PROOFREADING INSTRUCTIONS Microscopy and

1500017 11

PROOFREADING INSTRUCTIONS

North American Branch 32 Avenue of the Americas New York, NY 10013-2473, USA Tel : Fax : E-mail:

212 337-5978 212 337-5959 [email protected]

Dear Microscopy and Microanalysis Contributor: Attached is a PDF page proof of your introduction/article/book review scheduled to be published in:

Microscopy and Microanalysis Please follow these procedures: 1. Proofreading: Proofread your article carefully. Check especially the spellings of names and places as well as the accuracy of dates and numbers. Please answer all queries that may appear on a separate page. 2. Text: Changes in the text are limited to typographical and factual errors. Rewriting or other stylistic changes are not permitted. Contributors may be charged for excessive author alterations, and publication of your article may be delayed. 3. Corrections: Please respond with an e-mail message to Morrell Gillette at [email protected], identifying the correction by page number, column, paragraph, and line. Please indicate the present errant copy followed by the correct copy. The corrections to the proofs should be sent within 2 days of receipt. Corrections can also be entered into the PDF as embedded comments. 4. References: If the query involves a reference entry, please arrange the new entry into the correct format. In order to prepare the HTML full-text files for online viewing, the complete information including correct spelling of author names, year, titles, publisher, city of publication, page range, and so forth are needed. 5. Figures: Review the figure reproductions on the page proofs to see if important features have been well represented. If something seems out of order, indicate the errant features in a cover letter. Should it be necessary that new electronic copy of the figures (in PDF, TIFF, or EPS) or text (Word or LaTeX) will have to be provided, please indicate which application is being used. 6. Offprints or Bound Copies: Free offprints are longer being offered. However, a message with a link to access a free PDF of your paper will be sent to you. To order reprints or offprints of your article or printed copies of the issue, please visit the Cambridge University Reprint Order Center online at: www.sheridan.com/cup/eoc 7. Delay in response: Please note that failure to respond in a timely fashion may delay publication of your article or may require publication without your corrections. Thank you for your prompt attention to these proofs. If you have any questions, please feel free to contact Morrell Gillette by e-mail: [email protected], or at (212) 337-5978. Thanks. Best Regards, Morrell Gillette, Production Editor

QUERY FORM MAM Manuscript ID

[Art. Id: 1500017]

Author Editor Publisher

Journal: Microscopy And Microanalysis

Author :- The following queries have arisen during the editing of your manuscript. Please answer queries by making the requisite corrections at the appropriate positions in the text.

Query No

Nature of Query

Q1

The distinction between surnames can be ambiguous, therefore to ensure accurate tagging for indexing purposes online (eg for PubMed entries), please check that the highlighted surnames have been correctly identified, that all names are in the correct order and spelt correctly.

Q2

Please limit the no. of keywords to a maximum of five.

Q3

Please check the change made in the sentence “Although somewhat sluggish in response … portability of a smartphone.”

Q4

Please provide Acknowledgement section.

Cambridge University Press

1

Microsc. Microanal. 00, 1–11, 2015 doi:10.1017/S1431927615000173

2 © MICROSCOPY SOCIETY OF AMERICA 2015

5

A Review of Color Blindness for Microscopists: Guidelines and Tools for Accommodating and Coping with Color Vision Deficiency

6

Douglas R. Keene*

7

Shriners Hospital for Children, Micro-Imaging Center, 3101 SW Sam Jackson Park Road, Portland, OR 97239, USA

3 4

Q1

8 9 10 11 12 13 14 15 16 17 18

Q2

Abstract: “Color blindness” is a variable trait, including individuals with just slight color vision deficiency to those rare individuals with a complete lack of color perception. Approximately 75% of those with color impairment are green diminished; most of those remaining are red diminished. Red-Green color impairment is sex linked with the vast majority being male. The deficiency results in reds and greens being perceived as shades of yellow; therefore red-green images presented to the public will not illustrate regions of distinction to these individuals. Tools are available to authors wishing to accommodate those with color vision deficiency; most notable are components in FIJI (an extension of ImageJ) and Adobe Photoshop. Using these tools, hues of magenta may be substituted for red in red-green images resulting in striking definition for both the color sighted and color impaired. Web-based tools may be used (importantly) by color challenged individuals to convert red-green images archived in web-accessible journal articles into two-color images, which they may then discern.

