Category Archives: Data Collection

Input on iOS!

The previous post covered a promising new tool named Input for data collection. As a reminder, this tool involves setting up a QGIS project and using the Mergin plugin to transfer the project and data to and from your computer and smartphone. It is also a free app. Another aspect of Input is that until now it has only been available for Android.

Yesterday Lutra Consulting announced the release of the beta version for iOS. It is available via an app named TestFlight!

From there is can be used just as was described in the previous post where setting up a project, using Mergin and collected data with Input was described. Once I have had a chance to experiment with it I will follow up with a summary on that. In the meantime, if you are an Apple iOS user, download the app and follow the previous post to give it a trial run.

Using Input/Mergin and QGIS for Field Data Collection

This post will show the basic steps for working with an alternative to Fulcrum for community data collection with a smartphone. The app is named Input. It’s a mobile app developed by Lutra Consulting. The app is free. Until now it was only available on Android. However, Lutra Consulting is working on a version for iOS! The Input workflow is based in QGIS. This means you set up your data layers in QGIS, along with the logic for your data collection. (This will mean a steeper learning curve for those not already familiar with QGIS. But don’t let that discourage you, it is not that much steeper!)

You then use their Mergin  cloud service to synchronize your QGIS project and data between your desktop and phone. There is a Mergin QGIS plugin making this quite simple. All the symbology you set up with your data in QGIS will be honored by the Input mobile app. With the combination of QGIS and Input/Mergin you can mimic all the field data collection functionality covered in a typical Community Health Maps workshop!

Getting started

Here I will walk through the steps in setting up the same type of data collection in CHM workshops. (Note: this post was inspired by the well done video by Dr. Hans van der Kwast on using Input). To get started:

  • Create a folder for your project
  • Open QGIS (in this example I’m using v 3.8).
  • Install the QuickMapServices plugin and add the OSM Standard and Google Satellite basemaps.
  • Zoom in to your study area.

    2019-08-23_131506.png

This step is optional but helpful. You will set up a layer as the study area boundary.  From the QGIS menu bar choose: Layer | Create layer | New Geopackage Layer.

  • Save the Database  (*.gpkg) into your folder.
  • Table name = study_area
  • Geometry typePolygon 
  • Click OK.

Now that the study area has been created, the polygon for the study area can be digitized and then styled. Right-click on the layer and choose Toggle Editing from the context menu. Use the Add Polygon Feature tool on the Digitizing toolbar to trace your study area. 

Styling the Study Area

Clicked F7 to open the Layer Styling Panel.  Here I gave the polygon a Fill Style of No Brush, a Stroke Color of bright red and a Stroke Width of 1.26. This study area covers a school and park near my office.

2019-08-23_135059.png

Creating the Infrastructure Points Layer

This is more important. Create a second layer into which you will record points out in the field. Repeat the above steps to create a point layer and save it into the same GeoPackage. Name the Table infrastructure and set the Geometry type to point. Before clicking OK you will add columns to record the data you will collect. For each new column, enter the name, set the Type , Maximum length and click the Add to Fields list button. (Although not in the screenshot below you can also choose to add a date column – type = DateTime.) When you have added each field click OK to create the layer. 

2019-08-23_143834.png

When saving, choose Add New Layer so the layer is added to the existing GeoPackage.

Now symbolize the points. Here I have used to Simple markers to create a purple target icon.

2019-08-23_144154.png

Save your QGIS project as a QGS file into the same folder. (At the moment Input does not support the zipped default QGZ project file format.)

Creating Offline Basemaps

If you have a cellular connection the basemaps will work fine. However, if you anticipate losing cellular connection out in the field, you can use the Generate XYZ Tiles (MBTiles) processing tool to create offline versions of each basemap. You can set the extent to that of your study area layer. You will need to experiment with the zoom level settings. When they have been created you can use the Browser Panel to add these to your map. 

2019-08-23_145458.png

Working with Map Themes

You can also set up Map Themes. These allow you to have different views of your map. Clicking the eye icon at the top of the Layers Panel will open the Map Themes menu. For example, to set up a map theme for just the Offline Satellite layer the study area and your point layer, you would just turn those three layers on. Then choose Add Theme and name it Offline Satellite from the Map Themes menu. Here I have set up several themes. With these I can easily toggle between these different views while in the field. 

