X-Message-Number: 5133
From:  (Brian Wowk)
Newsgroups: sci.cryonics
Subject: Re: Beyond 2000 and Mind Uploading
Date: 5 Nov 95 17:51:19 GMT
Message-ID: <>

References: <478p8t$> 
<>


In <>  (Anders 
Sandberg) writes:

>David L Evens wrote:
>>The reason why MRIs are of limited resolution is we can only extract 
>>limited ammounts of data from the signal (since there is substantial 
>>noise and computing power is finite).  I expect that by the time we can 
>>build a computer powerful enough to upload to that would be cheap enough 
>>for individuals to actually afford, we will probably be able to build an 
>>NMRI system capable of extracting the neccessary data.

	Sorry, David, but I design MR brain imaging methods for a
living, and what you say here is hogwash.  If you know a way that
simply more computer power can overcome intrinsic signal-to-noise
limits in MRI (or any other noisy data stream), you better write up your
method for Science Magazine right away.  Your Nobel Prize is waiting.

	Anders is absolutely right about field gradients being a problem
also.  The twin problems of SNR (inversely related to the sensitive
area of the receiving coil) and gradient strength mean that MRI will
NEVER give nanometer resolution of the brain non-invasively.  Such
resolution could only be acheived with probes in micrometer-proximity
to the tissue being imaged.

---Brian Wowk  


Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=5133