1、Designation: F3294 18Standard Guide forPerforming Quantitative Fluorescence IntensityMeasurements in Cell-based Assays with WidefieldEpifluorescence Microscopy1This standard is issued under the fixed designation F3294; the number immediately following the designation indicates the year oforiginal ad
2、option or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. Asuperscript epsilon () indicates an editorial change since the last revision or reapproval.1. Scope1.1 This guidance document has been developed to facilitatethe collection
3、of microscopy images with an epifluorescencemicroscope that allow quantitative fluorescence measurementsto be extracted from the images. The document is tailored tocell biologists that often use fluorescent staining techniques tovisualize components of a cell-based experimental system.Quantitative c
4、omparison of the intensity data available in theseimages is only possible if the images are quantitative based onsound experimental design and appropriate operation of thedigital array detector, such as a charge coupled device (CCD)or a scientific complementary metal oxide semiconductor(sCMOS) or si
5、milar camera. Issues involving the array detectorand controller software settings including collection of darkcount images to estimate the offset, flat-field correction,background correction, benchmarking of the excitation lampand the fluorescent collection optics are considered.1.2 This document is
6、 developed around epifluorescencemicroscopy, but it is likely that many of the issues discussedhere are applicable to quantitative imaging in other fluores-cence microscopy systems such as fluorescence confocalmicroscopy. This guide is developed around single-color fluo-rescence microscopy imaging o
7、r multi-color imaging where themeasured fluorescence is spectrally well separated.1.3 Fluorescence intensity is a relative measurement anddoes not in itself have an associated SI unit. This documentdoes discuss metrology issues related to relative measurementsand experimental designs that may be req
8、uired to ensurequantitative fluorescence measurements are comparable afterchanging microscope, sample, and lamp configurations.1.4 This standard does not purport to address all of thesafety concerns, if any, associated with its use. It is theresponsibility of the user of this standard to establish a
9、ppro-priate safety, health, and environmental practices and deter-mine the applicability of regulatory limitations prior to use.1.5 This international standard was developed in accor-dance with internationally recognized principles on standard-ization established in the Decision on Principles for th
10、eDevelopment of International Standards, Guides and Recom-mendations issued by the World Trade Organization TechnicalBarriers to Trade (TBT) Committee.2. Referenced Documents2.1 ASTM Standards:2E131 Terminology Relating to Molecular SpectroscopyE284 Terminology of AppearanceE2186 Guide for Determini
11、ng DNA Single-Strand Damagein Eukaryotic Cells Using the Comet AssayE2642 Terminology for Scientific Charge-Coupled Device(CCD) DetectorsE2719 Guide for FluorescenceInstrument Calibration andQualificationE2825 Guide for Forensic Digital Image ProcessingF2944 Test Method for Automated Colony Forming
12、Unit(CFU) AssaysImage Acquisition and Analysis Methodfor Enumerating and Characterizing Cells and Colonies inCultureF2997 Practice for Quantification of Calcium Deposits inOsteogenic Culture of Progenitor Cells Using FluorescentImage AnalysisF2998 Guide for Using Fluorescence Microscopy to Quan-tify
13、 the Spread Area of Fixed Cells2.2 ISO Standards:3ISO 13653 Measurement of relative irradiance in the imagefieldISO/IEC 10918-1:1994 Digital compression and coding ofcontinuous-tone still images: Requirements and guidelinesISO/TR 12033:2009 Guidance for the selection of documentimage compression met
14、hods1This guide is under the jurisdiction of ASTM Committee F04 on Medical andSurgical Materials and Devices and is the direct responsibility of SubcommitteeF04.46 on Cell Signaling.Current edition approved Oct. 1, 2018. Published October 2018. DOI: 10.1520/F3294-18.2For referenced ASTM standards, v
15、isit the ASTM website, www.astm.org, orcontact ASTM Customer Service at serviceastm.org. For Annual Book of ASTMStandards volume information, refer to the standards Document Summary page onthe ASTM website.3Available from American National Standards Institute (ANSI), 25 W. 43rd St.,4th Floor, New Yo
16、rk, NY 10036, http:/www.ansi.org.Copyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United StatesThis international standard was developed in accordance with internationally recognized principles on standardization established in the Decision on Princ
17、iples for theDevelopment of International Standards, Guides and Recommendations issued by the World Trade Organization Technical Barriers to Trade (TBT) Committee.12.3 Other Documents:SWGDE/SWGIT Glossary SWGDE and SWGIT Digital however, pixel sizes of 8, 13, 16, and 20 m square are alsoavailable. E
