Life Sciences Commons^™

Gpu Deconvolution Wow Result On 2048x2048x32 Plane Z-Series, George Mcnamara

George McNamara

GPU Deconvolution WOW result on 2048x2048x32 plane Z-series ... formerly bad academic code ("you get what you pay for") now impressive

Alternative title: "instant gratification quantitative deconvolution fluorescence microscopy".

http://works.bepress.com/gmcnamara/55/

Please see "74"

http://works.bepress.com/gmcnamara/74/

for 32-bit images from this project (bepress file size limitation prevented me from including them in this ZIP archive).

//

Summary: Deconvolution microscopy has historically been painfully slow. The early vendors were:

- Scanalytics (Carrington and Fay), commercialized to try to sell expensive, specialized array processors made by CSPI (the CSPI box likely had less computing power than a first gen smartphone).

- Applied Precision (Sedat …

32-Bit Deconvolution Files Accompanying 55 Page, George Mcnamara

George McNamara

ZIP file with 32-bit images accompanying

http://works.bepress.com/gmcnamara/55

Bepress apparently has an ~1 Gigabyte file upload limit so I have placed the 32-bit images of greatest interest here.

See Word document inside for details.

Mcnamara 2010 Metamatters Newletter Series - Rogers Pmn The Chase And Temporal Area Maps, George Mcnamara

George McNamara

My four 2010 MetaMorph MetaMatters newsletter articles on "The Chase" (video available for download here on my bepress site) and history / use of temporal area maps.

Light Microscopy Reflection Focusing Cells And Coverglass, George Mcnamara

George McNamara

Light Microscopy Reflection Focusing Cells and Coverglass 20150520Wed

Many light microscopes have the capability of using camera based brightfield, phase contrast, or fluorescence to find the most contrasty specimen focus. When I worked for UIC (1992-1997) we offered this with a combination of Ludl MAC2000 controller, Z-drive, video card thingy, and video camera (most customers at the time used analog video cameras).

Many automated light microscopes now have clever -- and pricey -- autofocus systems based on an NIR LED structured illumination reflecting light off the bottom of the coverglass.

I encourage a simpler approach - my hardware details are …

Focus On Microscopy Fom 2006-2015 Abstracts, George Mcnamara

George McNamara

Focus On Microscopy FOM 2006-2015 abstracts

FOM's are cutting edge microscopy meetings. Their abstracts are posted online soon after the meeting. FOM2015's abstracts are at http://www.focusonmicroscopy.org/2015/program.html

I have been downloading the single PDF abstracts, and making a single large PDF. every year. I figure the "annual organized" PDF may be useful to others, so am posting a ZIP file here containing: FOM 2006 FOM 2007 FOM 2008 FOM 2009 FOM 2010 FOM 2011 FOM 2012 FOM 2013 FOM 2014 FOM 2015 I note the abstracts do not contain copyright notices, but anyone using should "give credit where credit is due" …

Gadph Single Molecule Mrna Fish 20141104tue, George Mcnamara

George McNamara

GADPH single molecule mRNA FISH 20141104Tue

Raw data on NVidia TITAN GPU card deconvolution using original Bruce and Butte 2013 Optics Express software running inside Fiji ImageJ.

B&B GPU Deconvolution Reference:

http://www.opticsinfobase.org/oe/fulltext.cfm?uri=oe-21-4-4766&id=249375

Acquired at 63x1x using leica plan apo 1.4 NA objective lens on Cooper lab Leica DMI6000 microscope, SOLA (initial release) LED lamp using only "green" LED, 554x Semrock exciter filter (in ASI ex wheel), Leica N3 (Cy3) filter cube, center 600 nm emission wavelength.

Hamamatsu FLASH4.0 sCMOS, 100 ms exposure time so 6.4 seconds acquisition time (64 planes X 4 seconds). Acquired with MetaMorph 7.8.8.

XY 100 nm …

Gpu Deconvolution Good Result On 2048x2048x32 Plane Z-Series (Update From Dataset 56), George Mcnamara

George McNamara

GPU Deconvolution Good result on 2048x2048x32 plane Z-series (update from dataset 56)

This web page and download is

http://works.bepress.com/gmcnamara/62

Marc Bruce sent me the file here (web page 62) that is a preliminary bug fix (apparently of NVidia CUDA FFT code!!!) from the original 'raw' Z-series that is part of the

http://works.bepress.com/gmcnamara/56

web page download. Please see that web page for details on the project (GADPH single molecule mRNA detection with Stellaris FISH probe sets).

