CORRESPONDENCE

npg

© 2014 Nature America, Inc. All rights reserved.

PSFj: know your fluorescence microscope To the Editor: The performance of a fluorescence microscope depends largely on the quality of its optical elements, predominantly its objective lenses. However, this quality is not easy for users to assess. We wrote PSFj (http://www.knoplab.de/psfj/ and Supplementary Software) to enable automated analysis of a microscope’s performance with respect to resolution, chromatic aberrations and planarity across the entire image field. Modern fluorescence microscopy methods demand state-of-the art instruments that perform at their diffraction limit (the maximum theoretical resolution a given microscope is able to deliver when equipped with a specific set of optical elements) and that provide a flat image field over a wide range of wavelengths. Although high-performance objectives are optimized to fulfill these demands, variation in the quality and alignment of components in the microscope as well as potential ill-usage (scratches, dirty lenses, etc.) will unavoidably lead to deviations from optimal performance. The performance of objective lenses is typically assessed by measuring their point spread function (PSF), i.e., by employing published protocols1 to image a point-like object (usually a subresolution-sized fluorescent bead). PSF width and centroid position at different wavelengths and positions in the field of view (FOV) provide measures of resolution, chromatic aberrations and image field planarity and are usually analyzed using either open-source software (such as MetroloJ, http://imagejdocu.tudor.lu/doku.php?id=plugin :analysis:metroloj:start) or custom routines1. However, no protocol exists for the rapid quantification of these measures throughout the FOV, making it generally difficult and tedious to judge whether a microscope is performing optimally. As a consequence, a microscope’s performance is assessed only occasionally (if at all) and is often based on a few PSF measurements. These measurements are

i

ii

typically made in the center of the FOV (where microscopes generally perform best), providing only a glimpse—an often misleading one—of the actual performance. To profit from the development of cameras with large complementary metal-oxide semiconductor (CMOS) sensors that can capture most of the FOV, including the edges where objective lens performance often deteriorates, users must conduct a careful assessment of each instrument’s performance. PSFj provides an automated full-FOV analysis of resolution, chromatic aberrations and planarity based on image stacks acquired from typical PSF test slides with multicolor, subresolution fluorescent beads. Using robust routines, PSFj identifies all beads in the FOV and analyzes their fluorescence intensity distributions (Supplementary Note). For each individual intensity distribution, it retrieves a number of parameters (Fig. 1) including the centroid position (x0, y0, z0), the smallest and largest full-width at halfmaximum (FWHM) in the xy plane (FWHMmin and FWHMmax, respectively), the angle (q) between the x axis and the direction of FWHMmax, and the axial FWHM (FWHMz). FWHMmin and FWHMmax serve as measures for lateral resolution, whereas their ratio A = FWHMmin/FWHMmax and q describe the lateral resolution anisotropy. The axial positions (z0) of all the beads, which are attached to the surface of a coverslip, provide a measure of the planarity of the image field. In addition, dual-color analysis allows for the determination of wavelength-dependent differences in bead centroid position (Δx0(l1, l2), Δy0(l1, l2), Δz0(l1, l2)), which provides a measure for lateral and axial chromatic aberrations. For dual-camera systems, this also quantifies the degree of misalignment between the cameras. Keeping different applications of such measurements in mind, we designed PSFj to offer three reporting options (Supplementary Methods). Results are visualized directly in the form of pseudocolored heat maps (Fig. 1), providing a quick and comprehensive performance

iii

Elliptic Gaussian PSF fit

FWHMmin FWHMmax

FWHMz

A

θ

z0

Δρ(λ)

Δz(λ)

iv

z0 y0

M min

H FW

z y

x

Image stack(s) of fluorescent beads

x0

M min FWH ax A = FWHM m

M max

FWH

PSF analysis for all identified beads

Visualization of microscope performance for full FOV

v

Processing pipeline e.g., image registration and restoration

Documentation and data export

Figure 1 | Analysis of microscope performance using PSFj. (i) First, PSFj analyzes the image stack and retrieves subvolumes containing the images of individual beads. (ii) On the basis of these individual bead images, subsequent analyses then determine the optical performance of the system. (iii) Finally, measured parameters—resolution (FWHM), planarity and chromatic aberration—are visualized using heat maps that represent the full FOV. (iv,v) A report can be generated (iv), and data can be exported for additional applications such as image registration and restoration (v).

NATURE METHODS | VOL.11 NO.10 | OCTOBER 2014 | 981

npg

© 2014 Nature America, Inc. All rights reserved.

