11/11/25 - Episode 9 - SNAPSHOT

Scott Isaacs:  Welcome to the ninth webinar in our series. We took a brief hiatus from producing these, but we're looking forward to releasing these monthly again, starting with this one. In this webinar, we will be presenting a demo of Snapshot. I'm Scott Isaacs, the Chief Architect and Senior Innovator here at URMC Wilmont Cancer Institute's Technology and Innovation Group and leading today's demo is our Lead Engineer] and Principal Developer, Lisa Smith. Lisa has been with us for years now in the Technology Innovation Group and has developed many cool apps like Cohort Builder and Snapshot, that you will see today. As a reminder, these webinars will be about 15 minutes long and will feature a few previously submitted questions at the end. And without further ado, take it away, Lisa.

Lisa Smith:  Thanks, Scott. Okay, as Scott said, I'm Lisa Smith and I'll be walking you through Snapshot, our award-winning self-service geospatial analytics platform. We'll start with the motivation behind the tool, followed by an overview of its functionality and key features.  Next up, I will demonstrate Snapshot live and go over recent enhancements to the product. I'll introduce you to the development team and provide sources for more information on the tool and the Technology and Innovation Group here at Wilmot and we'll have time for questions at the end.

The motivation for Canvas Snapshot came from the need to understand the patient population at Wilmot in the context of their region and identify possible patterns, trends, or areas of interest. Our goal was to enable users to independently and explore multiple regional factors alongside live, validated, Wilmot data from a geospatial perspective and potentially aid an allocation of our limited resources. For example, how do the regional cancer incidents and mortality rates relate to the diagnoses we see at Wilmot for the same region?  We can identify patterns or clusters. Excuse me, can we identify patterns or clusters that indicate the need for a new clinic or a new clinical trial? Are there regions of the catchment more prone to the cancer risk behaviors or areas that need more services in general due to other factors? And are these relationships if they exist changing over time?

What is Snapshot?

As I said, Snapshot is a self-service geospatial analytics platform, which provides users the ability to explore live, validated Wilmot data alongside regional public health data with virtually no barrier to entry. Once a user has access to the tool, they can create up to four maps on which to visualize Wilmot data in the form of counts of new patient visits, new diagnoses, patient deaths and clinical trial accruals broken down by disease, patient age, sex, race group and time period of interest. At the same time, they can view relevant public health data from those regions, including cancer incidents and mortality, behavioral risk factors and population demographics from SIR, BRFSS and ACS data sets. They can also select their geographic region size, county or zip code and at the state or catchment level. The user interface of the tool includes one default map at startup. The users can select their desired geographic region, the data source of interest and the specific variables of interest from that data source.

Features:

While the map focuses on the catchment, the Zoom feature allows users to adjust their view as needed. All maps share a standardized legend to enable consistent statistically sound comparisons and by clicking the Add Map button, users can explore data on up to four maps at once. If the user wants to explore internal data, additional filters are revealed so they can tailor that data set across diseases and associated disease subgroups, a period of time, and the sex, race and age group of the patients of interest. There's also a Hide or Show button for filters, which is especially helpful when comparing multiple maps.  Once the user defines their cohort of interest, they can click the Update Map button, and the data is pulled live, at the moment and rendered to the map. While the map is loading, a dynamic title is generated that describes what's being displayed. And users can click the camera icon to print or save their maps.

In this example, we explore lung cancer. We use four maps and we first consider community smoking prevalence for the whole state by county and then focus on the catchment by zip code. Next, we compared the community data to the new diagnoses of lung cancer at Wilmot in the first three quarters of last year. And finally, we can add a map to look at the same diagnoses for the first three quarters of this year. We can review, alter and compare these graphs on one screen with just a few clicks.