2019-08-23_160237.png

Setting Up Field Widgets for the Points Layer

The next steps are key for mimicking the data form functionality in Fulcrum. You will open the Layer Properties for the point layer, and switch to the Attributes Form tab. Here you can select each Field, and set up custom widgets which will control the editing behavior in the field. Below are the fields typically used in a CHM workshop.

  • fid – you don’t need to see this field, so select it and choose a Widget type of Hidden.
  • Infrastructure type – Select a Widget type of Value map. Here you can enter a series of choices with 1-x values. Under Constraints check Not Null which makes this required data.

2019-08-23_150426.png

  • Bike rack capacity –  Select a Widget type of Text Edit. This allows the data collector to type the number for this. Since this only needs to be answered if the infrastructure type = Bike rack you can enter that as an expression in the Constraints section.

2019-08-23_150739.png

  • Num bikes – set this up the same as the bike rack capacity.
  • Tree type  Widget type = Text Edit and you can set up the appropriate Constraint expression for this.
  • Sign type – Widget type = Text Edit and again set up the appropriate Constraint expression.
  • PhotographWidget type = Attachment. In the Path section click Relative paths. Then scroll down to find the Integrated Document Viewer section. Set the Type to Image.
  • Date – Set the Field Format to Date. Uncheck Calendar popup. Under Constraint check Not Null. Then in the Defaults section enter $now which will default to today’s date. 

Close Layer Properties and save your project. 

Project Properties

Now there is just one more setting to create. From the Project menu choose Properties and switch to the Data Sources tab.  Here check the box making the study_area Read only. This layer is just for reference and will not be edited in the field.

Save your project.

You are done with the set up and will now use the Mergin plugin to migrate your data to the server. In QGIS install the Mergin plugin.

Setting Up Mergin

Visit the Mergin web page (https://public.cloudmergin.com) and set up an account. With the free account you get 100Mb of data storage. (You can inquire with Lutra Consulting about getting additional cloud storage space.) 

In the QGIS Browser panel you will see a new Mergin data provider. Click on it and when prompted enter your credentials.

2019-08-24_111300.png

In the Browser panel again right-click on the Mergin icon and choose New Project. Fill in the details and click OK. The data will upload to your cloud account.

2019-08-23_153025.png

Collecting Data with Input

Now that the project is uploaded to the Mergin server you can switch to your smartphone. Install the Input app. Log in with your credentials and you will be able to access your project from the Projects list. You can then use the More option to switch between different map themes.

Screenshot_20190826-091357.jpg

When you are ready to collect a point, click the Record button and then the Add Point button. The form will open allowing you to collect all the data for that feature including a photograph. The points which have a condition preventing them from being entered will be unavailable and red. For example, the animation below shows a Stop sign being collected. The fields related to other things (bike racks and trees) are unavailable because a condition has been set within QGIS. Once the information has been entered simple click Save and move to the next data collection point.

collectingapoint.gif

Downloading/Synchronizing the Data

When you are ready to download the data you can either 1) Use the Mergin data provider via the QGIS Browser panel to Synchronize your data, or 2) Synchronize the data via the Mergin website.

merginprovider.gif

You can then open the updated project from the same Mergin provider and begin to work with your data!

2019-08-26_131902.png

This is still a very new app but shows immense promise for Community Health Mappers. I encourage Android users to try it out. iOS users stay tuned for future developments.

Pacific islanders Dive Deep into Community Health Maps Workflow

The First Vector Borne Disease Surveillance Workshop

0607191556-e1560536068216.jpg

On June 8 & 9, 2019, twelve pacific island public health professionals met in Honolulu, HI to participate in a Community Health Maps training specifically designed to demonstrate how to collect and work with geographic data related to vector borne diseases, i.e. those that are transmitted to humans via other animals such as mosquitoes. Attendees represented: American Samoa, the Federated States of Micronesia, Guam, the Commonwealth of the Northern Mariana Islands, the Republic of the Marshall Islands and the Republic of Palau.

