privacy.

Large de-identified datasets from the United States and the United Kingdom are available for AI research, but generally, Canada lags behind these countries in making health data available for this purpose.

Dr. Alison Paprica, Vice-President of Health Strategy and Partnerships at the Vector Institute, agreed that data should be used for research, but argued that de-identification alone does not eliminate risk.

“You’re not just giving a dataset to anybody,” she said. “You’re giving a dataset to people who are extremely skilled at finding relationships and patterns and maybe piecing together information in ways that most people couldn’t. So I think there’s going to be heightened sensitivity around re-identification risk.”

Society must manage this risk and balance it against the benefits. “How do we balance that?” Paprica asked. She suggested that consulting all involved stakeholders could help strike that equilibrium.

Advice for scientists aiming to use AI in their research

So what advice did the panelists have for scientists hoping to harness the power of AI in their own research?

Ghassemi stressed the importance of knowing what you’re doing: researchers have created many tools that make AI research easy to implement, but conscientious scientists need to know the statistical and training principles behind the methods.

“If you’re not aware of how these things are trained,” she said, “it’s really easy to misuse them. Like, shockingly easy to misuse them.”

Other panelists advised users to take care when choosing data to train the algorithms. “A learning algorithm can’t overcome bad data that goes in, or can’t completely overcome it,” said Lerch.

Moderator Dr. Shreejoy Tripathy summed up a key takeaway on applying AI to health care: “Understand your data… And understand your algorithms.”

Tenured and tenure stream female faculty earn on average 1.3 per cent less than their male counterparts, according to the findings of a committee convened by U of T Vice-President & Provost, Cheryl Regehr.

In response to the committee’s findings, the university has announced that it will increase the salaries of “all women faculty members who are tenured or in the tenure stream” — around 800 women — by 1.3 per cent.

The pay bump builds on previous university initiatives, including the increased hiring of women into the tenure stream, providing bias training for faculty and tenure appointment committees, and support for starting salary equity.

The report was carried out by the Provostial Advisory Group, a committee of staff, faculty, and educational administrators convened in 2016 as the product of a pre-grievance mediation between the university and the University of Toronto Faculty Association.

The report was carried out by faculty members specializing in statistics, one staff member, and a graduate student. In addition to analyzing salary differences among tenured professors, the group examined the male-female divide in professorships and calculated pay gaps for teaching stream faculty.

Men drastically outnumber women in professorships

Using numbers from 2015–2016, the group found that of the 965 tenured or tenure stream professors then employed at U of T, only 27 per cent were female. The numbers increased for associate professors and assistant professors, with women respectively accounting for 45 per cent and 43 per cent.

The “raw” salary difference within this group was 12 per cent, meaning that the average tenured or tenure stream male made 12 per cent more than his female counterpart. However, after considering academic rank, years since highest degree, and field of study, that difference was lowered to 1.1 per cent.

This large initial difference of 12 per cent was chalked-up to the tendency for women to hold junior positions and specialize in lower paying fields. In the relatively high-paying Economics faculty, for example, there is only one woman for every eight men.

The final model incorporated three more variables, including whether the person had received a Canada Research Chair, and if the person was holding, or had held, an administrative position. Taking these into account with the original model, they determined that gender was responsible for roughly 1.3 per cent of the salary gap.

Working by percentage, the corresponding increase will affect salaries differently. A woman earning $350,000 will receive an extra $4,550 per year, while a woman earning $100,000 will only take in an extra $1,300.

Ratio even, salaries equal in teaching stream

The committee also investigated pay differences within the teaching stream faculty, comprised of professors who are expected to exclusively teach. The data, taken from 2016–2017, showed that women accounted for 50 per cent of the 336 full-time teaching stream faculty.

Unlike the 1.3 per cent divide within the tenured and tenure stream, the committee found no discernible gap between male and female salaries within the teaching stream.

Still more to come, promises Regehr

According to Regehr, U of T will be “conducting a similar analysis for librarians in continuing appointments.”

In an interview with U of T News, Regehr also recommended that the university check in on salaries periodically “to ensure that a gender-based pay gap does not reappear over time.”

The Varsity’s own analysis of professors who earned over $100,000 in 2017 showed that women were not only largely under-represented in high-paying positions but also faced various pay gaps.