Marc is:

MARC BRUCE, PHD

CTO and Co-Founder

MICROVOLUTION, LLC

http://www.microvolution.com/

//

The 730 Mb ZIP download includes 3 files:

* 32-bit TIFF series sent …

Timelapse Data - Gm Imaging Cytometer Needs Stitching Part 1 Of 2, George Mcnamara

George McNamara

Timelapse data - GM imaging cytometer needs stitching part 1 of 2

Imaging challenge: align multiple time series over time to overcome "stage slop" in the imaging cytometer used in this experiment.

http://works.bepress.com/gmcnamara/53 (this)

and

http://works.bepress.com/gmcnamara/54 (next)

are zip files containing 8-bit stacks (MetaMorph multiplane TIFF files) acquired on an imaging cytometer.

"53" is the brightfield data, raw and automatic aligned stacks in MetaMorph (Apps menu, MM7.8.6).

"54" is GFP fluorescence, raw and automatic aligned stacks in MetaMorph (Apps menu, MM7.8.6). The GFP fluorescence decreases over time due to photobleaching. GFP disappears from cells when they die (soluble GFP diffusing …

Timelapse Data - Gm Imaging Cytometer Needs Stitching Part 2 Of 2, George Mcnamara

George McNamara

Timelapse data - GM imaging cytometer needs stitching part 2 of 2

Imaging challenge: align multiple time series over time to overcome "stage slop" in the imaging cytometer used in this experiment.

http://works.bepress.com/gmcnamara/53 (previous) and http://works.bepress.com/gmcnamara/54 (this)

are zip files containing 8-bit stacks (MetaMorph multiplane TIFF files) acquired on an imaging cytometer.

"53" is the brightfield data, raw and automatic aligned stacks in MetaMorph (Apps menu, MM7.8.6). "54" is GFP fluorescence, raw and automatic aligned stacks in MetaMorph (Apps menu, MM7.8.6). The GFP fluorescence decreases over time due to photobleaching. GFP disappears from cells when they die (soluble GFP diffusing …

Pathscan Enabler At Md Anderson Cancer Center, George Mcnamara

George McNamara

McNamara 20140703 - Additional Pathscan and Tiki_Goddess related resources

http://works.bepress.com/gmcnamara/52

http://home.earthlink.net/~tiki_goddess/

http://works.bepress.com/gmcnamara/1/

http://works.bepress.com/gmcnamara/11/

https://www.linkedin.com/in/georgemcnamara

Our Pathscan Enabler IV, delivered July 2014, uses the QuickScan software to scan a histology slide in "two clicks" (after loading the

slide):

1. Push the QuickScan button on the front of the Pathscan Enabler.

2. Click the Scan button in the QuickScan pop-up application (optional: change scan area in the Prescan image window).

The image gets saved as a TIFF file to the Windows 7 (64-bit) desktop.

I have been using Pathscan Enabler's since version I in 2000. See Chantrain et al 2003:

Chantrain CF, …

Tattletales And T-Bow Update 20140602mon, George Mcnamara

George McNamara

Tattletales and T-Bow Update 20140602Mon

http://works.bepress.com/gmcnamara/42

Please see also http://works.bepress.com/gmcnamara/26

Tattletales: multiplex fluorescent protein biosensors by spatial localization with TALE-FPs, Cas9-FPs, ZF-FPs, LacI-FPs, TetR-FPs, etc.

T-Bow: Rainbow T-cells and Tumor cells (and ES cells, iPS cells, other cells and organisms). You can think of this as "Brainbow meets TALENs/Cas9/ZFNs/other DNA sequence specific binding proteins".

If not familiar with Brainbow, see

http://en.wikipedia.org/wiki/Brainbow

If not familiar with TALENs, Cas9, etc, see

http://www.addgene.org/genome_engineering/

Big idea: localizing fluorescent proteins - and/or Nano-Lanterns (Take Nagai) - to tandem repeat arrays - is a great way to improve signal to noise ratio compared to the usual …

Pubspectra To Do Data, George Mcnamara

George McNamara

PubSpectra raw data that I had in my "to do" folder to add to the PubSpectra data set but never did.