This was the first of two, two-day, workshops aimed specifically at tackling the spread of diseases like Dengue fever, West Nile and Zika viruses. It was a team effort. The training was organized by the Association of State and Territorial Health Officials (ASTHO). Participants attendance was funded by CDC’s National Center for Environmental Health. Travel for the trainers was funded by the National Library of Medicine, (funding for the workshop is provided under a sub-award from the National Library of Medicine to ICF International). This particular CHM workshop was taught in conjunction with ASTHO’s Insular Area Climate and Health Summit.

After an introduction to the Community Health Maps project – it’s origins, workflow and examples of past projects – participants learned to create a data collection form and use their smartphones to map features (trees, signs, benches etc…) around the IMG_20190610_214340_314_1_previewconference site using Fulcrum

Participants saw how this particular part of the workflow could be applied in their home regions to digitally locate areas of standing water and/ or sand pits that are some examples of breeding habitat for mosquitoes. Others anticipated mapping salt water resistant taro, households where infections have occurred and other geographic factors that contribute to the spread of vector borne diseases.

For the remainder of the first day the group took the data they created earlier and imported into QGIS, a sophisticated geographic visualization desktop software. In this section they became familiar with QGIS and how to symbolize layers and make a print map.

In addition to the data collected on site, we worked with mosquito data acquired courtesy of Dr. Chris Barker  covering Madera County, CA. The data included mosquito trap results over five years, virus testing, mosquito biting complaints, storm drains, parcel boundaries, roads and a hypothetical case of Dengue fever.

IMG_1909_1_preview

The second day focused on generating vector borne disease surveillance products. Kurt Menke developed a curriculum to demonstrate how a GIS can create maps and statistical charts that transform simple text and numbers in a database into intuitive graphics that communicate information quickly and accurately. The previous blog post has more detail about the specific vector borne disease surveillance products participants learn to create.

20190607_114226(0)_1_previewThe attendees had a wide range of GIS skills from introductory to advanced capabilities. We experienced many of the common technical difficulties when working in a hotel conference room, older and newer computers and variations with different operating systems (Windows and Macs) as well. Despite all the differences, all of the participants: A) collected data with their smartphones, B) exported their data to a desktop GIS, C) used prepared data to create geographically accurate statistics, D) generated heatmaps of mosquito populations, E) calculated the minimum infection rate per year for West Nile and St. Louis Encephalitis viruses, F) identified potential sources using buffer operations with distances specific species are know to be able to travel, G) identified parcels at risk due to their proximity to a fictional outbreak of Dengue Fever and H) generated trend graphs of mosquito populations through time via the QGIS Data Plotly plugin. All participants received official QGIS certificates.

The skills required to complete these tasks are not always simple and straight forward. The participants of this workshop expressed great enthusiasm and persistence in figuring it all out… making mistakes and trying again. Many expressed a need for more training and a desire to have more specialized trainings on site specifically related to projects they are already working on.

The second workshop in the series will be taught next week in Providence, RI at the GIS Surveillance Workshop. This will be attended by State based health officials.

This vector borne disease surveillance version of the Community Health Maps workflow showcases the analysis and data visualization capabilities of QGIS, as well as, the data collection capabilities of Fulcrum. It represents perhaps the greatest potential for applied use of Community Health Maps to date.

These workshop materials will part of the suite of https://communityhealthmaps.nlm.nih.gov/resources/ available through the Community Health Maps program in the near future.

If you are interested in having this taught for you or your colleagues contact Kurt Menke (kurt@birdseyeviewgis.com)

Community Health Maps at Rising Voices 7

Last week Community Health Maps traveled to Boulder, Colorado to teach a pre-conference workshop at Rising Voices 7. The theme was Converging Voices: Building relationships and practices for intercultural science. The conference was hosted by the National Center for Atmospheric Research (NCAR). The workshop was well attended with about two dozen participants representing numerous AI/AN tribes and other organizations. These workshop was part of the Community Health Maps project and was funded by the National Library of Medicine (funding for the workshop was provided under a sub-award from the National Library of Medicine to ICF International).

The goal of Rising Voices is to “advance science through collaborations”. Participants learn how indigenous and western scientific knowledge systems can compliment one another and advance our understanding of important issues in our communities. The focus is on climate.

