Deconvolution of Transmission Images
Locked
 
|
Richard Cole |
Deconvolution of Transmission Images
|
*****
To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy ***** You are correct, transmitted images are formed completely differently than fluorescent images, therefore the algorithms are separate. Having said that, we use Autoquant for both types of images. For DIC and Phase image we have also used a simple summed projection for qualitative purposes. Rich Richard Cole Research Scientist V Director: Advanced Light Microscopy Core Unit Wadsworth Center Research Assistant Professor Dept. of Biomedical Sciences School of Public Health State University of New York P.O. Box 509 Albany N.Y. 12201-0509 518-474-7048 Phone 518-474-4430 Fax Email <mailto:[hidden email]> [hidden email] Website <http://www.wadsworth.org/cores/alm/index.htm> www.wadsworth.org/cores/alm/index.htm IMPORTANT NOTICE: This e-mail and any attachments may contain confidential or sensitive information which is, or may be, legally privileged or otherwise protected by law from further disclosure. It is intended only for the addressee. If you received this in error or from someone who was not authorized to send it to you, please do not distribute, copy or use it or any attachments. Please notify the sender immediately by reply e-mail and delete this from your system. Thank you for your cooperation. |
|
|
*****
To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy ***** You cannot really deconvolve transmission images, because the image of a point is not independent of its neighbours, so each point is not in fact convolved with a psf. You can however isolate in-focus information from out of focus information and there are several algorithms for this. The basic idea is than in-focus detail will change a lot if we blur it whereas out of focus information won't change much. So you can extract 3D information, or construct an extended-focus image, but it isn't deconvolution. Guy -----Original Message----- From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Richard Cole Sent: Friday, 23 March 2012 1:32 AM To: [hidden email] Subject: Deconvolution of Transmission Images ***** To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy ***** You are correct, transmitted images are formed completely differently than fluorescent images, therefore the algorithms are separate. Having said that, we use Autoquant for both types of images. For DIC and Phase image we have also used a simple summed projection for qualitative purposes. Rich Richard Cole Research Scientist V Director: Advanced Light Microscopy Core Unit Wadsworth Center Research Assistant Professor Dept. of Biomedical Sciences School of Public Health State University of New York P.O. Box 509 Albany N.Y. 12201-0509 518-474-7048 Phone 518-474-4430 Fax Email <mailto:[hidden email]> [hidden email] Website <http://www.wadsworth.org/cores/alm/index.htm> www.wadsworth.org/cores/alm/index.htm IMPORTANT NOTICE: This e-mail and any attachments may contain confidential or sensitive information which is, or may be, legally privileged or otherwise protected by law from further disclosure. It is intended only for the addressee. If you received this in error or from someone who was not authorized to send it to you, please do not distribute, copy or use it or any attachments. Please notify the sender immediately by reply e-mail and delete this from your system. Thank you for your cooperation. |
|
|
Jim Mansfield |
Re: Deconvolution of Transmission Images
|
|
*****
To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy ***** Hi Guy, Maybe I'm just being overly fussy, or maybe I am not understanding some part of using deconvolution to resolve focal planes, but of course one can do real and actual deconvolution on brightfield images. The DeltaVision products all do widefield (i.e., non-confocal) deconvolution (albeit in FL) and that is very similar to brightfield deconvolution. Besides, confocal images are not completely independent; real point-spread functions do go to infinity on both sides of the focal plane (follow the direct light path through the pinhole from source to detector). Generating the PSF for brightfield deconvolution is much more difficult than for confocals, and so it is considerably more problematic than for a confocal, and there can be serious errors generated that way, but of course one can, in fact, really deconvolve brightfield images! Jim -----Original Message----- From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Guy Cox Sent: Thursday, March 22, 2012 2:58 PM To: [hidden email] Subject: Re: Deconvolution of Transmission Images ***** To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy ***** You cannot really deconvolve transmission images, because the image of a point is not independent of its neighbours, so each point is not in fact convolved with a psf. You can however isolate in-focus information from out of focus information and there are several algorithms for this. The basic idea is than in-focus detail will change a lot if we blur it whereas out of focus information won't change much. So you can extract 3D information, or construct an extended-focus image, but it isn't deconvolution. Guy -----Original Message----- From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Richard Cole Sent: Friday, 23 March 2012 1:32 AM To: [hidden email] Subject: Deconvolution of Transmission Images ***** To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy ***** You are correct, transmitted images are formed completely differently than fluorescent images, therefore the algorithms are separate. Having said that, we use Autoquant for both types of