13:13:19 So I'm going to go ahead and hit record and then we will get started.
13:13:23 Okay.
13:13:24 So.
13:13:25 We are now recording. So welcome to our second seminar.
13:13:35 So, yeah,
13:13:36 we're geo health network and we wanted to collaborate with canoe to
13:13:39 bring you a series that discussed.
13:13:42 Data options and canoe has a host of data options.
13:13:45 And once we have Caroline join.
13:13:48 I'll be able to.
13:13:55 Carolyn will be able to speak to what canoe is doing,
13:13:58 but in brief,
13:14:00 Let me see here.
13:14:08 So, yeah, so we're offering free workshops and seminars for students.
13:14:11 And mostly our focus is to bridge the gap between health and geography
13:14:15 knowledge.
13:14:16 And we have a few of our executives here today. So we have,
13:14:18 we're not all students anymore.
13:14:20 We have Megan O'Neill who did her master's of public health at Dahl
13:14:24 Alanna.
13:14:25 Megan.
13:14:26 Maybe you can.
13:14:31 Wave at us.
13:14:32 And we have a Naomi Schwartz who just finished her PhD in the
13:14:35 department of geography and planning.
13:14:37 Hi, Naomi.
13:14:38 And.
13:14:40 Yeah.
13:14:48 And I'm in my PhD, in the department of geography and planning at UFT.
13:14:51 And we're so lucky to be here today.
13:15:04 Thanks to donations from the UFT school of cities and the
13:15:07 U of T department of geography and planning.
13:15:10 So thanks to those donations today,
13:15:11 we're giving away one prize of a hundred dollars cash. So yes,
13:15:15 we're just going to mail you a check.
13:15:16 For a hundred dollars and this is open to anyone. Who's a student.
13:15:19 So a college student, undergrad master's or PhD student.
13:15:22 And we're gonna put a link in the chat.
13:15:25 And you can fill out that form and be entered and we're going to
13:15:28 randomly draw a name and we'll get in touch with you on how we can
13:15:32 connect to get you that excellent prize.
13:15:36 So today from canoe,
13:15:38 we have Carolyn and Carolyn did her PhD at U of T with the Southern
13:15:41 Ontario center for atmospheric aerosol research.
13:15:46 And then did her post-doc with canoe in 2018 on air pollution
13:15:49 modeling.
13:15:51 And now Carolyn's with the world health organization and part-time
13:15:53 with canoe.
13:15:54 And Carolyn studies, air pollution, exposure, assessment methods,
13:15:57 and inequalities and environmental exposures.
13:16:00 So Carolyn is an excellent resource and is here to answer.
13:16:03 Any questions you have about canoe and canoe data.
13:16:10 And we are so thrilled to have me on the show here.
13:16:13 Miele is a research assistant with the UBC scent.
13:16:16 UBC faculty of medicine, sorry. In Vancouver.
13:16:27 Miele just finished her.
13:16:28 Master's in geography at UBC in the lab for advanced spatial analysis,
13:16:32 studying the spatial distribution and environmental determinants of
13:16:35 pediatric inflammatory bowel disease.
13:16:44 So meal's going to go ahead and tell us about her excellent project
13:16:47 using canoe data and administrative health records,
13:16:50 which I think is a really important learning
13:16:53 opportunity. And right after her presentation,
13:16:56 we're going to have a Q and a.
13:16:58 And then we're going to do breakout sessions by interest group.
13:17:00 And you're going to have a chance to meet some other students that are
13:17:03 interested in the same health geography topics as you,
13:17:05 and maybe work on a little question together.
13:17:08 And, and then we'll wrap up.
13:17:10 So I will turn it over to you.
13:17:14 Yeah.
13:17:15 Well, so you should be able to share.
13:17:17 Excellent.
13:17:18 Well, thank you so much for that introduction.
13:17:19 Excited to be here.
13:17:21 I'm just going to share my slides real quick.
13:17:26 Here's the introductions.
13:17:29 We probably should have had a couple of minutes ago, but.
13:17:31 Here's the information now.
13:17:38 So I'm going to be talking to you today about my master's thesis.
13:17:43 Rather than going into all of the nitty-gritty about my results,
13:17:45 conclusions, that kind of a thing.
13:17:55 All that is available in exhaustive detail in my thesis.
13:17:58 So I'm going to talk a bit more about how I selected my methods.
13:18:00 I'm going to do a little bit of a live coding demonstration to show
13:18:03 kind of what I do and my methods.
13:18:11 I tried to keep things a little bit more interesting,
13:18:13 but before I begin,
13:18:14 I do want to say an acknowledge that I'm from occupied Duwamish lands,
13:18:16 which is now Seattle.
13:18:18 And I'm joining you today from the traditional ancestral and unceded
13:18:20 territory of the muscular in Squamish and Tsleil-Waututh peoples,
13:18:22 which is South Vancouver.
13:18:23 A brief agenda for Shea.
13:18:24 So I'm going to talk a bit about my education and my career
13:18:26 background.
13:18:27 And then my research and my workflow I'll show some of the coding that
13:18:30 I use to produce some really simple but effective interactive maps.
13:18:38 I'll take it back and talk about kind of big picture results and
13:18:40 lessons that I've learned during my research.
13:18:42 And then we'll have plenty of time for a Q and a after that.
13:18:44 And then I learned mentioned we're going to do some breakout groups
13:18:47 with some topics of discussion,
13:18:48 and then hopefully we'll have time to come back and share a little
13:18:50 bit.
13:18:51 From our breakout groups.
13:18:52 So first my background, I have a BA in human geography from UBC.
13:18:57 My focus was on GIS and cartography, which was at the time,
13:18:59 not an official major at UPC.
13:19:01 I just took all of the GIS photography and some stats courses that I
13:19:04 could get my hands on, on a few remote sensing courses as well.
13:19:16 And that I immediately transitioned into my physical geography, MSC.
13:19:19 Physical geography is a bit of a misnomer because really my area of
13:19:22 specialty would be more considered medical geography, epidemiology,
13:19:25 environmental health,
13:19:26 but at UVC there's either a master's in human geography or in physical
13:19:28 geography. So I was in the physical geography stream.
13:19:31 And then during my BA and my masters,
13:19:33 I was doing some kind of side freelance work.
13:19:35 JS photography data work.
13:19:37 So I worked for the recent campus and community planning group,
13:19:39 and I did some work for PhD student in the school of community and
13:19:42 regional planning at UPC as well.
13:19:45 And these weren't necessarily directly related to my thesis or my
13:19:47 research,
13:19:48 but it gave me a lot of experience working with all kinds of GIS data
13:19:51 analysis, visualization.
