Home

Analyze - Web App

Preview ScienceIO's PHI identification and redaction capabilities using our new web-based app.

Identifying and redacting protected health information (PHI) from healthcare data is useful in a number of settings. ScienceIO has developed a best-in-class model for separating PHI from medical concepts using AI.

Our new web-based app has made the technology accessible to anyone seeking to evaluate our capabilities in this area. If you like what you see, be sure to let us know! New API endpoints and SDK support will be available in an upcoming release so that you can test it with your code.

1512

📘

This app is available to anyone with a ScienceIO account, including those on the free tier.

If you don't yet have an account, get started for free at https://app.science.io/signup.

Why Try It?

The Analyze web app allows you to test our technology on your healthcare data without having to write code or make a call to the API (we will do that for you). You can then analyze the results in a user-friendly format to determine how well ScienceIO has identified or redacted PHI for your needs.

Use this app if:

  • You want to evaluate our PHI identification or redaction capabilities without having to get API keys, write code, or execute a call
  • You do not have a coding background but want to quickly try out some PHI technology on your data
  • You want to test our PHI identification and redaction capabilities against another tool in the market, or against your own internal capabilities

The results may inform your decisions about how you may be able to use our API to:

  • Identify and redact HIPAA-protected information in your data, so that you can safely use the file within your organization or outside of it
  • Work toward creating a structured, comprehensive patient profile/record from multiple documents by identifying, aggregating, and disambiguating the PHI

Supported Files

You may submit healthcare data to the web app using text files (.txt) with your own healthcare data, or by copying/pasting text.

📘

Note

Support for the JSON file type is coming soon, as are ScienceIO sample healthcare data files.

PHI Types (Preparing Your File)

Files for upload should contain unredacted PHI. You will get the best results by ensuring a wide mix of PHI that includes any of the following:

* Age
* City
* Country
* Date
* Device
* Doctor
* Email
 * Fax
* Health plan
* Hospital
* ID number
* Medical record
* Organization
* Patient name
 * Phone
* Profession
* State
* Street
* URL
* Username
* Zip code

How to Use the App

  1. Visit app.science.io and click Login.
    *If you don't yet have an account, get started for free at https://app.science.io/signup.

  2. Click the Upload button to upload your test file, or copy/paste text into the text box.

  3. Select either Identify (to identify all PHI) or Redact (to remove all PHI).

  4. Click Analyze to view your results.

  5. To run another analysis, click Start New Analysis to start over.

❗️

We do not save your results.

If you close the window or click Start New Analysis, you will have to start over to see your results again.

Understanding the Results

The Output Screen

The web app outputs an easy-to-read version of a JSON file that contains the results of the PHI identification process. Each piece of PHI is shown in blue, and clicking on any piece of blue text will display the type of PHI that was identified.

Technical users:
If you would like to see the original JSON file that was generated by the API from your text, click the .JSON tab.

Troubleshooting

Upload Errors

Most upload errors are caused by incorrect file formats or damaged files.

  • Make sure you are uploading the correct file type; the file must be a .txt file.
  • If you suspect a problem with the file or the upload process, try copying the text from the file into the text box instead. Then try again to get your results.

Incomplete or No PHI Identification

PHI should be identified in all text that contains it. If this does not happen:

  • Make sure your file included unredacted PHI.
  • Make sure you have included a variety of PHI types.
  • Make sure the information is actually PHI; medications like Benadryl, for example, are not PHI.

Remember that only information with identifiers tying it to an individual is considered PHI.

  • "Type II diabetes" is health information, not PHI.
  • "Sally Jones has type II diabetes" includes a patient identifier and therefore is PHI; removing the patient name turns it back into health information.

Seeing unexpected results? Send us an email to tell us about it.

Deleted File or Output

Because this tool was created for evaluation purposes only, we do not save your file or the output of the PHI identification/redaction. It is only read and evaluated for the purposes of creating a web-based output for immediate viewing.

If you wish to see the output again, please re-upload the file or copy/paste the text again.

Cannot Access the Web App

In this release, the web app is not yet compatible with mobile devices. Be sure to use a desktop browser to try the app.

Feedback

We'd love your feedback! Tell us what you think about this product. Email us at [email protected].

Updated 6 days ago

What’s Next