PubSpectra has over 2000 spectra in it (last updated 2006) in Excel "big file" format - the XLSX file inside the zip file at

http://works.bepress.com/gmcnamara/9/

Anyone is welcome to add and organize new data yourselves. I used Un-Scan-It from Silk Scientific (www.silkscientific.com).

More importantly, I strongly adopting my PubSpectra format of 1 nm wavelength intervals, with the row number corresponding to nm. For example, 400 nm data appears in row 400.

Much of Urse Utzinger's "Spectra Arizona" data,

http://www.spectra.arizona.edu/ …

Timelapse Data - Gm Vandana 20140502fri Part 2 Of 2, George Mcnamara

George McNamara

Timelapse data - GM Vandana 20140502Fri part 2 of 2

(positions 4 and 5 of 5).

Data from a project with Vandana Kaul, Univ Houston. Vandana and GM know the details of this experiment. Nikon BioStation IM, 20x effective magnification (Nikon 40x lens with de-zoom), 800x600 pixels (by 2x2 binning). Five positions (P1 ... P5), acquired at 1 frame per minute, playback 30 fps.

Videos 1, 2 and 3 are in "part 1" at

http://works.bepress.com/gmcnamara/48/

Timelapse Data - Gm Vandana 20140502fri Part 1 Of 2, George Mcnamara

George McNamara

Timelapse data - GM Vandana 20140502Fri part 1 of 2

(positions 1, 2 and 3).

Data from a project with Vandana Kaul, Univ Houston. Vandana and GM know the details of this experiment. Nikon BioStation IM, 20x effective magnification (Nikon 40x lens with de-zoom), 800x600 pixels (by 2x2 binning). Five positions (P1 ... P5), acquired at 1 frame per minute, playback 30 fps.

Image Z-Series Wrt Nvidia Titan 6 Gb Gpu Cuda Deconvolution, George Mcnamara

George McNamara

Image Z-series wrt NVidia Titan 6 Gb GPU CUDA Deconvolution

This ZIP file contains raw and deconvolved data series from George McNamara, Laurence J.N. Cooper lab, M.D. Anderson Cancer Center. I have also included our lab instructions, and a screenshot of the current CUDA deconvolution settings (for dry objective lenses, Manish Butte wrote me to use the immersion medium refractive index, air = 1.0).

Instrument:

Leica DMI6000 inverted fluorescence microscope, Lumencor SOLA LED light source, Leica "L5" GFP filter cube (480/40ex, 505dm, 527/30em).

Leica 20x/0.75 NA objective lens, 1.6x optovar, for 325 nm XY piel size. For this dataset, I …

Cell Morphometry, George Mcnamara

George McNamara

Cell Morphometry ZIP content by George McNamara

http://works.bepress.com/gmcnamara/41

Robert Murphy's TypIC (typical cell Chooser,

http://murphylab.web.cmu.edu/services/TypIC/

now superseded by PSLID and SLIF

http://murphylab.web.cmu.edu/services/ ), was a "game changer" for me with respect to cell shape analysis. Rather than trying to compute the average of (say) a triangle and a pentagon ... which might result in a square, or a rectange, or some bizarre quadrilateral ... R.M. advocated using the median. OK, in a 2 member dataset this would result in averaging the two shapes (if use the standard way to calculate median of even number datasets), but this could be avoided …

More Fish (1 Of 4), George Mcnamara

George McNamara

More Stellaris FRISH and GPU Deconvolution, 20140313Thur - part 1 of 4

Please see other content in the FISH Imaging section of

http://works.bepress.com/gmcnamara/subject_areas.html

for details on the microscope. Briefly: Leica DMI6000 inverted microscope, SOLA lamp, Hamamatsu FLASH4.0 sCMOS camera, 100 nm XY, 200 nm Z pixels, 32 planes (16 um Z), MetaMorph 7.8 acquisition software.