13:19:52 So I found those experiences really helpful in shaping my work,
13:19:54 even though they weren't in the same.
13:19:56 Necessarily stream or topic as my work.
13:19:57 And then now I'm a research assistant with the UC faculty of medicine
13:20:00 at BC children's hospital.
13:20:02 So a brief overview of inflammatory bowel diseases,
13:20:04 which were not something that I knew anything about before I started
13:20:06 this research about two years, two and a half years ago.
13:20:08 Many of you may have heard of Crohn's disease or ulcerative colitis,
13:20:11 and those are the two main inflammatory bowel diseases.
13:20:26 It's essentially a set of diseases that are characterized by chronic
13:20:29 inflammation of the gastrointestinal tract.
13:20:31 And they are thought to arise from a complex interaction of genetic
13:20:34 and environmental factors, which can include diet, lifestyle,
13:20:36 and many, many aspects of the physical environment.
13:20:38 So they're very complex chronic illnesses with multiple causes.
13:20:43 A brief overview of the spatial spread of IBD.
13:20:46 So it's more common generally in industrialized countries,
13:20:49 in urban areas and areas with higher socioeconomic status.
13:20:51 So it is kind of a rare disorder.
13:20:53 That's actually more common in middle or upper middle class families.
13:20:56 Canada actually has some of the highest pediatric incidence of IBD in
13:20:58 the world.
13:21:00 So for my master's research.
13:21:01 I kind of wanted to do a spatial component because I am a geographer.
13:21:10 And there's been very little mapping work done of IPT.
13:21:13 And I think in general, there's more happening now,
13:21:15 but chronic illnesses maybe were a little bit later to the party of
13:21:18 spatial analysis.
13:21:19 And now there's a lot more spatial work in chronic illnesses.
13:21:21 In addition to infectious diseases.
13:21:22 So my first research objective was to identify and describe spatial
13:21:25 temporal patterns in pediatric IBD in BC.
13:21:35 My second objective was to model associations between pediatric IBD
13:21:38 and air pollution in the form of nitrogen dioxide, residential green,
13:21:41 miss vitamin D adjusted solar UV radiation.
13:21:44 So essentially the vitamin D dose that you would get from exposure to
13:21:46 solar radiation in that location.
13:21:48 And also some area ethnicity.
13:21:49 So to do this.
13:21:52 I did a spatial cluster analysis using more ans I,
13:21:55 so if you're not a medical geographer or a GIS person,
13:21:58 that's kind of the, one of the bread and butter spatial statistics of.
13:22:00 Geography.
13:22:01 I used that as a local indicator of spatial association,
13:22:03 although a little bit into that more later.
13:22:05 Essentially, I was trying to detect disease.
13:22:06 Clusters of high incidents were low incidents in the province.
13:22:14 And then I wanted to model associations between my environmental
13:22:17 determinants and IBD using case control study.
13:22:19 So comparing cases with IBD to controls who did not have IBD.
13:22:22 So my data, I was very lucky to get data from BC children's hospital.
13:22:34 So I partnered with Dr.
13:22:35 Kevin Jacobson and the division of gastroenterology there,
13:22:37 and they maintain a register of patients who are diagnosed with IBD in
13:22:40 the province, pediatric patients.
13:22:42 So my cases were patients who had IBD.
13:22:43 My controls had received the diagnostic test for IBD and were
13:22:46 confirmed to not have IBD.
13:22:47 And I matched them with my cases based on the age.
13:23:12 So for my environment data I got from canoe,
13:23:14 which if you attended the talk in February, you know,
13:23:16 a bit more about, but can you is an amazing,
13:23:18 amazing resource of environmental exposure through data for health
13:23:20 research. So I got air pollution, UV, vitamin D green,
13:23:23 this measuring what the normalized difference, vegetation index,
13:23:26 and then material and social deprivation and disease.
13:23:27 As I control variables,
13:23:29 all of this one could do at the six digit postal code level ready to
13:23:31 use right out of the gate, which is incredible.
13:23:33 And then from the Canadian census at the dissemination area level,
13:23:36 I got South Asian ethnicity, Jewish ethnicity,
13:23:38 and area indigenous identity.
13:23:40 And these are from the UFT computing and the humanities and social
13:23:42 sciences.
13:23:43 So for my workflow,
13:23:44 if you've worked with health data before you know that the ethics
13:23:46 application.
13:23:47 Process can be quite lengthy.
13:23:49 I was lucky. Mine only took about three months,
13:23:50 but this is something to certainly budget for.
13:23:52 When you're working on a project,
13:23:53 it can take quite a while to get ethics approval.
13:23:55 And I did have to do several rounds of revisions of my ethics
13:23:57 application.
13:23:58 So essentially what happens is I've sent them an application with what
13:24:00 I wanted to do. All of the methods for data security, privacy.
13:24:03 And then the analysis methods I was going to use.
13:24:05 And then they sent it back,
13:24:06 essentially asking me if I was willing to aggregate my data and
13:24:08 abandoned some of the spatial aspects of my project.
13:24:12 Because when you work at the six digit postal code level insidious,
13:24:15 that's essentially a city block.
13:24:17 So it is quite a specific measure of spatial location.
13:24:19 Essentially do know where the patients live,
13:24:21 or at least the general area where they live.
13:24:22 So generally to protect privacy, they don't like to approve.
13:24:24 Projects that involve such close work with patient location.
13:24:31 However,
13:24:32 I was able to argue with them and justify that it was really important
13:24:36 to the research that I had, that really specific location data,
13:24:38 both for accurate spatial analysis and for assigning exposures
13:24:41 accurately.
13:24:45 But argue is kind of a misnomer. They were very willing.
13:24:47 The research ethics board is really willing to work with me.
13:24:49 I just had to justify to them and show them the methods that I was
13:24:51 going to use to make sure that patient privacy is going to be
13:24:53 protected as much as possible without losing the benefit to the
13:24:56 patient.
13:24:57 And then once I had all of my data in hand,
13:24:59 All of the rest of my project essentially took place in the,
13:25:01 our environment. And I'm going to show you that in a second,
13:25:03 kind of what it looks like.
13:25:04 So cleaning, joining all of my data together,
13:25:06 analysis and visualization. I do all of that in AR.
13:25:08 So now I'm going to share.
13:25:11 Heart with you and show you how to make a couple of quick,
13:25:13 easy interactive maps.
13:25:18 Is that working for, if you could just give me a thumbs up,
13:25:20 if you can now see our, that would be very helpful. Excellent. Okay.
13:25:24 So here is our studio, which is a very common.
13:25:27 Free open-source statistical software.
