Predicting Mortality of ICU Patients: The PhysioNet/Computing in Cardiology Challenge 2012 1.0.0
(2,395 bytes)
function [ pafi ] = pnPaFi(data)
%PNPAFI Extract PaFi, ensuring they are measured at the same time point
% [ pafi ] = pnPaFi(data) extracts the PaFi ratio, or the ratio between
% PaO2 and FiO2 (PaO2 / FiO2) for syncronously measured PaO2 and FiO2
% values. FiO2 is interpolated using sample-and-hold as detailed in
% pnFIO2.
%
% Inputs:
% data - Cell array of data inputs in standard format:
% Column 1 - Subject IDs
% Column 2 - Time stamp vectors for each subject
% Column 3 - Feature name vectors for each subject
% Column 4 - Data value vectors for each subject
% Empty cells indicate no available measurements (missing data).
%
% Outputs:
% pafi - Cell array of only pafi data in standard format:
% Column 1 - Subject IDs
% Column 2 - Time stamp vectors for each subject
% Column 3 - Feature name vectors for each subject (only contains
% 'PaFi')
% Column 4 - Data value vectors for each subject
% Empty cells indicate no available measurements (missing data).
%
% Example
% %=== Load data in
% load('data_processed_cell.mat');
%
% %=== Calculate score
% [ pafi ] = pnPaFi(data);
%
% See also PNFIO2 PNEXTRACTFIELD PNLOADTEXTFILESCELL PNIANDREW PNMAIN
% References:
% Physionet Challenge 2012
% http://physionet.org/challenge/2012/
% Copyright 2012 Alistair Johnson
% $LastChangedBy: alistair $
% $LastChangedDate: 2012-04-24 22:39:24 +0100 (Tue, 24 Apr 2012) $
% $Revision: 342 $
% Originally written on GLNXA64 by Alistair Johnson, 24-Apr-2012 14:02:09
% Contact: alistairewj@gmail.com
%=== PaFi = PaO2 / FiO2
pao2 = 'PaO2';
pao2 = pnExtractField(data,pao2);
%=== Extract sample and hold FiO2 values
[ fio2 ] = pnFiO2(data);
%=== For each record, extract PaO2 times and calculate PaFi at them
pafi = cell(size(pao2));
pafi(:,1) = pao2(:,1); % subject ID
N = size(pao2,1);
for k=1:N
if isempty(pao2{k,2})
continue;
else
%=== Get times for pao2 and fio2
t1 = pao2{k,2}; t2 = fio2{k,2};
%=== Intersection
[t_impute,ti,tj] = intersect(t1,t2);
pafi{k,2} = t_impute;
pafi{k,3} = repmat({'PaFi'},numel(t_impute),1);
pafi{k,4} = pao2{k,4}(ti) ./ fio2{k,4}(tj);
end
end
end