Limits...
eSIP: A Novel Solution-Based Sectioned Image Property Approach for Microscope Calibration.

Butzlaff M, Weigel A, Ponimaskin E, Zeug A - PLoS ONE (2015)

Bottom Line: However, modern microscope systems being as complex as they are, require very precise and appropriate calibration routines, in particular when quantitative measurements should be compared over longer time scales or between different setups.Furthermore, due to the fit of the complete profile, our method is less susceptible to noise.Generally, the extended SIP approach represents a simple and highly reproducible method, allowing setup independent calibration and alignment procedures, which is mandatory for advanced quantitative microscopy.

View Article: PubMed Central - PubMed

Affiliation: Cellular Neurophysiology, Center of Physiology, Hannover Medical School, Hannover, Germany.

ABSTRACT
Fluorescence confocal microscopy represents one of the central tools in modern sciences. Correspondingly, a growing amount of research relies on the development of novel microscopic methods. During the last decade numerous microscopic approaches were developed for the investigation of various scientific questions. Thereby, the former qualitative imaging methods became replaced by advanced quantitative methods to gain more and more information from a given sample. However, modern microscope systems being as complex as they are, require very precise and appropriate calibration routines, in particular when quantitative measurements should be compared over longer time scales or between different setups. Multispectral beads with sub-resolution size are often used to describe the point spread function and thus the optical properties of the microscope. More recently, a fluorescent layer was utilized to describe the axial profile for each pixel, which allows a spatially resolved characterization. However, fabrication of a thin fluorescent layer with matching refractive index is technically not solved yet. Therefore, we propose a novel type of calibration concept for sectioned image property (SIP) measurements which is based on fluorescent solution and makes the calibration concept available for a broader number of users. Compared to the previous approach, additional information can be obtained by application of this extended SIP chart approach, including penetration depth, detected number of photons, and illumination profile shape. Furthermore, due to the fit of the complete profile, our method is less susceptible to noise. Generally, the extended SIP approach represents a simple and highly reproducible method, allowing setup independent calibration and alignment procedures, which is mandatory for advanced quantitative microscopy.

No MeSH data available.


Related in: MedlinePlus

Extended section imaging property charts from layer and solution-based calibration samples.Comparable eSIP analysis using the same microscopic system with a homogenous fluorescent layer (A) and the solution-based calibration sample (B). Axial profiles (z profile) are shown in 5 different spatial positions (shown as a scheme within the plot). The 'Var over Mean' plot is scaled to maximum frequency in each Intensity bin to account for the different intensity distributions for layer and solution. The conversion factor (CF) is obtained by a linear fit (black line), which is 527.87±0.17 for the layer and 511.02±0.22 for the solution-based approach. The regression coefficient R2 is given to describe the goodness of fit. All eSIP parameters are shown as 3 dimensional plots: Intensity (A) in percent of the maximal photon number (pmax), axial position (z0) in μm, FWHM or steepness (ωFWHM) in μm, offset (I0) in digital levels (DL), skewness (s) or length constant (LC) in 1/μm and the Lorentz-Gauss fraction (mL). Additional scanning parameters: excitation 440 nm with 2% laser power, emission Channel 529 nm centre wavelength, pinhole 0.5 AU, detector gain 700V, 354 x 354 μm field of view with 512 x 512 pixels, pixel dwell time 1.58 μs, and axial spacing is in (A) 0.2 μm and in (B) sequentially 0.1 μm (10 μm around the glass/solution border), 1 μm (for adjacent 45 μm) and 5 μm (for adjacent 200 μm).To give an impression on the computational load: The analysis of the data from (B) took about 3 minutes using a scripting language (MATLAB) on a good equipped office calculator (Intel Core i7 3.2GHz 64 bit system with 32 GB). An analysis of the data from (A) using the SIP approach published by Brakenhoff et al. can be found in S2 Fig.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4526552&req=5