13:25:37 And all of the code you see on the screen,
13:25:38 this has everything to make an interactive map of canoe data.
13:25:40 That's all the code I used. So the only thing that I've done,
13:25:42 I downloaded the commute data and I did.
13:25:48 Clip it to restrict it just to the area of Vancouver so that it will
13:25:50 load a little bit faster,
13:25:51 but otherwise this has canoe data straight out of the box.
13:25:53 I haven't done anything to it.
13:25:54 There's not any secret code that I used beforehand to set things up.
13:25:56 So I'm going to load by packages, read in my canoe data.
13:26:00 And make sure everything is joined together.
13:26:01 And then here's code. I'm just going to create oops.
13:26:04 I'm going to create a spatial object here.
13:26:06 And then this is the team app package, which I really like.
13:26:09 Which makes it really easy to make interactive maps.
13:26:17 So here is the code for interactive habits. Just those four lines.
13:26:21 It's going to load here on the right. It might take a second.
13:26:29 So now I have an interactive map of the canoe data. So on the right,
13:26:32 I have nitrogen dioxide, air pollution.
13:26:34 If you've ever been to Vancouver, this is Stanley park right here.
13:26:36 This is downtown Vancouver. I'll just do me a little bit.
13:26:43 So what's great is you can see these multiple screens.
13:26:45 It's gonna zoom and show the cursor at all the mass at the same time.
13:26:48 So it's great for comparing multiple areas.
13:26:50 As you can see here in downtown Vancouver.
13:26:51 We have this corridor here, a very low greenness,
13:26:53 kind of in the heart of downtown when the densest urban area there.
13:26:56 And you can see that's quite a high area of mentioned backs that air
13:26:59 pollution, which makes sense is.
13:27:00 It's mostly a traffic associated pollutant.
13:27:03 So, this is a super, super simple, quick way.
13:27:05 I use this all the time for analysis,
13:27:06 or just when I first get my hands on some data,
13:27:08 just to kind of see what I'm working with.
13:27:09 I like to use this method.
13:27:19 And then I was trying to think of a way to show you how to do some of
13:27:21 the, the GIS cluster mapping that I was doing. And unfortunately,
13:27:24 I can't show you using my actual data because it's all private and I
13:27:28 don't want to get fired. So.
13:27:36 What I did instead was took publicly available COVID data at the
13:27:39 health service delivery area level in BC.
13:27:41 So this is released by the BC government.
13:27:42 You can all go to their website and download this and work with it.
13:27:44 If you want to it's available. I shave files and a spatially,
13:27:47 and I've already loaded some of this just because it takes a minute to
13:27:49 load.
13:27:50 But I'm just going to read it all of my COVID incidence data and join
13:27:53 it to the spatial data.
13:27:54 And I'm going to show you an interactive map of COVID incidents in BC.
13:27:56 So this is average daily incidents of COVID per a hundred thousand
13:27:59 people in March of 2021.
13:28:01 And it's just going to take a second to load here.
13:28:03 So this package team app, you can get as fancy as you want with it,
13:28:05 but this is, this is kind of the boiler plate quickest.
13:28:07 Preset method.
13:28:08 So here we have here's BC.
13:28:11 If you're not very familiar with BC,
13:28:12 this is Vancouver down at the bottom here.
13:28:15 So most of the people of the province live like right here.
13:28:17 In the Metro Vancouver area.
13:28:22 But as you can see, we have kind of higher incidents around BC,
13:28:25 South of BC.
13:28:26 I'm at a Northern, a little bit in Northern BC as well.
13:28:39 So then this is very useful to see,
13:28:40 but I wanted to actually apply the method I used in my thesis to get
13:28:43 statistically significant clusters of high incidents or low incidents
13:28:45 and see what kind of spatial patterning we're working with.
13:28:47 And if we can quantify that.
13:28:48 So I'm going to run.
13:28:49 Some things right here and then I'll show you what I get. So.
13:28:51 We can see.
13:28:54 If you don't understand Ram.
13:29:07 XY or a lot of spatial statistics. That's totally fine. Essentially,
13:29:10 what we're seeing is we have sort of moderate spatial clustering,
13:29:13 so not a high level of spatial clustering,
13:29:15 but there is some detectable,
13:29:16 spatial clustering going on in this dataset.
13:29:19 By the way, this is not, this is not my code.
13:29:21 I can put the link in the chat.
13:29:22 I think for anyone who wants to see the tutorial that I got this from,
13:29:24 I didn't write this code from scratch myself.
13:29:26 I adopted it for what I was doing.
13:29:29 And then I'm going to run the local rants, I,
13:29:30 which is going to allow me to visualize the clusters are hot spots and
13:29:34 cold spots of disease.
13:29:36 And you have to kind of manually set these clusters.
13:29:37 I'm going a little faster, so that makes sure we stay on time here.
13:29:40 And then I'm going to map it.
13:29:44 So, unfortunately there aren't actually any, unfortunately I.
13:29:48 It is what it is,
13:29:49 but there aren't any statistically significant clusters in this
13:29:52 dataset.
13:29:53 So I'm just going to map everything.
13:29:54 That all the clusters aren't statistically significant.
13:29:58 But we can at least get an idea of how things are looking.
13:30:00 So in this map, on the right.
13:30:10 Here's BC. Again, what we have is in red areas, we have high,
13:30:13 high clusters. So that's areas of high COVID relatively high.
13:30:16 Not necessarily, there's not an objective measure of high necessarily,
13:30:18 but.
13:30:19 Relatively high for the province.
13:30:21 COVID incidents located near other areas of high COVID incidents.
13:30:24 So you can see.
13:30:25 Kind of Northern BC and then in the lower mainland as well,
13:30:27 for the most part.
13:30:28 Around Vancouver.
13:30:29 We have high instance look at in your high incidents.
13:30:37 This light blue color is a low high cluster.
13:30:39 So it's a spatial outlier where we have an area of low incidents.
13:30:42 That's much lower than this. Most of the surrounding areas, incidents.
13:30:45 So we have in central BC, we have this low cluster.
13:30:47 And then here, this blue color is low, low.
13:30:49 So this means areas of lower incidents surrounded by low incidents.
13:30:51 So we have out in the Okanagan.
13:30:52 I'll use if you've been out East and BC,
13:30:54 and then we have been Cooper Island here as well.
13:30:55 In March, their average daily COVID incidence is on the lower end,
13:30:58 located near areas that are also on the lower end.
13:31:08 So I'll put that link in the chat.
13:31:10 So you can go to the tutorial that has a bunch of other additional
13:31:12 information as well, but this is just kind of a basic idea.
13:31:14 This is what I use for my master's thesis.
