meta00 / vital_sqi

A package for physiological signal quality control
MIT License
31 stars 11 forks source link

App dashboards #27

Open meta00 opened 3 years ago

meta00 commented 3 years ago

Dashboard1: SQI Computational Dashboard. Allow user to input a list of medical signal segments and computes a list of all sqi scores. Input type: Multiple files (CSV)Functions: 1) Load multiple files. Option to preprocess data (filtering - detrend) for the user to decide2) Call all sqi computational feature from sqi package and save as a pandas 3) Save the sqi score file to local machine in DataFrame. Output: A csv-file save all of sqi scores of each segment. Rows are a list of segments. Columns are segment names and their relevant sqi scores.

Dashboard 2: SQI Decisional RulesAllow user to: select sqi to preserve and defined the threshold. Arrange the order of sqi rules, which sqi score will be applied first.Input type: sqi-scores csv file - obtained from the dashboard 1 Functions: 1) Get the check boxes of selected sqi scores. Default is to check all 2) Get the user defined threshold of selected sqi from the textbox. Check input type is numeric and also display tooltip of the normal range 3) Get the user defined rule order from text box. Check input type is numeric; more than one rule can have the same order.  4) Generate a rule-based tree from 1-3 settings. Save it as a XML/JSON file 5) Download the rule fileOutput: list of selected sqi, threshold for confirmation

Dashboard 3: Display the list of good and bad signal in different colors. Input: The decision rules and relevant sqi scores Function: Apply the rule base from 2.1.2 to predict the quality of signal Output: Decision of each segment - good or bad

Dashboard 4: Editing signal. Remove or keep bad quality prediction. Input: The list of bad signal. Function: remove invalid. Save the good quality segment and export to csv