20

Key words: accommodating color blindness, illustration, figure, simulate vision deficiency, Photoshop, Image J FIJI, journal archive, daltonize, protanopia, deuteranopia

21

I NTRODUCTION

19

22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46

With the prevalence of diverse colors in histological stains, fluorescent labels, and multicolored illustrations presented in the life sciences, consideration of the viewer’s ability to discriminate different colors is appropriate. “Color blindness” affects up to 8% of males and 0.4% of females. This article alerts the microscopy community to the prevalence of color vision deficiency within our audience and suggests the use of readily available tools to simulate how a multicolored image is perceived by affected individuals. With these tools the author may prepare data using a color scheme that allows striking definition to both color sighted and color challenged individuals. With the realization that many images within published archives are intangible to color deficient individuals, “real time” tools that convert an image to one perceptible to virtually all individuals are presented. The term “Color blindness” is unquestionably a misnomer. The term leads one to believe that color blind individuals perceive their surroundings only in grayscale, but in fact the disorder has many nuances and degrees. To understand these nuances, a basic understanding of the anatomy of the human retina is helpful. Among the photoreceptors of the retina, there are approximately seven million cone cells and 120 million rod cells (Fig. 1). Cones are found mainly in the central area of the retina (fovea), while rods are found in the peripheral retina. These photoreceptors are Received November 11, 2014; accepted January 27, 2015 *Corresponding author. [email protected]

composed of an inner segment and an outer segment, as well as a cell body and synaptic terminal. The photo pigment in rods is contained within flattened, internalized discs within the outer segment, whereas in cones the photo pigments are contained within membrane infoldings of a much shorter outer segment. Cone cells are responsible for color vision and function best in high levels of illumination. Rod cells function in low levels of illumination, triggered by just a few photons. Since cones are not stimulated by low light levels, night vision is primarily a function of rods and is mostly devoid of color; at higher light levels rods sense intensity, an important secondary role in color vision. Within photoreceptors, the photosensitive pigments are the opsins, a group of light-sensitive 35–55 kDa membranebound G protein-coupled receptors of the retinylidene protein family. Rod cells contain only rhodopsin. Cone cells contain various ratios of three opsins, distinguished by differences in their amino acid sequences, which result in differing light-absorption curves. Long-wave sensitive opsins (OPN1LW; MIM*300822) have a maximum absorbance at 561 nm, medium-wave sensitive opsins (OPN1MW; MIM*300821) at 531 nm and short-wave sensitive opsins (OPN1SW; MIM*613522) at 430 nm (Bowmaker & Dartnall, 1980; Baylor et al., 1987; McIntyre, 2002). These maximal sensitivities are very close to the primary colors red, green, and blue (RGB), respectively, and are therefore often referred to by those colors. Interestingly, red-green color deficiency is a sex-linked recessive trait, with the OPN1LW (red) and OPN1MW (green) genes mapped to the distal part of chromosome X. Women may carry a defective X chromosome

47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76

2

Douglas R. Keene

Figure 1. Scanning electron micrograph of human rod (gray) and cone cells (magenta) adjacent to the outer nuclear layer (orange). This colorized image is viewable to those with red, green, or blue color deficiencies [Image courtesy of Stephen Gschmeissner (www.theworldcloseup.com)].

Figure 2. Two colored images are often presented in red-green, with overlapping regions in yellow. The normal sighted would perceive the colors red, yellow, green, and blue as in column “N”. Column “P” simulates the same colors as perceived by a person with protanopia, column “D” deuteranopia, and column “T” tritanopia. Only Tritanopes would be able to distinguish all these color dots as separate, albeit in different hues.