13:31:16 And what's really amazing about this workflow is that.
13:31:18 I wrote a lot of these codes during my thesis.
13:31:20 And during my.
13:31:36 Freelance work. And then now when I'm working at children's hospital,
13:31:38 I have all of these codes that I've already developed, so I can use,
13:31:41 I can reuse them and go back.
13:31:42 And what's really useful for me is that if I created the data set a
13:31:44 year ago,
13:31:45 I can actually go back to the code and see exactly the specifications
13:31:48 of how I created it so I can reuse it.
13:31:49 And also I kind of have a built-in documentation to everything that
13:31:52 I'm doing.
13:31:53 So really, really quick.
13:31:54 Essentially what's happening at my actual research is for IBD.
13:31:58 This is, this is Vancouver right here.
13:32:00 Blue is lower than average incidents.
13:32:01 The darker Brown is higher than average incidents.
13:32:03 So we have higher incidents here in Delta and Surrey,
13:32:06 lower incidents here in Vancouver.
13:32:11 And then these lighter areas are close to the provincial average.
13:32:13 So when I run the actual hotspot and cold spot detection method,
13:32:16 you can see,
13:32:17 I have a statistically significant hotspot here in the province,
13:32:21 and that's using exactly the technique that I just showed you in B
13:32:24 interactive mapping.
13:32:25 So I'll just go through this very, very briefly.
13:32:33 When I was modeling the environmental determinants using canoe data,
13:32:36 I essentially wanted to compare cases.
13:32:38 So people who are diagnosed with IBD, with controls,
13:32:40 who were age matched to patients without IBD on their exposure.
13:32:43 So I wanted to see if there's a difference between people who have IBD
13:32:45 and similar people who don't have IBD.
13:32:47 Based on their exposure to air pollution, UV, freeness,
13:32:50 and then area in the city.
13:32:51 Which we're using as a proxy.
13:32:53 Individual ethnicity, which is not ideal,
13:32:54 but we don't have patient individual ethnicity.
13:32:56 So that's kind of the best that we could do.
13:32:58 And so I fit, I fit models, essentially using some control variables,
13:33:00 and then he just put my variables of interest.
13:33:01 Very basic logistic regression modeling also.
13:33:04 In our.
13:33:06 So from this, you can see on the right here, SBC,
13:33:08 here's the full results of my cluster detection. So we have.
13:33:10 The blue is cold spots. So areas of low.
13:33:19 Disease incidents. So in Northern BC,
13:33:21 and then also in Southeastern BC here, some of this cold spot,
13:33:24 maybe due to patients that are going to Calgary to receive care,
13:33:26 that definitely does happen.
13:33:28 Our dataset is fairly complete.
13:33:29 What we probably are missing a few people.
13:33:30 And then we have a couple hotspots on Vancouver Island.
13:33:32 This is Karen Mayo is kind of in central Oakenoggen BC.
13:33:34 And then here's Vancouver.
13:33:50 So we found a possible lower Crohn's disease risk in areas with a
13:33:53 higher indigenous population,
13:33:54 which has been found by other researchers as well in Canada.
13:33:57 And all of the associations that I'm finding were quite moderate.
13:33:59 They're very small numbers, which is what I was expecting.
13:34:01 Most epidemiological research on IBD. You end up with fairly small.
13:34:04 Small effect sizes because it's such a complex disease.
13:34:06 There are so many different causes.
13:34:07 So it's not necessarily disheartening to find a small effect.
13:34:09 I wasn't expecting to find.
13:34:10 Like the key to the whole disease that no one's ever ever looked at
13:34:13 before.
13:34:28 And then I did also see a higher incidence in lower mainland,
13:34:30 but actually not in Vancouver, which is interesting because I,
13:34:33 B D is generally associated with urban areas.
13:34:35 So a lot of people consider it an urban disease.
13:34:37 And so I would expect that the densest,
13:34:38 urban areas of the province would probably have the highest IVD rates,
13:34:41 but that's not true. Nowhere in Vancouver,
13:34:42 at least during my study period had IVD.
13:34:44 So I think that's very interesting. That's a little bit different.
13:34:46 That a lot of the literature.
13:34:47 And there's very little literature,
13:34:48 basically none that distinguishes between different types of urban
13:34:51 areas. So basically.
13:34:52 A lot of more clinical researchers are thinking about urban versus
13:34:55 rural. This is that dichotomous category.
13:34:57 There's no differences between different urban areas or different
13:34:59 rural areas that might be important.
13:35:00 So I think this is a really key area for future research.
13:35:03 So my key takeaways from the research process.
13:35:06 So very nitty gritty things.
13:35:07 If you're kind of in the thick of your thesis or other kinds of work
13:35:10 right now, these might be more helpful to you.
13:35:12 Having a reproducible adoptable workflow.
13:35:20 Like I did was so, so helpful. It's so helpful for me now,
13:35:23 professionally in my career to have those codes that I developed,
13:35:25 it was helpful at the time too.
13:35:26 It's a lot easier to fix mistakes than manually going through and
13:35:29 setting everything again, if there's a problem.
13:35:31 So I found that to be really, really helpful.
13:35:32 Definitely at the time, it took me to learn how to,
13:35:34 how to create this workflow. I saved a lot more time later in the end,
13:35:37 by having invested that time at the beginning to develop it.
13:35:39 Control selection is very important. So if you're comparing presence,
13:35:41 so the presence of diseased absences, the disease.
13:35:43 Who you compare your patients to is really important. That's really,
13:35:46 that really determines what results you're going to find. So.
13:35:56 Our, because our controls are people that didn't have IBD,
13:36:00 they were great controls,
13:36:01 and that they're really confirmed to not have the disease.
13:36:03 They have a high level diagnostic procedure to make sure that they
13:36:06 don't have the disease. So in that sense,
13:36:07 they're really excellent control.
13:36:09 But they had to have some kind of health or GI issue to prompt that
13:36:13 scope in the first place. So in a sense,
13:36:15 they're not necessarily healthy controls.
13:36:16 So you could argue that they're not,
13:36:17 they're not necessarily the best possible controls for this study.
13:36:20 And that could have also contributed if,
13:36:21 if the health issues or disease that they ended up having,
13:36:24 but it's not IBD.
13:36:33 It works through similar pathways, diabetes.
13:36:35 So if there are similar things in the environment that are also
13:36:37 causing whatever health issues they have, they're not going to,
13:36:39 we're not going to be able to see an effect that differentiates them
13:36:41 from people with IBD.
13:36:42 So that's something to keep in mind control section was really,
13:36:44 really important or whatever kind of absence data you're looking at is
13:36:46 really, really important.