pone.0134980.g002: Extended section imaging property charts from layer and solution-based calibration samples.Comparable eSIP analysis using the same microscopic system with a homogenous fluorescent layer (A) and the solution-based calibration sample (B). Axial profiles (z profile) are shown in 5 different spatial positions (shown as a scheme within the plot). The 'Var over Mean' plot is scaled to maximum frequency in each Intensity bin to account for the different intensity distributions for layer and solution. The conversion factor (CF) is obtained by a linear fit (black line), which is 527.87±0.17 for the layer and 511.02±0.22 for the solution-based approach. The regression coefficient R2 is given to describe the goodness of fit. All eSIP parameters are shown as 3 dimensional plots: Intensity (A) in percent of the maximal photon number (pmax), axial position (z0) in μm, FWHM or steepness (ωFWHM) in μm, offset (I0) in digital levels (DL), skewness (s) or length constant (LC) in 1/μm and the Lorentz-Gauss fraction (mL). Additional scanning parameters: excitation 440 nm with 2% laser power, emission Channel 529 nm centre wavelength, pinhole 0.5 AU, detector gain 700V, 354 x 354 μm field of view with 512 x 512 pixels, pixel dwell time 1.58 μs, and axial spacing is in (A) 0.2 μm and in (B) sequentially 0.1 μm (10 μm around the glass/solution border), 1 μm (for adjacent 45 μm) and 5 μm (for adjacent 200 μm).To give an impression on the computational load: The analysis of the data from (B) took about 3 minutes using a scripting language (MATLAB) on a good equipped office calculator (Intel Core i7 3.2GHz 64 bit system with 32 GB). An analysis of the data from (A) using the SIP approach published by Brakenhoff et al. can be found in S2 Fig.

Mentions: Unless otherwise noted, experiments were carried out on an inverted and motorized microscope (Axio Observer Z.1) equipped with a 40x/1.20 W C−Apochromat objective. The attached laser scanning unit (LSM 780, Zeiss, Jena, Germany) enabled confocal imaging. For excitation, 440 nm and 488 nm lasers were used. Unless otherwise noted pinhole was set to 0.5 AU. Detailed scanning parameters are listed in the corresponding figures / figure captions. Images were recorded at 16-bit pixel depth. For analysis, they were processed in MATLAB scripts utilizing the eSIP formulas described above. As start parameters for the fitting we used the parameters obtained from the lookup approach [11]. The fluorescein/perylenediimid-layer reference layer, thickness ≈ 110 nm, nD ≈ 1.59 was a generous gift by J. M. Zwier. The fluorescent solution used in Fig 2 is a dilution of fluorescein in distilled water (Uranine AP, λem = 516 nm, OD = 1.37 for 1 cm at 487 nm). To account for a wider emission wavelength range needed for Fig 3 Rhodamine 6G (λem = 555 nm) was added and adjusted to an equal fluorescence intensity (both fluorophores from Applichem, Darmstadt, Germany). The solution was imaged in an ibidi μ-Dish with 170±5 μm glass bottom (ibidi GmbH, Martinsried, Germany).


eSIP: A Novel Solution-Based Sectioned Image Property Approach for Microscope Calibration.

Butzlaff M, Weigel A, Ponimaskin E, Zeug A - PLoS ONE (2015)

Extended section imaging property charts from layer and solution-based calibration samples.Comparable eSIP analysis using the same microscopic system with a homogenous fluorescent layer (A) and the solution-based calibration sample (B). Axial profiles (z profile) are shown in 5 different spatial positions (shown as a scheme within the plot). The 'Var over Mean' plot is scaled to maximum frequency in each Intensity bin to account for the different intensity distributions for layer and solution. The conversion factor (CF) is obtained by a linear fit (black line), which is 527.87±0.17 for the layer and 511.02±0.22 for the solution-based approach. The regression coefficient R2 is given to describe the goodness of fit. All eSIP parameters are shown as 3 dimensional plots: Intensity (A) in percent of the maximal photon number (pmax), axial position (z0) in μm, FWHM or steepness (ωFWHM) in μm, offset (I0) in digital levels (DL), skewness (s) or length constant (LC) in 1/μm and the Lorentz-Gauss fraction (mL). Additional scanning parameters: excitation 440 nm with 2% laser power, emission Channel 529 nm centre wavelength, pinhole 0.5 AU, detector gain 700V, 354 x 354 μm field of view with 512 x 512 pixels, pixel dwell time 1.58 μs, and axial spacing is in (A) 0.2 μm and in (B) sequentially 0.1 μm (10 μm around the glass/solution border), 1 μm (for adjacent 45 μm) and 5 μm (for adjacent 200 μm).To give an impression on the computational load: The analysis of the data from (B) took about 3 minutes using a scripting language (MATLAB) on a good equipped office calculator (Intel Core i7 3.2GHz 64 bit system with 32 GB). An analysis of the data from (A) using the SIP approach published by Brakenhoff et al. can be found in S2 Fig.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4526552&req=5

