ClinicalVis is a medical record visualization that attempts to improve upon the baseline table-based medical record visualizations.
See the paper that describes and evaluates the visualization on arxiv.
This project contains two Polymer elements:
- clinical-vis: The updated medical record visualization
- clinical-table: A baseline table-based medical record visualization
First, make sure you have the
Polymer CLI installed. Then run
polymer serve to serve your element locally.
From this top-level clinical-vis directory:
$ polymer install
$ polymer serve
Then navigate to:
http://localhost:8081/components/clinical-vis/demo/index.htmlfor a demo of the clinical-vis visualizationhttp://localhost:8081/components/clinical-vis/demo/table-index.htmlfor a demo of the baseline clinical-table visualization
Each of the elements exposes a single bound parameter, 'json', which should be set to a JSON object containing the patient record to visualize. This is the expected format of the JSON object:
Top level key-value pairs:
- 'diagnosis': (string) Admitting diagnosis
- 'dob': (string) Date of birth (YYYY-MM-DD)
- 'ethnicity': (string) Ethnicity
- 'gender': (string) Gender
- 'icu_type': (string) ICU type
- 'notes': (array) List of notes in the medical record, see below.
- 'events': (array) List of events in the medical record, see below.
Each note in the 'notes' field is an object with the following key-value pairs:
- 'timestamp': (string) Number of seconds since UNIX epoch when the note was taken
- 'category': (string) Category of the note, such as "Nursing" or "Radiology"
- 'note': (string) Text of the note.
Each event in the 'events' field is an object describing a measured or observed value with the following key-value pairs:
- 'timestamp': (string) Number of seconds since UNIX epoch when the event was recorded
- 'key': (string) Name of the event, such as 'Temperature' or 'Heart Rate'
- 'category': (string) The category of the event, such as 'Vitals' or 'Labs'
- 'value_string': (string) If the event data is a string, such as the the value 'Clear' for the category 'Lung Sounds', then this field holds the value.
- 'value_num': (number) If the event data is a number, such as the the value 165 for the category 'Platelets', then this field holds the value.