108

109

110 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107

but do not show any sign of deficiency as the normal X chromosome acts dominantly. Although rare, color deficiency resulting from mutations in the gene for OPN1SW result in blue color insensitivities. The gene for OPN1SW is mapped to chromosome 7 and is therefore not a sex-linked trait (Nathans et al., 1986). OPN1SW, OPN1MW, and OPN1LW are all normally present in each of the three types of cones;which of the opsins predominates dictates the cones sensitivity to a different part of the color spectrum, defining the cone as red, green, or blue. An object whose color falls anywhere in the visible spectrum will therefore excite all three types of cones to a varying extent (Hurvich & Jameson, 1957; Shevell, 2003). Variation in perceived color among normal and color deficient individuals is most often caused by differences in amino acids involved in tuning the spectra of the red and green cone pigments. A non-null mutation in any one of the opsins may result in a slight shift of the maximal sensitivity of one of those three types of cones. This will result in less possible color mixtures and therefore a reduction in some portion of the perceived color spectrum (Rushton, 1966; Deeb, 2005). The disorder is referred to as anomalous trichromacy. Among the variations, protanomaly, deuteranomaly, and tritanomaly come from Greek and literally mean “deviation from the common” (anomaly) with the first (prot-), second (deuter-), or third (trit-) [cone], respectively (Fig. 2). Protanomaly results from a mutated form of the long-wavelength (red) pigment OPNILW, leading to less sensitivity to red light. The red portion of the spectrum is darkened, causing reds to reduce in intensity to the point where they can be mistaken for black. Although protanomaly

is a fairly rare form of color deficiency (1% of males and 0.1% of females), for the benefit of these individuals it is important to avoid dark red images in presentations; also red laser pointers will not be as visible green laser pointers. Deuteranomaly affects 6% of males and 0.4% of females (Nathans et al., 1986) and is the most common form of color deficiency. Deuteranomaly results from a mutated form of the medium-wavelength (green) pigment OPN1MW. The peak sensitivity is shifted from the green region toward the red region of the spectrum. Similar to protanomaly, deuteranomaly results in reduced discrimination within the red, orange, yellow, and green regions of the spectrum. In order to match a given hue of yellow light, deuteranomalous observers need more green in a red/green mixture than a normal observer. From a practical standpoint though, many protanomalous and deuteranomalous individuals have little difficulty carrying out everyday tasks that require normal color vision, and may not even be aware that their color discrimination differs from normal. As an aside, my friend (pictured with the green sail, to the right in the windsurfing picture, Fig. 3a) and I unfailingly disagree about the color of his sail. I insist that the color is green, and he insists that it is yellow. I showed him Figure 3, and to both our surprise he could not distinguish the “normal” from the “deuteranopia” image. As it turns out, the “color blind” tests discussed below suggest his vision deficiency as moderate deuteranomaly. There is a continuous gradation within “color blind” individuals; deuteranomaly can be everything between nearnormal color vision and deuteranopia, where OPN1MW is missing completely.

111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140

Composing and Viewing Colored Illustrations

3

Figure 3. Normal individuals perceive red-green images as in the left column (a, c), whereas deuteranopes (simulated in the right column, b, d) and also protanopes perceive the same images only in shades of yellow.

141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166

Tritanomaly is the rarest form of color deficiency, with an incidence of just 0.01% in both males and females (Nathans et al., 1986). Tritanomaly results from a mutated form of the short-wavelength (blue) pigment resulting in color perception that is shifted toward the green region of the spectrum. If one opsin is missing completely, there will be an “inability to see” (anopia) red (prot-), green (deuter-), or blue (trit-). These individuals have dichromatic vision, meaning that they must match any color they see with a mixture of just two primary colors. Protanopia and deuteroanopia each occur in about 1% of the male population (Nathans et al., 1986). For these forms of color deficiencies, both red and green are perceived as yellow and therefore cannot be distinguished (Fig. 2). Protanopes also perceive a reduced brightness that deuteranopes do not share, most pronounced in hues of red, which may be so severe that dimmer reds blend into black. They may confuse an active “red light” within a traffic signal as extinguished. These nuances are best appreciated by comparing a full spectrum color chart with one simulated for “color blindness” (Fig. 4). Monochromacy, resulting from a lack in all three opsins, is the only true color blindness as only shades of gray are seen. These individuals also have a severe light sensitivity, as there are only rods available to retrieve visual information.

There are currently no cures offered for human color blindness. However, studies in color blind monkeys that have provided the missing photo pigment by viral mediated gene transfer have successfully resulted in perception of the full color spectrum (Mancuso et al., 2009).