2019-05-20_164427

A photo taken by Dr. Angel Garcia (https://www.jmu.edu/geology/people/all-people/garcia-angel.shtml) during the workshop

At three hours the workshop was slightly shorter than normal. This allowed us to focus on field data collection with Fulcrum and web mapping with Carto. With a few minutes to spare Kurt Menke shared QGIS. Since he didn’t have time to really demonstrate the use of QGIS he focused on the open source aspect. As an open source project, QGIS is both GIS software and a community. As such it aligns with the ethics many communities try to foster at Rising Voices.

2019-05-22_080251

A field data collection selfie!

 

Community Health Maps in Michigan: Four Workshops in Four Days in Four Cities!

This past week Kurt Menke traveled to Michigan and taught four Community Health Maps workshops for the University of Michigan. These were organized by the University of Michigan Libraries and were done in four days on four different campuses: Ann Arbor, Flint, Detroit Center and Dearborn. These workshops were part of the Community Health Maps project and were funded by the National Library of Medicine (funding for the workshops was provided under a sub-award from the National Library of Medicine to ICF International)

Blog

A map showing the four workshops and the route taken between them.

Due to great outreach by the University of Michigan team of – Tyler Nix, Marisa Conte, Alexa Rivera, Justin Schell, Sara McDonnell, Troy Rosencrants, Kui-Bin Im and Claudia Walters – the workshops had great attendance and ran like clockwork.

The first was held at the University of Michigan – Ann Arbor at the Hatcher Graduate Library.  Twenty eight people attended. The audience was mainly a mix of faculty and representatives of local public health/community organizations with a few students. The final hour was reserved for group discussion and Justin Schell did a fantastic job moderating. Ideas for mapping park safety and identifying underserved populations were discussed as projects where mapping could help.

IMG_0204

Participants at the Ann Arbor workshop

The second was held the following day at the UM Flint. We started in the University Center Happenings Room and moved to Thompson Library for the afternoon QGIS session. Forty five people came out for this second training, quite a bit more than we expected. This included some walk-ins and late registrants. In addition to faculty and students there were quite a few representatives from local county health departments and non-profits.

20190403_134426

UM – Flint workshop participants working on QGIS.

The third was held at the UM Detroit Center on April 4th. Thirty people registered for this workshop. This was probably the most diverse group including some UM faculty, plus Wayne State faculty, county public health staff (including some as far away as Saginaw County) and non-profit public health workers.

20190404_142046

Kurt Menke describing Help Resources at the Detroit Center

The fourth was held at UM Dearborn on April 5th. This was the smallest of the four with 20 registrants. However, this allowed us to get a little farther into the capabilities of QGIS.

20190405_103209

Data collection at the UM Dearborn campus.

One of the most helpful components was a web page Marisa Conte set up for the workshops. It initially had all the important details for each workshop, including driving directions, links to the slides and workshop data. This alone set a new standard for workshop outreach and organization. During the discussion on the first day she began adding url’s for sites containing useful data and other resources identified during the afternoon discussion. During the week this  evolved into what is now a fantastic resource for everyone who attended. This will be helpful to anyone implementing CHM related projects or those who weren’t able to attend.

For the week  123 people were trained in Community Health Maps! This is a record that may never be broken. However what came out of the week more than the numbers, is the incredible potential for projects that can make a real difference in these communities. I’m looking forward to working more with everyone I met. Thanks again to UM Michigan for organizing such a successful series of workshops.

Mapping Field Photos in QGIS

We are on the eve of another QGIS release. Version 3.6 will be released any day now. This will mark an important milestone in QGIS development. QGIS 3.4 will become the first long-term release for the 3.x line. With 3.6 will come one specific change pertinent to this topic: the Raster Image Marker. Consider this post a small preview of what you can expect with v3.6!

In the typical CHM workshop attendees are shown how to use Fulcrum to download data collection points. These include photos. It is also possible to quickly and easily map any geotagged photo from your smartphone using just QGIS. For this example, I am using some photos I took yesterday during a hike with my iPhone.