Four zip files:

http://works.bepress.com/gmcnamara/43

http://works.bepress.com/gmcnamara/44

http://works.bepress.com/gmcnamara/45

http://works.bepress.com/gmcnamara/46

with two dataset (001, 002), raw and deconvolved using Bruce & Butte GPU deconvolution in Fiji ImageJ,

http://www.ncbi.nlm.nih.gov/pubmed/23482010

http://www.opticsinfobase.org/vjbo/fulltext.cfm?uri=oe-21-4-4766&id=249375 (free download, as are all Optics Express papers).

http://tcell.stanford.edu/software.html

http://lists.umn.edu/cgi-bin/wa?A2=ind1312&L=confocalmicroscopy&P=5534

Note: "001 raw" (bepress 43) contains my "imaging SOP" …

More Fish (2 Of 4), George Mcnamara

George McNamara

More Stellaris FISH and GPU Deconvolution, 20140313Thur - part 2 of 4

Please see other content in the FISH Imaging section of

http://works.bepress.com/gmcnamara/subject_areas.html

for details on the microscope. Briefly: Leica DMI6000 inverted microscope, SOLA lamp, Hamamatsu FLASH4.0 sCMOS camera, 100 nm XY, 200 nm Z pixels, 32 planes (16 um Z), MetaMorph 7.8 acquisition software.

Four zip files:

http://works.bepress.com/gmcnamara/43

http://works.bepress.com/gmcnamara/44

http://works.bepress.com/gmcnamara/45

http://works.bepress.com/gmcnamara/46

with two dataset (001, 002), raw and deconvolved using Bruce & Butte GPU deconvolution in Fiji ImageJ,

http://www.ncbi.nlm.nih.gov/pubmed/23482010

http://www.opticsinfobase.org/vjbo/fulltext.cfm?uri=oe-21-4-4766&id=249375 (free download, as are all Optics Express papers).

http://tcell.stanford.edu/software.html

http://lists.umn.edu/cgi-bin/wa?A2=ind1312&L=confocalmicroscopy&P=5534

Note: "001 raw" (bepress 43) contains my "imaging SOP" …

More Fish (3 Of 4), George Mcnamara

George McNamara

More Stellaris FISH and GPU Deconvolution, 20140313Thur - part 3 of 4

Please see other content in the FISH Imaging section of

http://works.bepress.com/gmcnamara/subject_areas.html

for details on the microscope. Briefly: Leica DMI6000 inverted microscope, SOLA lamp, Hamamatsu FLASH4.0 sCMOS camera, 100 nm XY, 200 nm Z pixels, 32 planes (16 um Z), MetaMorph 7.8 acquisition software.

Four zip files:

http://works.bepress.com/gmcnamara/43

http://works.bepress.com/gmcnamara/44

http://works.bepress.com/gmcnamara/45

http://works.bepress.com/gmcnamara/46

with two dataset (001, 002), raw and deconvolved using Bruce & Butte GPU deconvolution in Fiji ImageJ,

http://www.ncbi.nlm.nih.gov/pubmed/23482010

http://www.opticsinfobase.org/vjbo/fulltext.cfm?uri=oe-21-4-4766&id=249375 (free download, as are all Optics Express papers).

http://tcell.stanford.edu/software.html

http://lists.umn.edu/cgi-bin/wa?A2=ind1312&L=confocalmicroscopy&P=5534

Note: "001 raw" (bepress 43) contains my "imaging SOP" …

More Fish (4 Of 4), George Mcnamara

George McNamara

More Stellaris FISH and GPU Deconvolution, 20140313Thur - part 4 of 4

Please see other content in the FISH Imaging section of

http://works.bepress.com/gmcnamara/subject_areas.html

for details on the microscope. Briefly: Leica DMI6000 inverted microscope, SOLA lamp, Hamamatsu FLASH4.0 sCMOS camera, 100 nm XY, 200 nm Z pixels, 32 planes (16 um Z), MetaMorph 7.8 acquisition software.

Four zip files:

http://works.bepress.com/gmcnamara/43

http://works.bepress.com/gmcnamara/44

http://works.bepress.com/gmcnamara/45

http://works.bepress.com/gmcnamara/46

with two dataset (001, 002), raw and deconvolved using Bruce & Butte GPU deconvolution in Fiji ImageJ,

http://www.ncbi.nlm.nih.gov/pubmed/23482010

http://www.opticsinfobase.org/vjbo/fulltext.cfm?uri=oe-21-4-4766&id=249375 (free download, as are all Optics Express papers).

http://tcell.stanford.edu/software.html

http://lists.umn.edu/cgi-bin/wa?A2=ind1312&L=confocalmicroscopy&P=5534

Note: "001 raw" (bepress 43) contains my "imaging SOP" …

3b Project Data - Raw Tiff Series And Cropped Series, George Mcnamara

George McNamara

Fluorescence microscopy data acquired for 3B Bayesian localization microscopy.