COLOR MODELING As a caveat, there is continual variability in the population resulting in different sensitivities of the color spectrum. Colors will also appear differently to the same individual in the context of lighting, shadows, luminosity, and intensities (Shevell & Kingdom, 2009). The perception of a paint-chip color chart is different under incandescent light as compared with natural sunlight; it also appears altered in bright light compared with dim; on a glossy surface compared with matte. Color modeling seeks to standardize, catalog, and reproduce color from one medium to another, but in the end there will always be variation in the individualistic perception of those colors.

COLOR CODES The color of an object depends on both the physics of the object in its environment and the characteristics of the

167 168 169 170 171

172 173 174 175 176 177 178 179 180 181 182 183 184

185 186 187

4

Douglas R. Keene

Figure 4. Reprinted with permission from visibone.com, the left panel demonstrates the full spectrum of visible colors as perceived by those with normal vision. The right panel simulates how a person with deuteranopia would perceive the same spectrum. These charts are enormously useful for designing illustrations and web pages accessible to color challenged individuals.

188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214

perceiving eye and brain. In its purest sense, color results from spectral illumination which is reflected, emitted, or transmitted through an object; the remaining portion of the spectrum is absorbed by that object. Absolute color refers to a color space in which colors are explicit; the documentation of color in this space is colorimetrically defined without reference to external factors (Knudson, 1999). However, the mathematical relationship that defines absolute color must also factor the average human perception of color. These data also need to be translated into a quantifiable form that can be understood by imaging software, monitors, printers, and ultimately those who will view the image. CIELab and sRGB are examples of absolute color spaces. The numeric values in CIELab color models describe all the colors that a person with normal vision sees, based on one channel for luminance (L; values range from 0 to 100) and two color channels (a, the green–red axis and b, the blue–yellow axis). For example, the values L = 72, a = 59, and b = −40 describe the magenta color circled in the micrograph in Figure 5. Because CIELab describes how a color looks rather than how much of a particular colorant is needed for reproduction (using a device such as a monitor, desktop printer, or digital camera), CIELab is considered to be a deviceindependent color model. Color management systems use “Lab” as a color reference to predictably transform a color from one color space to a different color space, such as the different codes used by a monitor and a printer. Lab images

can be saved in Photoshop, Photoshop PDF, and Tiff, as well as other formats.

RGB COLOR MODEL The RGB color mode most closely mimics color vision in the human eye (using RGB cones to define color) and is the default mode for displaying an image on monitors. Each of the color channels has an intensity level range between 0 and 255 in an 8-bit image; combinations of these channels result in over 16 million possible colors. This mode assigns an intensity value to each pixel. In 8-bit images, the intensity values range from 0 (black) to 255 (white) for each of the RGB components of a color image. When the values of all three components are equal, the result is a shade of neutral gray. When the values of all components are 255, the result is pure white; when the values are 0, pure black. For example, the magenta color circled in Figure 5 has an R value of 253, a G value of 131, and a B value of 253. As in the eye, the RGB color model is also additive, meaning that colors become more brilliant as more light is added. RGB is the most vibrant of the color models and is supported by nearly all file formats. An unhappy analogy to color perception is that the range of colors displayed on different monitors will depend on variations among those monitors. However, if the RGB colors of the monitor are calibrated exactly (together with

215 216

217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238


5

Figure 5. Colors are assigned different numeric values that are unique to each color mode. A color table (b) showing all the hues in an open image (a) may be accessed in Photoshop by first converting the image to Indexed color mode (Image/Mode/Indexed color); then: Image/Mode/Color table. A “left click” on any color in the color table followed by “left click” on any color in the image will display the numeric color values (arrow, c). The circled color in the micrograph (arrow, a) is identified in HSB mode with hue, saturation, and brightness values of 300°, 48 and 99%, respectively. HSB mode seeks to define color as we see it, for example “bright reddish-pink.” In RGB mode the color circled in the micrograph is identified as 253 parts red, 131 parts green, and 253 parts blue. In Lab mode the circled region has a luminosity value of 72; it is at position 59 on the red-green scale and at position −40 on the blue-yellow scale. The percentage of cyan, magenta, yellow, and black inks used to print the image would be 19, 54, 0, and 0, respectively.

239 240

241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258

other properties of the monitor) then RGB values on that monitor can be considered absolute.

