How to upload HRM files
Future
The information on this page how to upload ASCII records refers to the currently used format. It is currently not possible to upload XHRM files.
Once supported by HRM vendors, only XHRM file will be used for upload and download to keep together data and metadata of patients and records.
Prerequisites
- Make sure that you have checked with you hospital's ethics committee when you make public any patients data. Contact John Pandolfino for an example of an application form ( This e-mail address is being protected from spambots. You need JavaScript enabled to view it ).
- No image files are required, only ASCII data and a single descriptor file containing details of all records (see below) should be uploaded. Images are created on-site and can be modified consistently when required. Currently, files in Sierra (extension:txt) and in MMS format (extension:asc) can be processed.
- Every uploader must register a prefix, e.g. usz or nwu (short: pref in this text). Several co-workers at an institution can share a prefix. Registration of a site prefix is done by This e-mail address is being protected from spambots. You need JavaScript enabled to view it on emailed request.
- Data file names must start with your pref, followed by an arbitrary code without spaces, and have extension .txt.
- Create a descriptor file with Excel having 4 or 5 columns headers as shown below. ";" means new column; don't write the ;, it will be created on export from Excel to csv. Please use the column names exactly as given; no additional columns, no "file number" instead of "name".
CSV Format for Descriptor File
Each HRM file is described by a line in the csv-file.
- name: The filename with HRM data. Use lowercase, no extension. So for a file usz001.txt, use usz001 as the name. There should be no underscore in the file name, because it will have a special meaning in the extension for multiple-swallow uploads for disorder classification.
- patid: A patient-ID unique within you institution. Do not use real patient numbers used in you hospital that would allow patient identification; keep a translation table of your anonymized IDs to the hospital's IDs instead. Number or strings with effectively 11 characters are acceptable. To avoid collisions between uploading institutions, the effective patient name internally will be generated by prepending your site's prefix, e.g. usz1459. For efficient statistics, it is absolutely important to have the patient number included, otherwise no within-between comparison is possible. Future expansions should probably include an instance variable, e.g. pre or post, to descrip multiple tests in one patient.
- comment: May be empty in Excel, or ;; (two semicolons or commas) in csv.
- replaceold: Normally, put a 0 here. For your protection, the upload will be rejected in case some other record with that name exists. When you put a 1 here, you force overwriting of the old record, e.g. when you want to apply a correction. Use it with caution, better contact me first when you need to change something.
- class: This column is optional. You can use it to add your own classification, avoiding do everything online. When you use this feature, you still have to go online to publish all records you uploaded, otherwise the classifications are not visit to the public. There is no strict format you have to follow; the only requirement is that the text contains one of the substrings normal, hypo, absent, hyper, spasm,elevated, paneso. So "PanESOphageal confusion" or "Spasmatic" both work.
- In Excel, use File/Save as Comma separated file (csv). File name of the csv-file is arbitrary, extension will be csv automatically. German versions of Excel use a semicolon as a column separator, others a comma. Stupid Excel, my program checks what the separator is, so both should be ok.
When you open the saved csv-file with Notepad, you should see the following.
name;patid;comment;replaceold
usz1;1459;Das ist usz1;0
usz2;1532;Das ist usz2;0
usz3;1532;Das ist usz3;1
or alternatively
name;patid;comment;replaceold;class
usz1;1459;Das ist usz1;0;normal
usz2;1532;Das ist usz2;0;paneso
usz3;1532;Das ist usz3;1;hypothetical
- Zip all data files plus one csv file. You cannot upload individual files, only zipped packages with a valid descriptor file. Check that in your zipped file there is no "path" given, all files must unpack into the same directory when you try it. Don't put more than 3 records into one zip-file for a starter, a brutal consistency check will be done on the files and data will be rejected on the slightest error. When the first upload worked, you can create packages of any size below 10Mb Zip to upload.
Uploading
- Log in under your name, and select Classification/Upload.
- Select the zip-file you prepared. The button is marked Browse or Durchsuchen, depending on your browser.
- Click Upload File. Uploading needs about 5 minutes for 50 swallows.
- Wait till you get the error messages or a general "successful" note. Contact me (mailto: This e-mail address is being protected from spambots. You need JavaScript enabled to view it ) when you do not understand the error lingo.
- After a successful upload, you will see a Busy... message. This means that interpolation and image generation is active. You can leave the page and come back later, or wait till you see the confirmation of your processing as shown below
| Next > |
|---|