Every year, the Office of Planning and Budget Office releases a report on the demographic data that U of T collects, including figures on international enrolment, the number of degrees awarded by field, and even the average number of wireless connections per day.

Notably, engineering and science degrees were heavily skewed toward male recipients, while education and physical education degrees were mostly given to female students.

The report also shows that an overwhelming amount of international students at U of T are from China, with other countries making up a small percentage in comparison.

Here’s a breakdown of what that data shows and what stood out.

Student gender balance

Of the 65,051 full-time undergraduate students last year, 55.7 per cent identified as female, 43.7 per cent as male, 22 students as another gender identity, and 341 students’ gender identities remained undisclosed. In its collection of data on student gender, the university only started including the category of “another gender identity” in 2017.

In comparison to figures from 2007, the university has maintained the ratio of female to male full-time undergraduate and graduate students.

Part-time undergraduates were 61 per cent female in 2007. The 2017 data shows a slight majority male student population among part-time undergraduates.

Part-time graduate students had the largest disparity in gender, with 64.4 per cent of the population identifying as female — two per cent up from 2007 numbers.

Data on the number of degrees awarded by field of study for the 2016 calendar year shows large gender disparities in the areas of engineering and physical sciences, education and physical education, and mathematics and physical sciences.

Engineering and physical science degrees overrepresented male students, with only 380 undergraduate degrees out of 1,186 being awarded to female students, amounting to less than 33 per cent.

Disparities are especially apparent in doctoral engineering and physical science degrees, where only 26 per cent of the 156 degrees awarded were to female students.

Among the 1,115 undergraduate mathematics and physical science degrees awarded in 2016, 39 per cent were to female students. These same disparities appear for doctoral degrees as well, with only 24 per cent of the 105 doctoral degree recipients and 31 per cent of the 118 master’s degrees being awarded to female students.

Education and physical education degree recipients also showed gender disparities, where female students are overwhelmingly represented. Across 1,287 undergraduate, master’s, and doctoral degrees awarded, three-quarters were female, with the largest disparity among the 759 masters students, where only 21 per cent were male.

International student enrolment

International students who attend U of T are overwhelmingly from China.

With 65.1 per cent of the undergraduate international student enrolment, the 10,463 Chinese international students made up 14.6 per cent of U of T’s total undergraduate population in 2017.

The second-highest international population was from India, with a comparatively few 677 students enrolled. Students from South Korea, the United States, Hong Kong, Taiwan, Saudi Arabia, Japan, and Nigeria made up the remaining international undergraduate student population with roughly 12.8 per cent share of total international undergraduate enrolment.

Trends remain similar for graduate international enrolment. Students from China made up 34.7 per cent of the graduate international student population — with students from the United States and India having made up 11.4 and 11.2 per cent of international graduate students, respectively.

By geographic region, undergraduate international enrolment has fluctuated. Enrolment from North America has increased from 281 to 449 students since 2013, while international students coming from the Caribbean and Latin America are on a rapid decline, with 2017 seeing about half of the 2014 enrolment. However, European international student enrolment maintained high levels, at around 800 students per year.

The Asia and Pacific region’s enrolment has seen a 68.9 per cent increase since 2013, more than any other regional division of international enrolment for undergraduate students.

Again, these trends are mirrored in the graduate student population. Of the 3,118 international graduate students in 2017, more than half were from Asia and the Pacific, with North America and the Middle East making up the next largest populations.

Donations

In the 2016–2017 school year, U of T received $274,854,977 in pledges and gifts, with 37 per cent of donations coming from alumni. Research grants also made up a large proportion of donations at $62,535,116. The university also received money from various corporations, foundations, and “friends.”

The largest donors are listed online. Donors who have gifted $25,000,000 or more include Paul and Alessandra Dalla Lana, Sandra and Joseph Rotman, John H. and Myrna Daniels, and Peter and Melanie Munk.

If an individual donates $1,827 or more, they can join the Presidents’ Circle club. The club holds special lectures and events presented by “the University’s most celebrated, insightful and inspiring professors.”