Live Demo:

So let me take a moment and actually see this tool in action. As you can see the tool loads, we have this one map. I'm gonna go ahead and open up four maps. To get us started, and they all open up the same. I'm going to start in the first map. Here you can see the geographic regions. I'm gonna just leave it where it is and look at the percent of our population that is at least 65. Well, this is somewhat informative. What is nicer is to drill down to the zip code so that we can get a more intimate look at this data. Now this is drawing, coloring and creating this map from the ACS data that we store on Hyperion. We can zoom in and focus ourselves in the region closest to Wilmot. If we click on any zip code, we can see that in this example in zip code 14580, 19.94 or roughly 20% of the population, is 65 or older.

 Now let's combine this with a look at some of our internal data. Specifically, new diagnoses of breast cancer. We could look at bilateral, left or right breast cancer. I'm just gonna stick with all types of breast cancer. And let's look at the diagnoses we've seen since the beginning of last year. For females over the age of 65. You can see the title I was discussing.  And while that loads, I'm going to take a minute and look at this same group of patients. Patients with breast cancer, all types. But this time I'm going to look at the trial accruals that we've seen from this group. We hide the filters so that it makes it easier to look at all the maps at once and we zoom in. We can look for patterns, we can see, is there an area of need? If we track backwards, say from 14424, or that zip code I talked about before, 14580, we can see a larger prevalence of new diagnoses and a larger prevalence of trial accrual. You can go back to the first map, you can look at other factors, maybe obesity or smoking, things like that that link to breast cancer as well.

So that's the basic walkthrough of the tool. So, I wanted to take a minute and tell you a little bit more about our team. This version of Snapshot is just the latest version in this type of surveillance tool that was originally designed by our director, Eric Snyder. And for this current version, I was the Lead Developer and worked with Scott Paoni, our Head of Product, on the core deliverables.

While I was responsible for the majority of the build, Emily Strong was instrumental in curating, validating, and refreshing the community data. And the user interface is the work of our team member, J.C. Conrad. For full overview of our team members and the work we do here at Wilmot, please visit our website and I've provided a QR code as well. For more information on Snapshot, you can read our article in CXO Tech Magazine by following this QR code, and we also have one down below here for our YouTube channel, which has demos of this tool and other technologies we have developed, as well as previous webinars and Q&A sessions. Thank you for spending some time with us today and I'll open it up for questions.

Scott Isaacs: Thank you for that, Lisa. Excellent work as always and a very nice tool that you've built there.

Q/A Session: 

Q: So, first question that we have that has been submitted was how often is this data refreshed?

 A: So, this tool is pulling data live from the Hyperion database, which updates the Wilmot patient at least daily. So, if a patient was newly diagnosed yesterday, they were probably in that data set we were just looking at with breast cancer. The external data sets are refreshed as they're released, but only after we validate that it's good data to release and include in the tool. So, it's internally validated by us.

Scott Isaacs: Great, so near real time on the internal data and then as soon as possible, once it's been validated for external. That's great.

Lisa Smith: Yes.

Q: And how easy would it be to add in an additional data filter?

 A: So, I'm excited to tell you, we just did some recent upgrades to Snapshot and I was able to port over a similar modular architecture that we use in a tool called Cohort Builder.  And so now it is a very light lift. We could say change age ranges or add another filter if needed and in fact, if any of you watching have any ideas or requests for improvements, feel free to contact us or put in a ticket if you are in our ticketing system.

Scott Isaacs: Good deal.

 Q:  And I think we've got time for one more question and that would be what makes Snapshot different from other traditional reporting tools or dashboards?

A: I would say definitely the live data delivered on demand. Most reporting tools or dashboards, they use static data with a reset schedule or a refresh schedule that's set up ahead of time at best. But with Hyperion, WCI data is continually updated as it comes in.
 So anytime you draw a map, you're seeing the most up-to-the-minute internal data that we can provide you. And I think that that's really crucial for decision making.

Scott Isaacs:  Absolutely. Having that data as soon as you can is definitely critical in a lot of these in dealing with cancer research and clinical operations. So Lisa, thank you so much. The tool looks fantastic and it's a lot of good work you've done on it as usual. So, I appreciate you taking the time to do this webinar. And for those of you that have been tuning in, thank you and have a great day.

 Lisa Smith:  Thank you!