images. For DIC and Phase image we have also used a simple summed projection for qualitative purposes. Rich Richard Cole Research Scientist V Director: Advanced Light Microscopy Core Unit Wadsworth Center Research Assistant Professor Dept. of Biomedical Sciences School of Public Health State University of New York P.O. Box 509 Albany N.Y. 12201-0509 518-474-7048 Phone 518-474-4430 Fax Email <mailto:[hidden email]> [hidden email] Website <http://www.wadsworth.org/cores/alm/index.htm> www.wadsworth.org/cores/alm/index.htm IMPORTANT NOTICE: This e-mail and any attachments may contain confidential or sensitive information which is, or may be, legally privileged or otherwise protected by law from further disclosure. It is intended only for the addressee. If you received this in error or from someone who was not authorized to send it to you, please do not distribute, copy or use it or any attachments. Please notify the sender immediately by reply e-mail and delete this from your system. Thank you for your cooperation. |
|
|
*****
To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy ***** One can only regard an image as the convolution of an object with a PSF if each point is imaged independently. Abbe showed that this is not so in regular widefield imaging. It IS so in fluorescence (Rayleigh imaging). In DIC it is even more obviously not so, since we are combining two beams, one through the point and one through its close neighbourhood, to form the image. As I said in my previous post, that doesn't mean we can't use algorithmic techniques to isolate in-focus information, but this is not deconvolution, it's image processing. In particular, it cannot improve resolution which, in principle, deconvolution can - Superresolution Three-Dimensional Images of Fluorescence in Cells with Minimal Light Exposure: Walter A. Carrington, Ronald M. Lynch, Edwin D. W. Moore, Gerrit Isenberg, Kevin E. Fogarty, Fredric S. Fay. Science, 268, (Jun. 9, 1995), pp. 1483-1487. I accept that most commercial software does not seek to give super-resolution, but we still need to distinguish what is deconvolution and what is not. Guy -----Original Message----- From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Jim Mansfield Sent: Friday, 23 March 2012 7:08 AM To: [hidden email] Subject: Re: Deconvolution of Transmission Images ***** To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy ***** Hi Guy, Maybe I'm just being overly fussy, or maybe I am not understanding some part of using deconvolution to resolve focal planes, but of course one can do real and actual deconvolution on brightfield images. The DeltaVision products all do widefield (i.e., non-confocal) deconvolution (albeit in FL) and that is very similar to brightfield deconvolution. Besides, confocal images are not completely independent; real point-spread functions do go to infinity on both sides of the focal plane (follow the direct light path through the pinhole from source to detector). Generating the PSF for brightfield deconvolution is much more difficult than for confocals, and so it is considerably more problematic than for a confocal, and there can be serious errors generated that way, but of course one can, in fact, really deconvolve brightfield images! Jim -----Original Message----- From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Guy Cox Sent: Thursday, March 22, 2012 2:58 PM To: [hidden email] Subject: Re: Deconvolution of Transmission Images ***** To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy ***** You cannot really deconvolve transmission images, because the image of a point is not independent of its neighbours, so each point is not in fact convolved with a psf. You can however isolate in-focus information from out of focus information and there are several algorithms for this. The basic idea is than in-focus detail will change a lot if we blur it whereas out of focus information won't change much. So you can extract 3D information, or construct an extended-focus image, but it isn't deconvolution. Guy -----Original Message----- From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Richard Cole Sent: Friday, 23 March 2012 1:32 AM To: [hidden email] Subject: Deconvolution of Transmission Images ***** To join, leave or search the confocal microscopy listserv, go to: http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy ***** You are correct, transmitted images are formed completely differently than fluorescent images, therefore the algorithms are separate. Having said that, we use Autoquant for both types of images. For DIC and Phase image we have also used a simple summed projection for qualitative purposes. Rich Richard Cole Research Scientist V Director: Advanced Light Microscopy Core Unit Wadsworth Center Research Assistant Professor Dept. of Biomedical Sciences School of Public Health State University of New York P.O. Box 509 Albany N.Y. 12201-0509 518-474-7048 Phone 518-474-4430 Fax Email <mailto:[hidden email]> [hidden email] Website <http://www.wadsworth.org/cores/alm/index.htm> www.wadsworth.org/cores/alm/index.htm IMPORTANT NOTICE: This e-mail and any attachments may contain confidential or sensitive information which is, or may be, legally privileged or otherwise protected by law from further disclosure. It is intended only for the addressee. If you received this in error or from someone who was not authorized to send it to you, please do not distribute, copy or use it or any attachments. Please notify the sender immediately by reply e-mail and delete this from your system. Thank you for your cooperation. |
«
Return to Confocal Microscopy List
|
1 view|%1 views
| Free forum by Nabble | Edit this page |