Donations are also accepted online, where various funds can be selected to specify where the donor would like their money to go. This includes funds specific to programs, institutions, campuses, and colleges. There is also a President’s Fund for Excellence, listed as part of the Boundless campaign’s “area of greatest need.”

Student Residences

New College had the most students in residence in 2017, holding 900 students with a 901 capacity.

Of the 6,616 residence spaces for students at U of T, 4,017 were occupied by first-year students. University College held the highest number of first-year students relative to its capacity. Besides graduate and family housing, Trinity College held the lowest number of first-year students among the 460 spaces available.

All residences at UTSG were operating at 95 per cent capacity or above in 2017. Chestnut Residence, University College, and Victoria College were all operating at 100 per cent capacity last year.

UTM’s undergraduate housing had a 1,462 student capacity with 642 first-year students. Residences at UTSC housed 754 students of its 767 spaces available, with 613 first-year students in residence.

Wireless connectivity

The university also collects data on the number of connections to U of T’s wireless network across all three campuses. Similar data also shows how students use university-provided web services such as ACORN, including the number of students changing or choosing academic courses, how many students have added bank information, and the number of credit card fee payments that declined.

The average number of connections to U of T’s Wi-Fi per day has doubled since 2013. In 2017, 59,636 unique users accessed U of T’s network per day, with an average of 95,578 devices connecting.

Data released by the Office of Planning and Budget shows a disproportionate overrepresentation of men in both tenured and non-tenured professor positions at U of T. Women make up only around 26 per cent of total full-time tenured and tenure stream Professors — a less than five per cent improvement from data collected in 2007.

This data goes along with an analysis by The Varsity of the Ontario Sunshine List, which showed clear gender pay gaps among the university’s top-paid professors.

In a breakdown by rank and gender, both full-time tenured and non-tenured Professors were overwhelmingly male in 2017.

Of 948 tenured and tenure stream faculty, 26.05 per cent were women. Of 1,091 full-time staff of professor rank, 26.12 per cent were women.

This aligns with historic trends, as the 2007 Facts and Figures book shows the same tenured and tenure stream faculty had 795 professors with 21.26 per cent women — the representation of women from 2007 to 2017 has increased just 4.8 per cent.

Nursing and the Rotman School of Management have the largest disparities in gender balance among tenured faculty. Out of 21 full-time tenured faculty in Nursing, 19 were female and two were male. Conversely, at Rotman, 87 out of 102 total tenured faculty were men.

The faculty with the largest number of tenured faculty, Arts & Science, was 34.32 per cent female among 679 professors.

Assistant Professor and Associate Professor positions have a better gender balance than the higher rank of Professor. Among tenured and tenure stream faculty, 46.64 per cent of Associate Professors and 37.54 per cent of Assistant Professors were women.

University responses

In a statement to The Varsity, Heather Boon, Vice-Provost Faculty and Academic Life, admits that the low percentage of female tenured and non-tenured Professors is due to historical hiring practices. However, Boon expects the number to increase as more women move up the ranks.

Boon also lists a number of initiatives that the university is working on to improve gender equity, including funding from the provost, appointment of a Provost’s Advisor on Women in STEM, and establishing mentorship programs for new faculty.

Also listed is an updated employment equity survey. While the university has committed to demographic surveys in the past, ambiguity still remains around their timeline.

Associate Professor in the Department of Organizational Behaviour and Human Resources Management at UTM Sonia Kang isn’t surprised by these results. Kang explains the systemic issue of gender imbalances where fewer women make it into traditionally male-dominated top ranks in many institutions. Expounding on her research, Kang describes the major hurdle that exists between the rank of ‘Professor’ and ‘Assistant’ or ‘Associate Professor.’

In her research, Kang describes how women have low rates of participation when put into opt-in competitive environments, such as tenure streams. However, when that same choice becomes an opt-out, meaning the decision has to be made to not apply, women reached an equal number of tenure into competitive environments as men.

Whereas tenure at U of T is an opt-out system, promotion to the full rank of Professor is opt-in. Kang suggests that an opt-out system for promotion from “Associate” to full-rank “Professor” could help in U of T’s underrepresentation of women in the higher ranks of academia.

“[Women] tend to display as being more risk-aware so they’re more aware of the risks of certain decisions so they might not take them. It can also be… women are socialized maybe to be less confident, but there’s a whole bunch of different reasons why you might lose women at that juncture.”