pone.0134980.g002: Extended section imaging property charts from layer and solution-based calibration samples.Comparable eSIP analysis using the same microscopic system with a homogenous fluorescent layer (A) and the solution-based calibration sample (B). Axial profiles (z profile) are shown in 5 different spatial positions (shown as a scheme within the plot). The 'Var over Mean' plot is scaled to maximum frequency in each Intensity bin to account for the different intensity distributions for layer and solution. The conversion factor (CF) is obtained by a linear fit (black line), which is 527.87±0.17 for the layer and 511.02±0.22 for the solution-based approach. The regression coefficient R2 is given to describe the goodness of fit. All eSIP parameters are shown as 3 dimensional plots: Intensity (A) in percent of the maximal photon number (pmax), axial position (z0) in μm, FWHM or steepness (ωFWHM) in μm, offset (I0) in digital levels (DL), skewness (s) or length constant (LC) in 1/μm and the Lorentz-Gauss fraction (mL). Additional scanning parameters: excitation 440 nm with 2% laser power, emission Channel 529 nm centre wavelength, pinhole 0.5 AU, detector gain 700V, 354 x 354 μm field of view with 512 x 512 pixels, pixel dwell time 1.58 μs, and axial spacing is in (A) 0.2 μm and in (B) sequentially 0.1 μm (10 μm around the glass/solution border), 1 μm (for adjacent 45 μm) and 5 μm (for adjacent 200 μm).To give an impression on the computational load: The analysis of the data from (B) took about 3 minutes using a scripting language (MATLAB) on a good equipped office calculator (Intel Core i7 3.2GHz 64 bit system with 32 GB). An analysis of the data from (A) using the SIP approach published by Brakenhoff et al. can be found in S2 Fig.
Mentions: Unless otherwise noted, experiments were carried out on an inverted and motorized microscope (Axio Observer Z.1) equipped with a 40x/1.20 W C−Apochromat objective. The attached laser scanning unit (LSM 780, Zeiss, Jena, Germany) enabled confocal imaging. For excitation, 440 nm and 488 nm lasers were used. Unless otherwise noted pinhole was set to 0.5 AU. Detailed scanning parameters are listed in the corresponding figures / figure captions. Images were recorded at 16-bit pixel depth. For analysis, they were processed in MATLAB scripts utilizing the eSIP formulas described above. As start parameters for the fitting we used the parameters obtained from the lookup approach [11]. The fluorescein/perylenediimid-layer reference layer, thickness ≈ 110 nm, nD ≈ 1.59 was a generous gift by J. M. Zwier. The fluorescent solution used in Fig 2 is a dilution of fluorescein in distilled water (Uranine AP, λem = 516 nm, OD = 1.37 for 1 cm at 487 nm). To account for a wider emission wavelength range needed for Fig 3 Rhodamine 6G (λem = 555 nm) was added and adjusted to an equal fluorescence intensity (both fluorophores from Applichem, Darmstadt, Germany). The solution was imaged in an ibidi μ-Dish with 170±5 μm glass bottom (ibidi GmbH, Martinsried, Germany).

Bottom Line: However, modern microscope systems being as complex as they are, require very precise and appropriate calibration routines, in particular when quantitative measurements should be compared over longer time scales or between different setups.Furthermore, due to the fit of the complete profile, our method is less susceptible to noise.Generally, the extended SIP approach represents a simple and highly reproducible method, allowing setup independent calibration and alignment procedures, which is mandatory for advanced quantitative microscopy.

View Article: PubMed Central - PubMed

Affiliation: Cellular Neurophysiology, Center of Physiology, Hannover Medical School, Hannover, Germany.

ABSTRACT
Fluorescence confocal microscopy represents one of the central tools in modern sciences. Correspondingly, a growing amount of research relies on the development of novel microscopic methods. During the last decade numerous microscopic approaches were developed for the investigation of various scientific questions. Thereby, the former qualitative imaging methods became replaced by advanced quantitative methods to gain more and more information from a given sample. However, modern microscope systems being as complex as they are, require very precise and appropriate calibration routines, in particular when quantitative measurements should be compared over longer time scales or between different setups. Multispectral beads with sub-resolution size are often used to describe the point spread function and thus the optical properties of the microscope. More recently, a fluorescent layer was utilized to describe the axial profile for each pixel, which allows a spatially resolved characterization. However, fabrication of a thin fluorescent layer with matching refractive index is technically not solved yet. Therefore, we propose a novel type of calibration concept for sectioned image property (SIP) measurements which is based on fluorescent solution and makes the calibration concept available for a broader number of users. Compared to the previous approach, additional information can be obtained by application of this extended SIP chart approach, including penetration depth, detected number of photons, and illumination profile shape. Furthermore, due to the fit of the complete profile, our method is less susceptible to noise. Generally, the extended SIP approach represents a simple and highly reproducible method, allowing setup independent calibration and alignment procedures, which is mandatory for advanced quantitative microscopy.

No MeSH data available.


Related in: MedlinePlus