Settings: 100 nm XY pixel size. 50 millisecond exposures, MetaMorph stream acquisition mode (as fast as possible). Stellaris FISH mammalian cells labeled with Quasar 670 fluorophores, one molecule per 20mer FISH probe, ~48 probes to TOP1 mRNA (Topoisomerase 1).

More on the 3B project and Stellaris FISH data at http://works.bepress.com/gmcnamara/38/

and at Susan Cox's web sites http://www.coxphysics.com/3b/ http://superresolved.com/

Time Series Data For 3b Image Processing, George Mcnamara

George McNamara

Time series data for 3B image processing

Four time series data sets of Stellaris single molecules RNA FISH (fluorescence in situ hybridization). All four datasets are 301 imaqe planes. FISH probes are to TOP1 mRNA (Topoisomerase I)in Saos osterosarcoma cells.

One dataset is 10 millisecond exposures (very dim), acquired in Streaming acquisition mode (no hardware overhead).

The other three datasets are a three consecutive planes of the same XY field of view.

I have included image processing results for:

PiMP - a fast method developed by Sebastian Munck et al 2012, http://www.ncbi.nlm.nih.gov/pubmed/?term=22357945

and greatly improved performance by Glen Macdonald (Seattle). …

Blood Vessel Z-Series Image Data Set From Confocal Microscopy 2013 Book, George Mcnamara

George McNamara

Blood Vessel Z-series image data set from Confocal Microscopy 2013 book chapter:

McNamara G, Yanai A, Khankaldyyan V, Laug WE, Boden J, Webster K, Li Y, Wen R. Low magnification confocal microscopy of tumor angiogenesis. Methods Mol Biol. 2014; 1075: 149-75.

doi: 10.1007/978-1-60761-847-8_6. PubMed PMID: 24052350.

See also:

Jeffrey Boden, Jianqin Wei, George McNamara, Hans Layman, Midhat Abdulreda, Fotios Andreopoulos, and Keith A. Webster. Whole-mount imaging of the mouse hindlimb vasculature using the lipophilic carbocyanine dye DiI. Biotechniques Rapid Dispatches, July 2012, pp. 1–4

http://www.biotechniques.com/rapiddispatches/Whole-mount-imaging-of-the-mouse-hindlimb-vasculature-using-the-lipophilic-carbocyanine-dye-DiI./biotechniques-333144.html

Supplemental file:

http://www.biotechniques.com/multimedia/archive/00182/BTNRD-BM-WebsterSUP_182355a.pdf

You may find useful to download Leica LAS AF Lite from

ftp://ftp.llt.de/softlib/LAS_AF_Lite/ …

Stellaris Fish Workshop Gadph And Dapi Z-Series, George Mcnamara

George McNamara

Stellaris FISH workshop GADPH and DAPI Z-series at MD Anderson Cancer Center, Houston, TX.

The zip file contains raw and GPU deconvolved image data from a workshop Biosearch Technologies conducted for MDACC researchers in December 2013. Image data was acquired on a Leica DMI6000 microscope with Lumencor SOLA light engine, DAPI and Cy5 filter cubes, Hamamatsu ORCA FLASH4.0 sCMOS camera (500 ms exposure time per plane for Quasar 670).

Pixel size 100 nm XY.

Z-step size 200 nm.

32 planes (power of 2 is optimal for GPU deconvolution). With 500 ms exposure time, the Quasar 670 GADPH FISH probes images …

Geoexes - My Dos Executables - Use At Your Own Risk, George Mcnamara

George McNamara

The downloadable ZIP file contains old Turbo Pascal programs I wrote. These ran in Microscoft DOS and in early Windows (ex. Windows 3.1) DOS shells. They may or may not work in recent (Windows 2000, XP, Vista, 7) cmd or command "shells". No warranty. Use at your own risk. I do not expect to ever program in Turbo Pascal again so if anyone wants to "take over" my old code (and adjust copyirght notices, warnings, etc), please email me. In particular, MATH.exe had 10,000 lines of pascal code in it. Another program is LUTSconv.exe which converted Fractint MAP lookup tables …

Stellaris Fishing 20131125mon Part 1 Of 2, George Mcnamara

George McNamara

Stellaris FISH dataset using three FISH probe sets. Slides courtesy of Biosearch Technologies,

https://www.biosearchtech.com/store/product.aspx?catid=224,318,324

see http://stellarisfish.smugmug.com/ for online gallery by Biosearch.