TESTING FOR COLOR D EFICIENCIES The best known test for red-green color deficiency was developed by Dr. Shinobu Ishihara at Tokyo University in 1917. It consists of 38 pseudoisochromeric plates, each with a pattern of differently shaded dots. Within each plate is a number or shape composed of dots with contrasting hues (Ishihara, 1972). To those with normal color vision, these dots will be dissimilar enough from the background to form a figure or recognizable pattern on each plate, whereas those with a red-green color vision defect will not detect a figure on all plates. The success and failures of each plate are considered together, resulting in a diagnosis of normal to a particular degree of color insensitivity. Dr. Ishihara was concerned that if the plates were not exactly reproduced misdiagnosis would ensue, to the extent that he controlled the production of the plates for many years. Today there are many online versions that will give some indication of the type and degree of color vision deficiencies (see Colblindor at

www.color-blindness.com). The accuracy of these tests may be diminished since most computer monitors are not calibrated; however, a general diagnosis is often made after viewing just a few plates (Fig. 6). Deemed among the more accurate of the tests, the Farnnsworth-Munsell 100 Hue Color test contains four distinct rows of tiles arranged based on different color families. The left- and right-most tile in each is anchored, but all other tiles may be arranged so that the nearest neighbors are closest in hue. The test result is based on the number of instances that a hue cluster is misplaced, or the severity of a tile displacement. The test result seeks to specifically classify individual color deficiency in terms of deut-, prot-, and tritseverities (Supplementary Figure 1). Similarly, Anomaloscope color blind tests allow the user to continuously vary the colors and hues on one panel to match the color presented on a second fixed panel (Nagel, 1907; Cole & Vingrys, 1982).

259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 277

Supplementary Figure 1

279

Supplementary Figure 1 can be found online. Please visit journals.cambridge.org/jid_MAM.

278 280 281

6

Douglas R. Keene

Figure 6. Take the Ishihara Test! Color sighted individuals will perceive the numbers 7 and 2 embedded in the dot matrix (left). Protanopia (center) and deuteranopia (right) visions are simulated; these individuals will not perceive the numbers.

282 283

FIJI

284

It is likely that the keystrokes defined here will change with newer versions of FIJI. The version of FIJI used for these descriptions was updated in late December 2014. Algorithms have been developed for normal sighted individuals wishing to simulate color as perceived by color deficient individuals. Algorithms for color deficient individuals to simulate normal color vision have also been developed (Brettel et al., 1997). Simulating how an image is perceived by an individual with a color vision deficit is particularly useful in determining the accessibility of an image before presentation or publication. This can be easily facilitated using FIJI, which is an extension of ImageJ. FIJI is an image processing package having an array of “plugins” of interest to the life scientist. It is available as a free download (FIJI.sc), installs easily, and has an automatic update function. FIJI includes an algorithm that simulates the perception of a color image by individuals affected with protanopia, deuteroanopia, or tritanopia (FIJI/Image/Color/Simulate Color Blindness). Alternatively, a plugin may be downloaded (currently at Vischeck.com) and added to FIJI, allowing continuous access to the simulation (Plugins/Vischeck Panel). This plugin produced the deuteranopia simulation of the full spectrum image in Figure 4. The comparison is arresting and beautifully demonstrates the utility of the protocol. This plugin was also applied to the images of windsurfers and cultured cells (Fig. 3), demonstrating that a person with deuteranopia would not be able to distinguish

285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310

Figure 7. Normal vision distinguishes the color dots as red, yellow, green, and blue (column “N”). Those affected with protand deuter-anopias cannot distinguish red and green from yellow. If the entire column “N” is opened as a single image in FIJI; then Image/Color/Replace Red with Magenta/, the resultant is column (R → M). Columns P, D, and T simulate how persons with protanopia, deuteranopia, and tritanopia would perceive column (R → M). Note that all color dots may now be distinguished by color challenged individuals. Note also that yellows are perceived as white; overlap in red-green images would be perceived as shades of white.

the sail colors or differences in cell component localization in these red/green images. Developed by Masataka Okabe (Jikei Medical School) and Kei Ito (U. Tokyo) and implemented by Johannes Schindelin (University of Wisconsin-Madison), a one-button

311 312 313 314 315


7

Figure 8. The red portion of the red-green image in Figure 3c is converted in FIJI to magenta, allowing striking contrast for both the color sighted (a) and also Deuteroanopia individuals (simulated in b).