13:36:48 Also having clean data.
13:36:49 So like the data that I got from canoe saved me months or years of
13:36:52 work, essentially, there's no way.
13:37:01 It would have been able to do this project in the two years that I had
13:37:03 for a master's degree,
13:37:04 if I didn't have all of this high quality exposure assessment data are
13:37:07 already given to me at the level of analysis that I needed,
13:37:09 like any one of those variables to create as once the work and the
13:37:11 fact that could be when other researchers have done that and made it
13:37:13 available,
13:37:14 saved me so much time and made this project so much more feasible on
13:37:17 the timeline that I had.
13:37:18 And then from a kind of bigger picture of something that I didn't
13:37:20 really consider going into this project was research ethics,
13:37:23 which is of course, very, very important.
13:37:24 It's sort of the key that allows us to use health data safely and
13:37:28 protect patients to make sure that their privacy is not being
13:37:30 violated. That there.
13:37:36 That, you know,
13:37:38 there really is a benefit to the small risks to the patient by using
13:37:41 their private information.
13:37:42 But the approval process can be very lengthy.
13:37:44 There's definitely a level of scrutiny to your research that if you're
13:37:47 in other disciplines that you don't have to do ethics approval,
13:37:49 you might not have.
13:37:50 I think that can be good because you really have to think about.
13:37:57 I'd be able to articulate the benefits of your research very on in the
13:38:00 project. So you can't just say, I want to try this out.
13:38:01 I'm not sure why this would be helpful. You have to say,
13:38:03 here's why this is useful. Here could be the benefits to the patients.
13:38:06 So it's a level of scrutiny that I think can produce stronger research
13:38:08 because you really have to be able to talk about why,
13:38:09 what you're doing is important and justify that to an outside person
13:38:12 who might not be really familiar with your work.
13:38:14 So my big kind of takeaways.
13:38:18 From my research process and from working as a medical geographer
13:38:21 health geographer in this field for a couple of years now,
13:38:23 and this is just for my own experience,
13:38:24 but I think a lot of you might be able to relate,
13:38:26 even if you're in just any interdisciplinary field or if you work in a
13:38:28 job where you wear a lot of different hats,
13:38:30 you have to talk to people who have different specialties.
13:38:33 You're often going to be the branch between a lot of different
13:38:35 disciplines. So I have to know about disease biology.
13:38:37 You have to know about geography.
13:38:39 I have to know about epidemiologists,
13:38:40 have to know about public health.
13:38:41 I have to know about environmental exposure and remote sensing,
13:38:43 all these different things.
13:38:44 Together.
13:38:45 And then be able to explain.
13:38:46 What I know about one thing or what other members of my research group
13:38:49 know about one thing to different people in the research group.
13:38:51 So I have to be able to explain.
13:38:59 The value of geography and the work that I'm doing and spatial
13:39:01 analysis to an audience who's not familiar to that. And vice versa.
13:39:04 I have to be able to explain,
13:39:05 we need to have a certain variable this way to a geographer,
13:39:07 because it's really important that because of the disease biology,
13:39:10 it has to be this certain way.
13:39:11 So in my experience,
13:39:13 a lot of more clinical researchers or healthcare providers,
13:39:15 don't have a lot of exposure to mapping and spatial analysis,
13:39:18 but they're really interested in it.
13:39:19 So it's really important for me to be able to explain the methods that
13:39:22 I'm using and their strengths and weaknesses,
13:39:24 and also the data and computing needs that I have that are very
13:39:26 specific to geographic data.
13:39:31 Two members of the clinical research team in a way that they'll
13:39:33 understand to make sure that we're all on the same page about what I'm
13:39:37 actually analyzing and that that's what everybody wants and needs.
13:39:39 And that I have the computing and data that I need to make that
13:39:41 happen.
13:40:01 Sort of related to that. So I know for me,
13:40:03 I feel out of my depth a lot of times,
13:40:05 because I'm not a super high level expert in one thing. So I think my,
13:40:09 my role and my skill is really making connections between a lot of
13:40:12 things. So breadth of knowledge,
13:40:13 rather than depth of knowledge is my value to my team that I know I do
13:40:16 know about spatial analysis and that's my area, but there's,
13:40:18 there's always gonna be,
13:40:19 I'll be able to find a medical geographer that knows more than me,
13:40:21 but I maybe know more about IBD and disease biology than someone who's
13:40:25 a.
13:40:26 Pure medical geographer.
13:40:27 So really it's easy to feel like to appeal Berry.
13:40:30 Out of your league,
13:40:31 I guess when you're in a room with all these people who are experts in
13:40:33 their specific thing,
13:40:34 but being able to connect all of these fields is absolutely its own
13:40:36 skill. Not everyone can do it.
13:40:38 So I'm trying to be confident in that and take that seriously as your
13:40:40 role, like your role is to connect all of these things.
13:40:42 And then finally this is sort of related to research ethics and just
13:40:45 having good re.
13:40:46 Good working relationships in general, but investing in your,
13:40:49 in your professional relationships is really important when you're
13:40:51 working in a field. That's so trust-based because people are,
13:40:54 don't want to just hand you health data. If they don't trust you,
13:40:56 if they don't.
13:40:57 Feel like you are taking it seriously and you're going to do a good
13:40:59 job.
13:41:00 So definitely investing in you in those research relationships and
13:41:03 making sure that you're building trust with the people that you're
13:41:05 collaborating with, I think is,
13:41:06 is really helpful for successful working relationships and success in
13:41:09 your career. So those we think of as kind of soft skills, but.
13:41:12 Being personable and friendly.
13:41:14 Trustworthy, I guess, with this,
13:41:15 with this really important and private data.
13:41:17 Is really important.
13:41:18 Okay, so that is all I know. That was a crammed it out a little bit.
13:41:21 Cause we started a little bit late here,
13:41:23 but hopefully you are able to absorb,
13:41:24 everyone could absorb at least something that was interesting to them.
13:41:26 And now I think we're going to open it up.
13:41:27 We're going to do a brief Q and a,
13:41:29 and then we're going to have some breakout rooms where you can all
13:41:31 talk to each other and meet each other.
13:41:32 And actually helped me with a question that we have in my field that
13:41:35 we're grappling with right now.
13:41:36 Thank you so much.
13:41:42 In a fabulous project. And thank you for showing us.
13:41:45 Maybe you can drop the link to the R package. Yes, I will do that.
13:41:51 In the chat. So if you have questions for me,
13:41:53 how can you use the Q and a button at the bottom of your
13:41:57 screen?
13:41:58 If you have questions about.
13:42:03 Geography or the health side,
13:42:06 the project health data ethics are, whatever it is. Go ahead.