CORRESPONDENCE check and inviting routine use. Second, the software can generate a summary report of the current system performance or a full report containing all individual PSF measurements and associated fitting parameters. Third, a table with the extracted resolution, planarity and colocalization data can be exported. This can be used for subsequent analysis, such as in an image processing or restoration pipeline. In addition, an average PSF from a user-selectable region of interest can be exported, for example, for image deconvolution. We used PSFj to quantify the performance of various high– numerical aperture (NA) objectives and to track day-to-day and system-to-system variation. The results showed substantial performance differences and allowed us to identify strengths and weaknesses of individual objectives as well as general shortcomings (Supplementary Figs. 1 and 2). In particular, we found that whereas lateral resolution performance generally fell short (~20–30%), axial resolution often met or exceeded expectations from the scalar approximation of the PSF commonly used in textbooks2 (Supplementary Note). Planarity was usually well corrected with variations over the FOV below the axial resolution and allowed for the detection of tilted slides caused, for example, by dust particles or misaligned slide holders or stages. Axial chromatic shifts were usually small, with little variation across the FOV (Supplementary Table 1). In contrast, chromatic shifts often showed circular symmetry and increased toward the edge of the FOV, which is a sign of lateral chromatic aberrations. Day-to-day performance variation of most objectives was relatively small (~2–6%) and comparable to single-measurement FOV variations (Supplementary Table 2). Furthermore, testing a limited number of identical objectives identified objective-toobjective and microscope-to-microscope variations of about 10% (Supplementary Tables 3 and 4). The PSFj software is open source and based on libraries from various sources, including ImageJ3 and µManager4, and it runs as a stand-alone application on the three major operating systems (using Java). Note: Any Supplementary Information and Source Data files are available in the online version of the paper (doi:10.1038/nmeth.3102). ACKNOWLEDGMENTS We thank H. Lorenz and C. Hoerth from the Zentrum für Molekulare Biologie der Universität Heidelberg (ZMBH) imaging facility, and F. Bestvater and D. Krunic from the Deutsches Krebsforschungszentrum (DKFZ) light microscopy facility, for assistance in acquiring some of the image stacks, and a number of friends for testing PSFj and providing expert feedback. This work was supported in parts by a grant from the German Research Foundation (DFG), Sonderforschungsbereich (SFB) 1036 (TP10). COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests.

Patrick Theer1,2, Cyril Mongis1 & Michael Knop1,2 1Zentrum für Molekulare Biologie, Universität Heidelberg, Heidelberg, Germany. 2Deutsches Krebsforschungszentrum, Heidelberg, Germany.

e-mail: [email protected]

1. Cole, R.W., Jinadasa, T. & Brown, C.M. Nat. Protoc. 6, 1929–1941 (2011). 2. Inoué, S. in Handbook of Biological Confocal Microscopy 3rd edn. (ed. Pawley, J.B.) 1–16 (Springer, 2006). 3. Schneider, C.A., Rasband, W.S. & Eliceiri, K.W. Nat. Methods 9, 671–675 (2012). 4. Edelstein, A., Amodaj, N., Hoover, K., Vale, R. & Stuurman, N. Curr. Protoc. Mol. Biol. 92, 14.20 (2010).Neuronal morphometry directly from bitmap images

982 | VOL.11 NO.10 | OCTOBER 2014 | NATURE METHODS

Neuronal morphometry directly from bitmap images To the Editor: Neuroscientists measure the tree-like structures of neurons in order to better understand how neural circuits are constructed and how neural information is processed. In 1953, Donald Sholl published his well-known technique for quantitative analysis of the complex arbors of dendrites and axons1, but conventional methods still require reconstruction of arbors via time-consuming manual or semi-automated tracing from microscopy images. To bypass this reconstruction step and perform the Sholl technique directly on images instead, we developed Sholl Analysis (http://fiji.sc/Sholl), an open-source program for ImageJ/Fiji2 (Supplementary Fig. 1). The plug-in employs an improved algorithm to retrieve data from two- or three-dimensional (2D or 3D) bitmap images in any format supported by the Bio-Formats library (Supplementary Methods). It pairs this data retrieval with curve-fitting, regression analysis and statistical inference so that users can automatically extract a collection of Sholl-based metrics of arborization1,3 (Supplementary Note). Using individual cortical pyramidal neurons in 3D images, we found Sholl Analysis to be accurate when benchmarked against corresponding manual reconstructions (Supplementary Fig. 2). The method was also resilient to image degradation by simulated shot noise (Supplementary Fig. 3 and Supplementary Software). To further assess accuracy, and to explore the utility of Sholl Analysis in tackling neurons that are particularly slow to reconstruct manually, we studied cerebellar Purkinje cells in mice, which have large and intricate dendritic arbors. From tiled 3D image stacks of cerebellum (Fig. 1a), we selected seven Brainbow2.1-expressing Purkinje neurons and isolated their morphologies (Fig. 1b and Supplementary Note). We then used the Sholl Analysis software to retrieve ten metrics and found they were indistinguishable from those retrieved from manual reconstructions of the same 7 cells (Fig. 1c,d and Supplementary Methods). To probe the sensitivity of the Sholl Analysis software, we asked whether its metrics could be used to distinguish closelyrelated neocortical interneuron subtypes. Parvalbumin-positive (PV) interneurons in layer 5 of visual cortex can be morphologically classified into two subtypes on the basis of their axonal morphology: type 1 PV cells have ascending axons arborizing in layer 2/3, whereas axons of type 2 cells remain in layer 5 (ref. 4). Because their dendritic arbors are indistinguishable4, these two cell types otherwise appear highly similar (Fig. 1e,f). Using the Sholl Analysis software, we retrieved 18 metrics directly from 3D image stacks of 12 PV interneurons. We then used Ward’s hierarchical clustering based on these metrics to independently classify these cells (Fig. 1g and Supplementary Fig. 4). The 12 cells segregated into two groups: one group of five neurons and another of seven. We found that all the neurons but two were correctly classified, with one cell assigned incorrectly to each class (Fig. 1g). Thus, our use of the Sholl Analysis software to quantify arborization directly from bitmap images correctly identified 80–86% of cells. In agreement, linear Sholl plots of type 1 cells indicated more branching than was found for type 2 cells at a distance of 225–300 µm from the soma (Fig. 1h), which corresponds to