Figure 9. Replacement of one color with another selected for its specific hue may be accomplished in Photoshop (Image/ Adjustments/Replace Color). In this figure the red colored vesicles (arrow, a) were selected and a specific hue of magenta substituted for red (arrow, b). A “live” simulation of deuteranopia or protanopia, as the hues are varied, may be toggled on and off (see the text for keystrokes). Given its ease and versatility, this method for color substitution is favored.

316 317 318 319 320 321 322 323 324 325

solution (FIJI/Image/Color/Convert Red to Magenta) to altering a red/green image so that it is discernible by all is offered in FIJI. Magenta is substituted for red. The resulting image is striking for both the color sighted and color impaired (Figs. 7, 8). This is a fantastic tool for illustrators in the life sciences. When applied to a red/green confocal image, regions of overlap can be distinguished as shades of white (Supplementary Figure 2). An argument may be made to unilaterally substitute magenta for red in all two-color images. We may hope that manufacturers will include magenta as an option in

assigning color to multicolored fluorescent images and also apply it to systems producing multicolored graphs.

326 327 329


331


332 330

The conversion of red to magenta will not work for three color images showing co-localization (Supplementary Figure 3).

335 334

333

336

8

Douglas R. Keene

Figure 10. Color deficient individuals are often significantly challenged by multicolored graphs and charts. Red and green lines cannot be distinguished by many individuals. Care should be taken not to rely on a “color key;” instead arrows should designate the identity of each plot. Alternatively, Photoshop may be used, as in Figure 8, to choose and adjust hues within the plots. In this figure, the red dashed line (arrow, a) was replaced with a magenta dashed line (arrow, b) using the “Replace Color” tool in Photoshop (c). A simulation of deuteranopia and tritanopia demonstrates that the colors black, yellow, green, magenta, and dark blue in the lower graph may be distinguished; however, Protanopes will still have difficulty with this color combination. The use of dotted, dashed (etc.) lines is highly recommended.

337 338 339 340 341

Here, the only solution is to represent two channels at a time, or grayscale versions of the three separated channels, which often benefits all viewers as grayscale images will demonstrate minor grades in intensity to best advantage (Supplementary Figure 4).

343 345

Supplementary Figures 3 and 4

344 346

Supplementary Figures 3 and 4 can be found online. Please visit journals.cambridge.org/jid_MAM.

347

348 349 350 351 352 353 354 355 356 357

A DOBE PHOTOSHOP Keystrokes may change with upcoming versions of Adobe Photoshop; the keystrokes defined here are available using either CS5.1 or CS6. When working in Photoshop, the analytical tools available are dependent on the chosen color mode [Image/Mode (select Indexed, RGB, CMYK, Lab color, or Multichannel)] one is working with. While editing an image, RGB is a good choice since the color range is the broadest; however, if the

image is to be printed it should be checked in CMYK mode, which is the code printers use to reproduce images. Adobe Photoshop allows the same adjustments and simulations available in FIJI. To simulate either protanopia or deuteranopia, open the image and then select (View/Proof Setup/Color Blindness/select either Deuteranopia or Protanopia). Then, selecting (View/Proof Colors) will toggle the image between the original and the simulation for deuteranopia or protanopia (depending on which was selected). This is very convenient, as it allows a quick and seamless transition between the original and the simulation, which may be toggled as one manipulates the image. Adobe Photoshop does not share the one-button conversion of “red to magenta” offered in FIJI. However, replacing one color with another is quite simple. To do this, open the image in RGB mode and then select (Image/Adjustments/Replace Color). This will open the “Replace Color” screen. Select the box for “Localized Color Clusters,” select the left eyedropper tool, set “fuzziness” to 200, select “Selection,” then single click on the color to be changed (Fig. 9). Most often adjustment of the “hue” slider will result in a wanted color replacement. One may wish to adjust the

358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379


9

Figure 11. An image may be directed to Colorblinds.org either by opening the image directly in colorblinds.org or using an extension to Mozilla Firefox (see text). The image may be Daltonized or displayed as the original. The original image is shown in (b), with (c) simulating the deuteranopia vision of (b). a: Simulates that deuteranopes distinguish two colors, with regions of co-localization, in the Daltonized image. Interactively, the cursor (arrow, a) may be positioned over any part of the image and the color (yellow) and hue (golden poppy) at that position within the original image will be reported.