13:42:10 And.
13:42:21 Our breakout rooms.
13:42:23 Weren't enabled part of our technical difficulties before the call.
13:42:26 So we won't be able to do that, unfortunately,
13:42:28 but maybe we can try to do that here together.
13:42:31 Yeah, absolutely.
13:42:33 Does sorry about that.
13:42:34 But does anyone have any questions for me all about her master's
13:42:37 thesis project?
13:42:40 I'm going to, sorry, I'm just going to send this link to everybody.
13:42:42 Thanks for the reminder.
13:42:43 Here we go.
13:42:45 Perfect. Thank you so much.
13:42:57 Neil, maybe you can put up on the screen again,
13:42:59 your cluster map. Yes, absolutely. That's a great idea.
13:43:05 What is the, the, our map or the, the hotspot map.
13:43:09 What do you think is.
13:43:15 Going on in the Okanogan Valley.
13:43:18 And then another question I had for you is.
13:43:20 Do you talk much in your lab?
13:43:25 About the health services or the primary care use with these
13:43:29 patients?
13:43:31 Access.
13:43:37 Yes, absolutely. Okay. So I'm hearing two questions.
13:43:39 So Okanagan Valley, I'm just writing this down.
13:43:40 Oakenoggen Valley one question and then kind of primary care use.
13:43:54 So Okanogan Valley, you know, I'm not, I'm not totally sure.
13:43:58 I think the pollution exposure there is a little bit higher,
13:43:59 but we actually didn't find many results for pollution exposure and
13:44:02 the results from other studies about air pollution and IBD are very
13:44:05 mixed. So I think there might be something about.
13:44:08 Urban environments there.
13:44:09 It's really hard to say I'm in our next phase of research,
13:44:11 we're trying to kind of tease out what it is about urban environments
13:44:14 or some of these high.
13:44:16 Higher incidence areas. So really I'm not sure.
13:44:18 And I'd actually love to hear any of your thoughts too.
13:44:20 If you want to put it in the chat.
13:44:21 If you have an idea of what could be happening in the Oakenoggen,
13:44:23 that might be kind of triggering more pediatric IBD.
13:44:26 I'd love to hear it.
13:44:36 A second question about primary care. So yes, I work with,
13:44:39 there's not that many pediatric gastroenterologists in the province.
13:44:42 And if you get diagnosed with IBD as a child in BC,
13:44:45 you pretty much have to come to Vancouver.
13:44:46 So that is a huge issue for diagnosis and for receiving care.
13:44:50 And that is definitely on the minds of the clinicians that I work
13:44:53 with. They want to make sure that all.
13:44:55 All of the patients in the province are getting diagnosed and are
13:44:56 getting the continuous care that they need.
13:44:58 And if you live in a rural area, and this is not unique to IBD.
13:45:00 Unfortunately in Canada,
13:45:01 it's a lot harder to access care to make that journey,
13:45:03 to get the tests and to get the procedures done that you need.
13:45:06 And IBD requires depending on the severity of disease. It's chronic,
13:45:09 it's like law. There's no cure.
13:45:11 It kind of has peaks and valleys and you can require quite a lot of
13:45:14 care, including surgery and a lot of medical management as well.
13:45:26 So I think that is definitely something that's on care provider's
13:45:28 minds, as well as the transition from pediatric to adult care,
13:45:32 because there is,
13:45:33 there are more adult gastroenterology resources in different areas of
13:45:36 the province than there are pediatric.
13:45:37 But no, that definitely is.
13:45:38 I'm not sure of all of the strategies that they're using and I think
13:45:40 there's always room for improvement, but that absolutely is.
13:45:44 Is a priority for care providers here that I know is making sure,
13:45:47 trying to find ways to make sure that rural patients do have a
13:45:50 comparable quality of care to patients that are living in Vancouver or
13:45:52 the Vancouver area.
13:45:54 That's so interesting. Yeah.
13:45:55 I think sounds like there's a whole nother project there for you to do
13:45:59 on.
13:46:00 The rural split.
13:46:01 And.
13:46:11 Then there's a,
13:46:12 there's a lot of conversations by the health authorities around.
13:46:15 Do you,
13:46:16 do you ask people to travel or is there a project to be done
13:46:20 around?
13:46:22 Re resourcing a specialist to go and do locums. Like, do you,
13:46:26 do you make patients travel or do you have the,
13:46:28 the care provider travel?
13:46:30 Dominica has a question.
13:46:32 So I'm going to take you off mute and you can go ahead.
13:46:34 Hello.
13:46:37 Hi. Hello. Hi. Sorry. I just.
13:46:47 Was difficult to write it out. So I'm sorry.
13:46:49 I missed the first 20 minutes.
13:46:50 So I may be asking a question that you've already answered in your
13:46:53 model.
13:46:54 Other than the nitrogen dioxide exposure. Did you include any.
13:46:58 Any information about water quality.
13:47:02 Or.
13:47:03 Food.
13:47:04 Nutrition diet something to that effect.
13:47:06 To account for.
13:47:07 The possible risk.
13:47:15 Of those in higher incidence rates in what looks
13:47:19 like to be serene Delta, as well as.
13:47:21 The freezer.
13:47:22 Valley area.
13:47:23 That's a great question.
13:47:31 So the short answer is no,
13:47:32 I didn't have diet is a huge factor that was missing. We definitely,
13:47:35 I can actually,
13:47:37 I'll go back to the slide where I have the bottles that I ran. So.
13:47:39 When I set out for this research,
13:47:40 I knew I was not going to be able to comprehensively model risk.
13:47:55 So actually I didn't have models where I combined all of my variables
13:47:57 of interest,
13:47:58 even because I knew I was never going to get a full picture of risk or
13:48:01 even approaching it because I, as you said, diet is very important.
13:48:04 I don't have diet. I don't have things like water, quality,
13:48:06 family history of IBD.
13:48:07 I don't have the patient's gender or their patient's ethnicity,
13:48:09 which are also important.
13:48:16 So,
13:48:18 because I knew I wasn't going to be bottling risk in a comprehensive
13:48:20 sense. I wasn't trying to say, like, here's my best predictive model.
13:48:23 It was more just to try to quantify the specific variables that I was
13:48:27 looking at, but yes, you're absolutely right.
13:48:28 It's definitely missing some important things.
13:48:30 And there is a lot of research on diet and IBD.
13:48:32 So I'm hoping that research group that I'm with.
13:48:34 It's kind of has a lot of chemicals.
13:48:36 There are people in diet. There are people in clinical research.
13:48:39 There are people in microbes.
13:48:40 Microbiome research, which is what I originally wanted to get into.