18、26423.1.15 radiant energy, nenergy transmitted as electromag-netic radiation. E2843.1.16 radiant flux, ,nthe time rate of flow of radiantenergy; radiant power. E2843.1.17 region of interest (ROI), nuser-defined portion ofthe image area in which data will be acquired. The remainderof the image area w
19、ill be discarded. E26424. Summary of Guide4.1 Wide-field fluorescence microscopy is an optical imag-ing technique that relies on illumination of the entire field ofview of a fluorescence microscope and simultaneous detectionof the emitted fluorescence from all or a sub-region of the fieldof view usi
20、ng a camera. The emitted fluorescence can bemeasured as an intensity value in fluorescence microscopy,which is computed by summing together the intensity valuesfrom a group of individual pixels in a digital image acquiredusing a digital camera, such as a CCD, sCMOS, or EMCCD.Arelative intensity meas
21、urement (RIM) is determined as the ratioof one intensity measurement to another, the result of whichshould be an accurate estimate of the ratio of the irradiancefrom part or all of a specimen to the irradiance from part or allof the same or another specimen.4.2 The quantitative comparison of RIMs ca
22、n be compro-mised or invalidated by many possible factors including thenon-uniformity of intensities across the field of view of themicroscope, the presence of an offset in the pixel values in therecorded digital image, the intensity signals in the imageexceeding the linear dynamic range of the came
23、ra, or theinaccurate recording of the pixel values in image data files dueto factors such as a lossy compression operation or unexpectedmodification of the pixel bit depth when saving each file. Thequantitative comparison of RIMs can also be compromised bylow signal-to-noise ratio of the measured li
24、ght intensities orinstability in the optical power of the illumination source.4Available from Scientific Working Group on Imaging Technology (SWGIT),http:/www.swgit.org5Available from U.S. Department of Health and Human Services, Food andDrug Administration (FDA), Center for Devices and Radiological
25、 Health (CDRH).https:/www.fda.gov/downloads/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/UCM435355.pdf. You may also send an e-mail request toCDRH-Guidancefda.hhs.gov to receive a copy of the guidance. Please use thedocument number 1400053 to identify the guidance you are requesting.
26、6Available from European Machine Vision Association. http:/www.emva.org/standards-technology/emva-1288/.F3294 1824.3 This guide provides a list of corrections and normaliza-tions that are required so that RIMs can be accurately com-pared within and between images. This guidance documentalso includes
27、 a list of references to peer-reviewed, publishedmethods that can aid in the qualification of instrument perfor-mance in Appendix X1. The collection of these resourcesshould be useful in the design of robust cell-based assays thatused quantitative fluorescence microscopy for data collection.5. Signi
28、ficance and Use5.1 Overview of Measurement SystemRelative intensitymeasurements made by widefield epifluorescence microscopyare used as part of cell-based assays to quantify attributes suchas the abundance of probe molecules (see ASTM F2997),fluorescently labeled antibodies, or fluorescence protein
29、re-porter molecules. The general procedure for quantifying rela-tive intensities is to acquire digital images, then to performimage analysis to segment objects and compute intensities. Theraw digital images acquired by epifluorescence microscopy arenot typically amenable to relative intensity quanti
30、ficationbecause of the factors listed in 4.2. This guide offers a checklistof potential sources of bias that are often present in fluorescentmicroscopy images and suggests approaches for storing andnormalizing raw image data to assure that computations areunbiased.5.2 Areas of ApplicationWidefield f
31、luorescence micros-copy is frequently used to measure the location and abundanceof fluorescent probe molecules within or between cells. Ininstances where RIM comparisons are made between a regionof interest (ROI) and another ROI, accurate normalizationprocedures are essential to the measurement proc
32、ess to mini-mize biased results. Example use cases where this guidancedocument may be applicable include:5.2.1 Characterization of cell cycle distribution by quanti-fying the abundance of DNA in individual cells (1).75.2.2 Measuring the area of positively stained mineralizeddeposits in cell cultures
33、 (ASTM F2997).5.2.3 Quantifying the spread area of fixed cells (ASTMF2998).5.2.4 Determining DNA damage in eukaryotic cells usingthe comet assay (ASTM E2186).5.2.5 The quantitation of a secondary fluorescent markerthat provides information related to the genotype, phenotype,biological activity, or b