380 381 382 383 384 385 386 387 388 389 390 391 392

“fuzziness” slider with “selection” checked, which will show like-colored regions in the image where the color will be replaced. If deuteranopia or protanopia have already been selected via the View/Proof Setup keystrokes, the View/Proof Colors option will allow live simulation of color blindness as the Hue slider is adjusted. Depressing “Ctl” on the keyboard will open a small image visualization of the original, nonadjusted image. The control of image color is so thoroughly interactive and adjustable that this has become a favorite mechanism for color replacement. Similarly, multicolored graphs and tables can also be problematic for color challenged individuals. As shown in Figure 10, good choices for colored lines include black, dark

yellow, green, magenta ,and dark blue. It is very helpful to designate the identity of each line with a symbol and arrow, alas not with a color key. In addition, valuable is the inclusion of differing solid, dotted, and dashed lines, making color a useful (to the color sighted) but unnecessary (for the color deficient) element in discerning one plot from another.

WEB-BASED TOOLS An extension to the “Google Chrome” browser dubbed “Chrome Daltonize” is invaluable to both color vision normal and impaired individuals. Once Google Chrome is

393 394 395 396 397 398

399 400 401 402

10

Douglas R. Keene

Figure 12. Hardware tools for those with color blindness include “oxy-iso” and “enchroma” lenses (left panel). Applications (apps) are available for smartphones that can be tuned for an individual’s particular deficiency (i.e., DanKam, right panel). Both tools will likely allow a color deficient individual to pass the Ishihara test and significantly aid in redgreen definition.

403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426

installed, the Chrome Daltonize extension may be added. A small color wheel will appear next to the address window of the Google Chrome browser (Supplementary Figure 5). A right click on the color wheel will open an options menu. From this menu one may choose to Simulate or Daltonize— prot, deuter- ,or trit-anopia subsequently opened websites. Chrome Daltonize can be toggled on/off by selecting the customize button at the farthest right side of the address window (settings/extensions/Chrome Daltonize). Daltonization is a technique of modifying hues by enabling specific algorithms designed to compensate for each color deficiency. When enacted, a red-green image that would not be discernible to one with red-green deficiency is transformed to an image that is discernible. It works much like the Fiji plugin (Convert Red to Magenta) described above. This is an enormously useful tool for those with color deficiencies wishing to discriminate two-color archived journal images, which are most often red and green. These images are prevalent in our journals as evidenced by Allred et al. (2014), reporting that among the papers published in Nature from January to April 2014 with at least one figure requiring color discrimination, 75% used red-green images. Even if red/ green images in future publications were eliminated, these archived images will exist for eternity.

428 430


429 431


432

434 433 435 436 437 438 439 440 441 442

An extension may be added to the “Mozilla Firefox” browser so that a web image may be Daltonized with hues specifically modified for protanopia, deuteroanopia, or tritanopia. The extension adds the menu item “open image on colorblinds.org” to the right-click menu as the mouse is hovered over a web image. This redirects the image to colorblinds.org where it is daltonized. An added feature of this site is the ability for color deficient individuals to identify the hues and specific colors in an image as it is visualized by

normal sighted individuals. This can be very useful for the color challenged individual wishing to identify specific colored regions in a fiber when directed by text (Fig. 11). Currently the extension may be downloaded from “Colorblinds.org.” Although somewhat more cumbersome for web-based image simulation, “Chrometric” will simulate anopias and anomalies from images stored on your computer, making it a more versatile than FIJI or Photoshop for simulating a wider variety of color disorders. Images may be directly entered into the browser window, but since the browser does not include a search engine URLs must also be entered directly. Supplementary Figure 6 demonstrates the utility of the simulator. The free browser may be downloaded from “vaadin.com.”