13:48:42 Yeah. So yes, there are. We're hoping in the future,
13:48:44 we're going to be able to combine a lot more kind of arms of the study
13:48:47 with the environmental exposure to try to build a better model
13:48:49 essentially of risk.
13:48:51 Yeah, because my feeling is, although the air pollutant.
13:48:53 Of nitrogen dioxide is very important in a lot of disease.
13:48:58 With a disease like IBD,
13:48:59 I'd be more concerned about the route of exposure, which is oral.
13:49:03 So that's why I'm wondering if you, if you had your hands on any.
13:49:06 A pollutant that is measured in the water.
13:49:09 If there's a data source for that.
13:49:15 I think that might be a strong contributor. Like for example,
13:49:18 if there's measurements on a, I don't know,
13:49:21 organophosphates or other sort of pesticides in the water.
13:49:24 That might be a good.
13:49:26 A good variable to add to your model.
13:49:28 No, that's a, that's a really great point.
13:49:29 And I'm hoping that kind of more data like that.
13:49:35 It's harder to do kind of national exposure metrics that are specific
13:49:38 to certain water systems.
13:49:39 So I see why that data is kind of limited right now.
13:49:51 But no, I think you're, you're totally right.
13:49:52 That kind of like more specific patient exposure is definitely a
13:49:55 direction that I want to go in.
13:49:56 I think we actually have someone from the child study on the call.
13:49:58 So something like that, where we're looking at microbiome,
13:50:01 like life exposure to environmental microbiome,
13:50:03 I think would be super interesting.
13:50:04 I'm really into microbiome research.
13:50:05 That's why I got into researching.
13:50:07 Because it's, it's hard to get into microbiome research.
13:50:08 If you're not a microbiologist or don't have like a,
13:50:10 kind of a hard science background like that.
13:50:13 But yeah, I think you're, you're absolutely right.
13:50:14 I'd love to do more, maybe a study that restricts the area.
13:50:17 So it's easier to get data,
13:50:18 like you're saying that's maybe specific to like a municipal water
13:50:20 supply or in fire, mental microbiome.
13:50:24 But doing a smaller study, if we can.
13:50:25 The reason we did all of BC is that because the disease is so rare,
13:50:28 we were trying to get as many people as we could into the study.
13:50:30 So when you start to get to smaller areas,
13:50:31 You restrict some of that. So it's easier to get better data for,
13:50:34 for a smaller area, but it's harder.
13:50:35 It's harder to get enough sample size. So, yeah, we definitely though.
13:50:43 I'm kind of trying to get more specific environmental exposures is
13:50:45 definitely on our radar.
13:50:46 And I think you bring up a great point with looking at kind of water
13:50:49 quality and different components of water quality. Absolutely.
13:50:51 That could be relevant.
13:50:52 I think Emil, you Dominica.
13:50:54 Thank you so much for that question and that.
13:50:56 That point and you bring up a really good point.
13:51:09 One thing I really appreciate about large scale
13:51:13 spatial analysis is it does a really good job of helping you target
13:51:17 resources for like finer scale research.
13:51:19 That's more intensive or like moving into like qualitative research
13:51:23 and.
13:51:24 So the work you've done.
13:51:25 Would be really good rationale too.
13:51:27 Get you more resources to dig into individual level data
13:51:31 collection for these hotspots you've identified or.
13:51:34 So it's.
13:51:35 Night creates a nice program of research.
13:51:38 So if anyone would like to come off mute and ask your question,
13:51:40 it certainly is a nice opportunity. Otherwise.
13:51:57 I'm going to go to the Q and a here.
13:51:58 So Andrea says that Andrew has been trying to learn arc GIS,
13:52:02 but it looks like a lot can be done in R is there anything that arc
13:52:05 GIS can do that are can't do,
13:52:07 especially since RJs is already on the computers on campus.
13:52:11 And it's.
13:52:12 And which are locked away until the end of the pandemic.
13:52:14 So it sounds like Andrea is saying that.
13:52:16 Our which is free to download right away on your machine.
13:52:19 Might be a bit more accessible. So, Mia, what do you think about arc?
13:52:22 A GIS versus R yeah. So I mean,
13:52:24 there definitely are things that RJs can do that are can't do.
13:52:33 Like, I don't think we could just abandon RJs and everyone use AR and
13:52:36 there'll be no issues with that.
13:52:37 But I will say there's pretty much nothing that I've come across that
13:52:40 I've wanted to do in my research that I couldn't figure out a way to
13:52:42 do in AR.
13:52:43 And you bring up a great point. So I.
13:52:52 Was in the final two semesters of my masters when the pandemic
13:52:56 happened last March. And I was,
13:52:57 I had to switch to working at home with almost no notice.
13:52:59 So the fact that I had all of my codes set up in our cause all I have
13:53:02 is a Mac book, laptop,
13:53:03 and I would completely melt if I tried to do like a dual boot to run
13:53:06 RJs.
13:53:07 The fact that I had,
13:53:08 I was able to transition my computing resources for my like nice lab
13:53:11 computer at UVC.
13:53:12 To my home Mac book was huge for me. I mean, that was a,
13:53:15 that was a real lifesaver.
13:53:16 I would've essentially had to start over if I,
13:53:17 if I had lost access to arc JS during that, I will say, I think.
13:53:20 I mean,
13:53:21 archery has definitely has a high learning curve at the beginning,
13:53:24 but I think so.
13:53:25 So does learning to program.
13:53:26 And the fact that RGS is really inherently visual, right?
13:53:29 From the beginning. So when you throw data into RGS,
13:53:31 you have to look at it. I mean,
13:53:32 it's set up to look at data immediately.
13:53:34 I think that does make it a little bit easier with RGAs at the
13:53:36 beginning.
13:53:38 But I will say it's fairly easy to learn to do visualizations in R and
13:53:40 there's so many resources online for our, for like the tutorial.
13:53:44 I sent out that people have made to do spatial work and are,
13:53:46 there's definitely a huge community of people who are doing.
13:54:05 Spatial statistics and our, and,
13:54:06 and some of the softwares like geo doc, but like, look,
13:54:08 insulin developed Geodon.
13:54:10 There's like the SPD package R which is a lot of the same
13:54:12 functionality as Geodon.
13:54:14 So there are people who are definitely trying to get at least parallel
13:54:16 functionality with our, and I do find it faster for a lot of things.
13:54:19 So definitely if it's, if you've been using RJs and you're.