While these problems may be conditional to competitive fields like academia, Kang also thinks that there are greater societal issues at play creating large disparities in gender representation.

Editor’s Note (November 26, 6:51 pm): An earlier version of this article misstated Sonia Kang’s title. It is Associate Professor, not Assistant Professor. This article has also been updated to provide additional context on tenure streams at U of T. 

A study released by the School of Graduate Studies revealed that PhD graduates are more likely to end up employed at post-secondary institutions than in all other sectors combined.

The study analyzed the employment status of 88 per cent of the 10,886 PhDs who graduated from U of T from 2000–2015, across all academic divisions. Of those analyzed, 59 per cent attained positions in post-secondary institutions after graduation, while the rest found employment in the private and public sectors, non-profit organizations, or in independent businesses. Only four PhD graduates were unemployed.

However, this trend was not uniform across the board. While close to eight in 10 Humanities PhDs ended up in academia, just less than half of Physical Sciences PhDs chose this path; 40 per cent were employed in the private sector. Of those PhDs who were employed in post-secondary education, about half were in tenure-track positions. Fourteen per cent of graduates were postdoctoral fellows, and approximately four per cent ended up in teaching-stream positions. Many ended up doing research and teaching at U of T, York University, Ryerson University, and McMaster University.

PhDs in the private sector either established their own enterprises or found jobs at Google, Intel, or the Royal Bank of Canada, among others. In the public sector, they mostly worked in hospitals or in the government, and in the charitable sector, they tended to work at health-related non-profits, such as the Ontario Institute for Cancer Research and the Ontario Brain Institute.

The data also revealed significant changes in the composition of the PhD student body. Between 2000 and 2015, the number of U of T PhD graduates per year almost doubled, from 494 to 901, and the amount of non-citizen PhD graduates also doubled. On the other hand, gender composition of the PhD student body over that time has been stable, varying slightly from a 50-50 split. The majority of PhD students — 85 per cent on average — were Canadian citizens and permanent residents.

The recent move by Canadian universities towards releasing demographic data is an encouraging sign and will hopefully promote inclusion on campuses.  

When walking around the University of Toronto, one of the first things a person might notice is the level of diversity. I originally come from Ottawa and I moved to London for my undergraduate degree at Western University. Although my personal experiences in both cities have allowed me to interact with a wide diversity of people and groups, U of T stands apart due to an even more visible, rich multiculturalism.

Diversity sets U of T apart, and it is a quality that most students and staff are likely proud of. Accordingly, it is encouraging that the promise of demographic data being released over the next five years gives us the opportunity to quantify that strength.

The release of this information will also help us identify whether there are certain demographics in particular that are not getting the same level of postgraduate opportunities. This will give us the knowledge needed to determine whether university admissions are claiming diversity for diversity’s sake, or whether the university community accurately represent an inclusive cross-section of Canadian society. Administrations and campus organizations can then adjust their recruitment strategies accordingly.

The fact that this step has been taken by universities across Canada is additionally encouraging. Minority populations in less diverse areas than Toronto will now be empowered with the information needed to point out any inequality they perceive or experience in admission and hiring practices.

Vidhant Pal is a graduate student at the Institute of Biomaterials and Biomedical Engineering. 

Between 2006 and 2016, there have been 26 Killed or Seriously Injured (KSI) incidents inside or on the periphery of U of T’s St. George campus, according to Toronto Police Service data. Of the injuries sustained in these incidents, one was a fatality, two were minimal injuries, and the rest were major injuries. The one fatality occurred at Harbord Street and Spadina Avenue when a bus struck a pedestrian on November 7, 2013.

The most recent incident in the police dataset occurred in October 2016, when a pedestrian was hit by a vehicle on St. George Street.

While four of the incidents involved someone who had been drinking, a plurality of the incidents involved an inattentive individual, often because a pedestrian or driver failed to yield the right of way. A TTC bus was involved in one incident in 2008 when it struck a pedestrian who was crossing without the right of way.

In the age breakdown of those who were injured, 15 of the 28 people who sustained injuries were between the ages of 15 and 29. Seven people were aged 30–49, and six were 50 or older.