This experiment was to evaluate the crosstalk between the Biosearch fluorophores:

Quasar 570

CAL Fluor Red 610 (CFR 610)

Quasar 670

DAPI (DNA counterstain)

Autofluorescence (green, but sometimes showing up in other channels).

and our lab's Leica DMI6000 fluorescence microscope with Leica filter sets:

DAPI

GFP (L5)

Cy3 (N3)

Texas Red (TxRed2)

Cy5 (Y5)

I also acquired green channel and red channel with exciter filters in our ASI excitation wheel:

GFP + 492 exciter

Texas Red (TxRed2) + 572 …

Stellaris Fishing 20131125mon Part 2 Of 2, George Mcnamara

George McNamara

Stellaris FISH dataset using three FISH probe sets.

Slides courtesy of Biosearch Technologies,

https://www.biosearchtech.com/store/product.aspx?catid=224,318,324

see http://stellarisfish.smugmug.com/ for online gallery by Biosearch.

This experiment was to evaluate the crosstalk between the Biosearch fluorophores:

Quasar 570

CAL Fluor Red 610 (CFR 610)

Quasar 670

DAPI (DNA counterstain)

Autofluorescence (green, but sometimes showing up in other channels).

and our lab's Leica DMI6000 fluorescence microscope with Leica filter sets:

DAPI

GFP (L5)

Cy3 (N3)

Texas Red (TxRed2)

Cy5 (Y5)

I also acquired green channel and red channel with exciter filters in our ASI excitation wheel:

GFP + 492 exciter

Texas Red (TxRed2) + 572 …

Leica Microscope Gpu Deconvolution Stellaris Fish Dataset #1, George Mcnamara

George McNamara

McNamara 20131101Fri Leica widefield microscope CUDA Deconvolution Stellaris FISH probe cultured cells dataset #1.zip

A text file in the zip archive has experiment details. I am posting this online so that researchers - whether academic or commercial - can evaluate the acquired data, the Bruce and Butte 2013 Optics Express ( http://www.opticsinfobase.org/oe/fulltext.cfm?uri=oe-21-4-4766 ) deconvolution result (note: I may not have used optimal settings), and to compare these deconvolution results to other methods. If anyone generates alternative spatial deconvolution output, such as from: * SVI Huygens * Media Cy AutoQuant * Agard's ER-Decon (Arigovindan 2013 PNAS) * Vicidomini SGP GPU Deconv …

Hamamatsu Flash4.0 Scmos Exposure Time Series, George Mcnamara

George McNamara

Hamamatsu FLASH4.0 scientific cMOS camera exposure time series are pairs of images of:

1 millisecond (00,001ms series)

10 millisecond (00,010ms series)

100 millisecond (00,100ms series)

1,000 millisecond (01,000ms series)

4,000 millisecond (04,000ms series)

10,000 millisecond (10,000ms series)

I also included:

* difference images (exposure 2 minus exposure 1 plus 100 intensity values).

* a series of eleven 1 second (1,000 ms) exposure time images in a multi-plane TIFF file (different images than the pair of 1,000ms images above).

* Stack Arithmetic: Median, Average, Minimum, Maximum, of the eleven plane series (Stack Arithmetic is a MetaMorph command).

These images were acquired …

Flash4 Dark Reference Images, George Mcnamara

George McNamara

Hamamatsu FLASH4.0 dark reference images, acquired with 10 second exposure times, no light to camera. Camera offset (set by Hamamatsu( is ~100 (the average intensity of the first image is always ~1 intensity level higher - an odd feature, but trivial in practice for a 16-bit camera).

George McNamara, Ph.D.

Single Cells Analyst at L.J.N. Cooper Lab

University of Texas M.D. Anderson Cancer Center