    1. Using the Processing Toolbox, search for and open the Import Geotagged Photos tool and set it up as shown in below, pointing the tool to the folder containing your photos and naming the output point file.ImportGeoTaggedPhotos
    2. The data are in geographic coordinates (latitude and longitude). It will be helpful to project the photopoints to a Cartesian coordinate system such as UTM or State Plane. To do this right-click on the photopoints layer in the Layers Panel and choose Export –> Save As from the context menu.
    3. In the Save Vector Layer As window save a new copy of the layer. If you need help determining which coordinate system to use feel free to reach out to Kurt Menke (kurt at birdseyeviewgis.com). He is happy to help!SaveVectorLayerAs
    4. Now you have a couple interesting ways to visualize these points. First you can generate something called a Wedge Buffer.  These are pie shaped polygons that you can set up to represent the field-of-view of the photograph.wedgebuffers
    5. Search the Processing Toolbox for the Create Wedge Buffers tool.  The Input Layer will be the reprojected photopoints. The Buffers output can be a shapefile in your photos folder. The real trick is using what is known as a Data Defined Override. In QGIS you can use values in attribute columns or expressions for tool inputs instead of putting in a single typed value. In the animation below, a Data Defined Override is being used for the Azimuth parameter. This determines which way the wedge will be pointed. The Import Geotagged Photos tool extracts more than just photo location. It also extracts direction, altitude etc. Here the Azimuth parameter is simply being pointed to the direction attribute column and this orients the wedge in the direction the photo was taken! The only other parameter to set is the Outer Radius. This determines how long the wedge will be. Here it is being set to 300 meters.wedgebuffers
    6. Now there are both photo locations as points, and wedge buffers showing the field-of-view! wedgebufferscreated

 

  1.  Next let’s improve the default symbology for the wedge buffers using a Shapeburst fill. You will open the Layer Styling Panel (F7) and set the wedge buffers as the target layer. Next select the Simple Fill component and switch the Symbol layer type from Simple fill to Shapeburst fill. You can then set the two colors to use. Here I am using red and transparent. You can then set a distance for the effect and play with other settings like Blur strength. shapeburst
  2. Finally you will learn a feature that will be released with QGIS 3.6: Raster Image Markers. Here the target layer is the photopoints layer. The Symbol layer type is being switched from Simple fill to Raster image marker. Then the Data Defined Override for the image is being set to the photo field, and the Rotation parameter is being set to the rotation field!rasterimagemarker
  3. Now the photos have been mapped. Their field-of-view is being represented by a wedge buffer, and the photo itself is added to the map with a Raster image marker!photosmapped

Stay tuned for more tips on mapping and updates about this latest QGIS release.

UMD Students Learn How to Map Health Issues

By Angela Kim & Colette Hochstein

On October 4, 2018, National Library of Medicine® (NLM) Research Assistant Julian Argoti and University of Maryland (UMD), School of Public Health intern Angela Kim spoke to the “Professional Preparation in Community Health” class at the University of Maryland, School of Public Health, College Park. The 75-minute class was attended by approximately 45 undergraduate students in the UMD Behavioral and Community Health (BCH) program. The presenters introduced NLM’s Community Health Maps (CHM) blog and facilitated a hands-on activity.

UMD1

UMD Fall 2018 “Professional Preparation in Community Health” Class

The presentation covered a basic introduction to the tools and workflow in CHM. The students were asked to use Fulcrum, a low-cost tool, to build a custom data collection form for the first step of the CHM workflow – data collection. They discovered first-hand how intuitive the tool is. After creating their own custom data collection form on the topic of their choice, the students left the classroom to collect data points around the School of Public Health building.

UMD3

Students testing functionality of water fountain.

On their return, the students explored maps of their data points on Fulcrum. Many noted that the process of collecting data points was easy and fun. UMD Professor James Butler mentioned that although drinking a good amount of water is emphasized at the School of Public Health, he had not previously noticed that there is no water fountain near the faculty lounge. His comment underscored that issues are often not observed until actively examined, as during the mapping process.

UMD2

Students collecting data points on water fountain locations

The class ended with Professor Butler concluding the class by reiterating how CHM can serve as a useful tool for visualizing many of the different health issues discussed in class.

The students were alerted to the free new online Community Health Maps online tutorial, a self-paced course from the NLM designed to help users gain the skills needed to use Community Health Maps.