34、iochemical features of a colony or cell(ASTM F2944).6. Measurement Bias6.1 Sources of bias in relative intensity measurements arelisted below:6.1.1 CCD BiasThe detectors used in widefield fluores-cence microscopy are typically scientific complementary metaloxide semiconductor (sCMOS) sensors, charge
35、-coupled de-vices (CCDs), electron multiplying (EM)-CCDs, or similartypes of arrayed photodetectors (2). Regardless of the detector,the recorded digital signal in the absence of incident light,called the CCD bias (a.k.a. bias current offset or dark counts),is often not zero. The CCD bias is an offse
36、t that is added toeach pixel in the digital image. Accurate determination of thevalue of the CCD bias is critical as it will need to be subtractedfrom the raw image.6.1.1.1 A CCD bias that is less than zero is problematic.Digital images are typically saved in files that store onlypositive integer va
37、lues. If the CCD bias is less than zero, anunknown offset is subtracted from each pixel value and relativeintensity comparisons will not be possible.6.1.2 Linear Dynamic RangeImages of the specimenunder evaluation must be collected with the signal within thelinear dynamic range of the detector. Sign
38、als that are at orbelow the noise floor of the camera will not be detected.Similarly, signals that are above the detector saturation are nolonger in the linear range of the camera and cannot be used inrelative intensity evaluations.6.1.3 Non-Uniform FieldThe intensities measured from auniformly fluo
39、rescent sample are typically not uniform acrossthe field of view. Field non-uniformities can arise from manyfactors, including non-uniform illumination, vignetting, andnon-uniformities in the detector (ISO 13653 and (3, 4). Thismeans that the measured fluorescence intensity is dependent onits positi
40、on within the field of view. If measurements are to bemade in a region of the field of view with uneven illumination,a flatfield correction should be applied.An appropriate flatfieldfield correction will result in measured intensities that are notdependent on their location within the field of view
41、and can becompared.6.1.4 Save Raw Images or Use Lossless CompressionMany of the software packages that are used for controllingimage capture from digital cameras offer the opportunity tosave images in a lossy compression format (e.g. jpeg). Thisform of compression can alter the intensity data in a n
42、on-linearfashion, leading to unpredictable biases in relative intensitymeasurements. It is best to save raw, non-compressed imagedata or a lossless compression format (e.g. tiff) for images thatare intended to be used to make relative intensity measure-ments.Additional modifications to the pixel val
43、ues of an imagecan occur due to unexpected settings on the image analysissoftware. For example, bit truncation or bit depth conversioncan occur on a saved image. It is worthwhile to evaluate thehistogram of pixel intensities on the saved image and compareit to the histogram of the collected image to
44、 ensure the imageis appropriately saved. A good resource for information onlossless and lossy image compression formats can be found inASTM E2825-12, ISO/TR 12033:2009, and ISO/IEC 10918-1:1994.7. Normalization Strategies for Sources of Bias7.1 CCD BiasTo measure the CCD bias, a dark framemust be co
45、llected in the absence of illumination incident on thedetector. This estimate for the CCD bias will only be accuratefor the detector settings for which the image was taken. If anydetector settings are changed, such as the temperature, gain, orbinning, the CCD bias estimate may no longer be valid and
46、may need to be remeasured. The CCD bias may change overthe course of a data acquisition run caused, for example, by the7The boldface numbers in parentheses refer to the list of references at the end ofthis standard.F3294 183camera reaching a steady-state operating temperature duringstart up. Therefo
47、re, consideration should be given to measuringthe CCD bias at the beginning and end of a data acquisition runto assure that no change in the CCD bias has occurred.7.1.1 A simple approach for estimating the CCD bias is toturn off the microscope epi-fluorescent illumination and ac-quire at least 5 ima
48、ges at the same exposure time and samemicroscope settings (objective, filter sets, etc.) as used for thequantitative fluorescence images of the samples.7.1.2 A more general approach is to turn off the microscopeepi-fluorescent illumination and acquire at least 5 images at 5or more exposure times. Pl
49、otting the mean intensity for eachdark frame versus exposure time can reveal the presence ofbackground light that is incident on the detector. A regressionanalysis of the mean intensity versus exposure time can be usedto estimate the CCD bias value for any exposure time. Anexample of the data and analysis that can be used to estimatethe dark frame counts is shown in Fig. 1.7.1.3 A careful measurement of local background in theimage of the sample is another approach that can be used tosubtract the CCD bias. In this case, the mean pixel value ofpixels near the obje