13:54:21 You might have to switch because of the pandemic where you're thinking
13:54:23 about switching,
13:54:24 I would say give it a try doing some of the things that you do in RJs,
13:54:27 like search online or search on stack exchange to find out how to do
13:54:30 it at our,
13:54:31 and see if you can implement something and you might actually be
13:54:33 pleasantly surprised by how simple it is, especially if you have some.
13:54:36 Baseline knowledge of RGAs.
13:54:41 Yeah. And just to add to that, I,
13:54:42 I like doing stats in our, like, I don't,
13:54:46 I don't go to arc to do stats.
13:54:49 It's too difficult actually,
13:54:51 but I would echo what me also said about the visual component.
13:54:55 Like ours is not really the place for cartography.
13:54:57 I'll say all the, all the visuals I produced for my thesis,
13:54:59 I did an AR.
13:55:00 I was thinking, actually, when you were showing those maps,
13:55:02 I was like, wow.
13:55:03 Some pretty nice maps in our, so it's definitely like,
13:55:06 it's probably more work to do it in our,
13:55:07 but it would have been way more work to switch back and forth.
13:55:09 I think.
13:55:10 And I didn't have access to RJs, but it's definitely doable.
13:55:13 Like RJs are obviously illustrator is like ideal.
13:55:15 If you really want to make some gorgeous.
13:55:16 Yes, but I would say like,
13:55:17 you can make like decent publication quality maps in our absolutely.
13:55:22 I would definitely agree that their publication quality.
13:55:24 So it depends on the audience, but,
13:55:27 so we'll take one more question.
13:55:29 So Robert says that you mentioned the lack of data differentiating.
13:55:36 Between urban and rural and looking at your rural region specifically,
13:55:40 do you know if there's any correlation between IBS and agricultural
13:55:43 communities?
13:55:44 Like yes.
13:55:45 All the resources like mining, logging, commercial fishing.
13:55:47 Excellent question, Robert.
13:56:01 Yeah, really good question. So I think there's a lot of data on urban,
13:56:03 rural. It's just not necessarily being used by IVD researchers.
13:56:06 So that's what I want to take my geography background and knowledge
13:56:09 and like urban environments knowledge and really apply that to see,
13:56:12 like there's not just one kind of urban environment.
13:56:13 There's so many and same with rural. That's a great question.
13:56:15 So actually,
13:56:16 Living in an agricultural area is,
13:56:18 has been found by a lot of studies to be protective and not just for
13:56:20 IBD, but for things like asthma allergies.
13:56:22 Being raised in like kind of an agricultural setting.
13:56:24 Or living close to an agricultural setting can be really protective
13:56:26 for a lot of these kinds of chronic inflammatory immune related
13:56:29 diseases that we're seeing.
13:56:30 Which are like really the, the big health challenge of our time.
13:56:33 Once the pandemic is over.
13:56:35 Fingers crossed.
13:56:36 That these like chronic inflammatory things like asthma,
13:56:38 heart disease, allergies, all these kinds of things.
13:56:40 This is like a huge, huge health challenge.
13:56:42 And these are all things that are very complex,
13:56:44 probably microbial linked, all kinds of stress linked.
13:56:46 And yeah. So the,
13:56:47 so agricultural areas haven't found to be protective for IVD,
13:56:51 which is a great point.
13:57:03 Looking at BC, but I do think like Delta has a lot of agriculture,
13:57:06 for example,
13:57:07 and that's actually one of the highest incidents areas in the
13:57:09 province, from what we found. So I think,
13:57:12 I think there should be more research on like the type of agriculture.
13:57:14 And like you said,
13:57:15 things like fisheries or mining or all of these other kinds of
13:57:17 resource.
13:57:18 Areas of kind of resource extraction or of,
13:57:20 of like intensive production of the land. I think different.
13:57:22 Different types of agriculture, probably important,
13:57:24 like whether you're raising livestock versus whether you have crops.
13:57:27 Versus like aquaculture or.
13:57:43 Berry production versus like wheat production.
13:57:44 I think all of these things are really important.
13:57:46 And one of the issues with IBD is that it's so rare.
13:57:48 It's a lot easier to do this kind of research with allergies or
13:57:50 asthma,
13:57:51 something that affects a lot more people because you have to have
13:57:54 enough people. This would be probably like a multinational study.
13:57:56 If you were going to try to,
13:57:57 or we're in a bigger country than Canada to try to differentiate like
13:58:00 different types of agriculture to get enough people in each of your
13:58:02 types of agriculture.
13:58:04 That you could really differentiate it.
13:58:05 We got definitely you're you're right on the money with,
13:58:07 with linking agriculture.
13:58:08 To IBD and these chronic inflammatory diseases in general.
13:58:10 And I think over the next 10 to 15 years,
13:58:12 Or even the next five years,
13:58:13 I think we're going to see like huge leaps and bounds in our
13:58:15 understanding.
13:58:16 Oh,
13:58:18 partially from all of the microbiome research that's going on right
13:58:20 now.
13:58:21 And these like big cohort studies where they're following kids from
13:58:24 birth onward, like the child study.
13:58:25 I'm looking at their microbiome.
13:58:26 I think we're going to start to see in the next few years,
13:58:28 a lot of really exciting.
13:58:30 Things come out in our understanding about how these environments and
13:58:32 agriculture, why that could be so important for our health.
13:58:35 Yeah, that's excellent. Thank you so much.
13:58:36 Well, thank you so much.
13:58:38 What a great project and yeah,
13:58:41 really actually just see linking.
13:58:43 Health data with, with canoe data and exposure data.
13:58:47 Maybe it will just ask you to go to the final slide. Yes.
13:58:51 And I have, I just have my email up here.
13:58:52 I can put it in the chat. If anyone has any questions later.
13:58:54 I'm going to talk to me about anything. Feel free to email me.
13:58:56 It's just my name@gmail.com. And if you could go with my name.
13:58:58 There's only one of me. So.
13:59:00 So find me somehow technical difficulties that we had at the start.
13:59:20 But we'll just remind you that we do have a lottery
13:59:23 if you're a student, so you can win a prize. If you fill that out.
13:59:26 And we're going to have our next seminar where we have
13:59:29 a nother graduate student who used.
13:59:33 Exposure data and health data.
13:59:35 To do an excellent project.
13:59:37 And the link.
13:59:38 Here's a link where you can register for that next project or that
13:59:41 next presentation.
13:59:43 And yeah, get in touch with me. I'll if you want to talk more,
13:59:45 there's so much to discuss. You can get in touch with us.
13:59:48 And we look forward to seeing you guys again.
13:59:51 Next month.
13:59:57 And thank you so much to canoe for this excellent collaboration and
14:00:00 thank you to U of T school of cities and the U of T department of
14:00:04 geography and planning. Who's been our biggest supporter.