Pedestrians were involved in 12 of the incidents, and cyclists were involved in nine others. Of the 26 KSI incidents, two took place on St. George and four took place around the Queen’s Park area.

U of T road safety plans for the future

According to Christine Burke, Director of Campus & Facilities Planning at U of T, the university understands students are concerned about road safety around Queen’s Park and St. George, and it is trying to improve road safety in both of these areas.

“We’ve brought that to the city’s attention and hopefully we can see what kind of recommendations come up to try to improve safety,” said Burke.

Anne Boucher, Vice-President External at the University of Toronto Students’ Union (UTSU) said that “road safety is definitely on the radar” of the UTSU.

“I’ve voiced concerns about a lack of pedestrian crossways and will continue to do so,” said Boucher. “They’ve been quite open to feedback, so I’m confident they’d be open to making some changes if it involves the safety of students.”

U of T is working with the city to implement safety changes in the proposed UTSG Secondary Plan. The proposals in the plan include improving crossings at Queen’s Park as well as making “the whole area between Harbord, St. George, Spadina, and College… much more pedestrian focused and more bike friendly,” according to Burke.

The university also hopes to increase road safety through the Landmark Project, which would pedestrianize the entirety of King’s College Circle.

This November, students on campuses across North America protested in solidarity with Black students who had been targets of racism at universities in the United States. These students presented a set of demands to their respective schools in order to address systemic racism. At U of T, the Black Liberation Collective asked the administration to commit to increasing Black faculty and student representation in order to better reflect Toronto’s population. 

Commendably, our administration has made a small stride in education equity by deciding to collect race-related data on campus. While the exact commitments are unclear at this moment, the broader policy of tracking quantitative information about racial representation in our faculty and student populace is a very welcome step in achieving educational equity.   

As a person of colour, I am acutely aware of the underrepresentation of racialized faculty and students. Some have the privilege of not seeing this reality. It can take less than a semester of classes for someone to note that we have a low number of Black and other racialized students in leadership positions. Without statistical data, however, students have long had trouble proving the problem exists. The collection of race-related data can help bolster causes for equity by simply providing evidence of an issue that students know anecdotally to be true.

It is not hard to find instances in which similar initiatives provided an impetus for positive social change. Back in 2004, for instance, the Toronto District School Board  — the largest school board in Canada — acknowledged the benefits of a census similar to the one our university may implement. They accounted for the race and socio-economic backgrounds of their students in order to  “identify and eliminate systemic barriers to student achievement.” With the data collected, the TDSB was able to quantify a considerable gap between economically marginalized and racialized students. With this information, a problem was identified and a solution more easily and meaningfully created. The TDSB was able to establish the Model Schools for Inner Cities initiative, which seeks to close “the opportunity gap to support equitable outcomes for all students.” 

For those who question the necessity of increasing diversity in university classrooms, it is important to remember several things. First, our society defines itself by principles of equality — as a subset of this, it is important for all persons to have the equal opportunity to attend post-secondary institutions, and it is thus our responsibility to actively investigate barriers that unfairly prevent certain students from succeeding.

Second, while the call to begin this type of census was brought on by students in the Black community, there are universal benefits to solving the problem of underrepresentation. From an equity perspective, diversity means there will be challenges to mainstream narratives, and thus a more wholesome understanding of society. In fact, in 2015, The Atlantic published an article that argued non-white educators “can help disrupt what are often one-sided portrayals of the world and offer invaluable insight to students from different backgrounds.” Comparably, a 2014 Scientific American article extensively traced how “[d]ecades of research by organizational scientists, psychologists, sociologists, economists and demographers” have shown how socially diverse groups enhance innovation and creativity. 

Lastly, race-related data collection at U of T is also an important symbol of administrative accountability. While the university has not yet released the specifics of this initiative, it is clear they are listening to students and engaging in dialogue with them. In doing so, our school is recognizing the legitimacy of student voices — significantly, those that have been historically marginalized by educational institutions — and reaffirming the importance of diversity. We should thus be cautiously optimistic about how this data will serve as a catalyst to create programming that works towards increasing the participation of all visible minorities, in faculty and within undergraduate and graduate student populations.

Milen Melles is a first-year student at Victoria College studying humanities.