There are 3 messages in this thread.
You are currently looking at messages 0 to 3.
Hi All, I am working with injecting contrast/dye and measuring the resulting enhancement (resulting from the injection) in a patient. The injection is a square input (e.g., 20 ml @ 4 ml/sec for 5 sec). Now, I measure the resulting enhancement (OUTPUT) at a certain location. Now, this output is related to injection via the transfer function of the patient. so, OUTPUT = TRANSFER FUNCTION of patient (convolved with) INPUT Now, if we have to find the enhancement in the patient for an input of 4 ml @ 4ml/sec for 1 sec in time domain using the above data in time domain. In other words, how can I find the impluse response of the patient in the time domain using deconvolution. How do we do it in Matlab? Is there any Matlab function that can do this. Does this make sense? Thankyou very much for taking your time!! Best Regards, NU This message was sent using the Comp.DSP web interface on www.DSPRelated.com______________________________
NU wrote: > Now, if we have to find the enhancement in the patient for an input of 4 > ml @ 4ml/sec for 1 sec in time domain using the above data in time domain. > In other words, how can I find the impluse response of the patient in the > time domain using deconvolution. To deconvolve atleast one of the two operands should be known. >How do we do it in Matlab? Is there any > Matlab function that can do this. Yes. There is a Matlab function for deconvolution. But again you must know one of the operands i.e. either transfer function or the input. The function is 'deconv'.______________________________
> OUTPUT = TRANSFER FUNCTION of patient (convolved with) INPUT This is only true if the patient is a linear system, which is unlikely. (This is a good example of the opposite of anthropomorphism, which was a thread here recently... the suggested name was mechanicism.)______________________________