443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 459


461


460 462

There are also hardware items that will aid the color impaired to better discriminate red/green images. Oxy-Iso lenses were developed for use by the medical community to identify veins and bruising that are otherwise undetectable to the naked eye. The lenses enhance red hues to the extent that people with red-green color deficiencies will likely pass the Ishihara test. However, the lenses are not a cure for color insensitivities as they filter out yellow and blues to the extent that a yellow light may appear extinguished. Another eyewear alternative are “EnChroma” lenses designed specifically for correction of either protanopia or deuteranopia. These lenses have ~100 coatings of dielectric material, each just a few nanometers thick, for filtering or reflecting specific portions of the spectrum between the primary colors to enhance reds and greens (Fig. 12). “DanKam” (Fig. 12) is an application available for smartphones that allows the user to display a scene through the camera and compensate for his or her color deficiency by

465 464

463

466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482

Composing and Viewing Colored Illustrations 483 484 485 486 487

Q3

488 489 490 491 492 493 494 495 496

Q4

497

498 499 500 501 502 503 504 505 506 507 508 539

using a continuously variable filter, allowing one to enhance various spectral colors. With this inexpensive app, a person with red-green color deficiency can see patterns in an Ishihara test (Fig. 7) that were previously not detectable. Although somewhat sluggish in response, the application allows “‘live’ image conversion” with the portability of a smartphone. In addition, there are some excellent websites that suggest guidelines for the preparation of figures and presentations accessible to color deficient individuals. It is enormously useful to consider the experience of color challenged individuals. Most notable among these sites is (M. Okabe and K. Ito. “How to make figures and presentations that are friendly to Color blind people.” Modified 2008: http://jfly.iam.u-tokyo.ac.jp/ color/index.html#checker (see also Wong, 2010, 2011).

REFERENCES ALLRED, S.C., SCHREINER, W.J. & SMITHIES, O. (2014). Color blindness: Still too many red-green figures. Nature 510, 340. BAYLOR, D.A., NUNN, B.J. & SCHNAPF, J.L. (1987). Spectral sensitivity of cones of the monkey Macaca fascicularis. J Physiol 390, 145–160. BOWMAKER, J.K. & DARTNALL, H.J. (1980). Visual pigments of rods and cones in human retina. J Physiol 298, 501–511. BRETTEL, H., VIENOT, F. & MOLLON, J.D. (1997). Computerized simulation of color appearance for dichromats. J Opt Soc Am A 14, 2647–2655.

11

COLE, B.L. & VINGRYS, A.J. (1982). A survey and evaluation of lantern tests of color vision. Am J Optom Physiol Opt 59, 346–374. DEEB, S.S. (2005). The molecular basis of variation in human color vision. Clin Genet 67, 369–377. HURVICH, L.M. & JAMESON, D. (1957). An opponent-process theory of color vision. Psychol Rev 64, 384–404. ISHIHARA, S. (1972). Tests for Colour-Blindness. Tokyo, Japan: Kanehara Shuppan Co., LTD. KNUDSEN, J.B. (1999). Java 2D Graphics. Sebastopol, CA: O’Reilly & Associates. MANCUSO, K., HAUSWIRTH, W.W., LI, Q., CONNOR, T.B., KUCHENBECKER, J.A., MAUCK, M.C., NEITZ, J. & NEITZ, M. (2009). Gene therapy for red-green colour blindness in adult primates. Nature 461, 784–787. MCINTYRE, D. (2002). Colour Blindness: Causes and Effects. Chester, UK: Dalton Publishing. NAGEL, W.A. (1907). Two cameras for Augenärzliche function test. Adaptometer and small spectrophotometer (Anomaloscope). J Ophthalmol 17, 201–222. NATHANS, J., THOMAS, D. & HOGNESS, D.S. (1986). Molecular genetics of human color vision: Genes encoding blue, green, and red pigments. Science 232, 193–202. RUSHTON, W.A. (1966). Densitometry of pigments in rods and cones of normal and color defective subjects. Invest Ophthalmol 5, 233–241. SHEVELL, S.K. (2003). The Science of Color. Oxford, UK: Elsevier. SHEVELL, S.K. & KINGDOM, A.A. (2009). Color in complex scenes. Annu Rev Psychol 59, 143–166. WONG, B. (2010). Points of view: Color coding. Nat Met 7, 573. WONG, B. (2011). Color blindness. Nat Met 8, 441.

509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538