- Email: [email protected]
Simple and inexpensive software designed for the evaluation of color
Simple and Inexpensive Software Designed for the Evaluation of Color
colored light so close together that they are seen as a single light source. The apparent color of that light source is determined by the relative intensities of its red, green, and blue (RGB) components.1 These components are known as the primary colors.2 The Windows operating system translates each combination of hue, saturation, and luminosity into varying degrees of energy directed at the spots of red, green, and blue phosphor.1 In RGB mode, various brightness values of red, green, and blue light combine to form the colors on the screen. The range of colors in the visible spectrum is represented by controlling the intensities of the individual RGB components. In Adobe Photoshop (Adobe Systems Incorporated, San Jose, California), the excellent standard software for handling pictures, an intensity value for each of the RGB components (ranging from 0 to 255) is assigned to each pixel. For example, a bright red color may have an R value of 246, a G value of 20, and a B value of 50. When the values of all three components are equal, the result is a shade of gray; when the value of each component is 255, the result is pure white; and when all components have values of 0, the result is pure black. These numbers are called color values.2 Hanson and associates digitized video-documented laryngoscopies and used Adobe Photoshop to analyze the level of redness of chronic laryngitis lesions.3 Adobe Photoshop measures average RGB color values of up to 5 ⫻ 5 pixels simultaneously2; so Hanson and associates made 10 random measurements in each of 5 areas.3 We thought this was a rational approach; so we developed software that was specially designed to facilitate the analysis of color value. During treatment with Xalatan (latanoprost; Pharmacia Corporation, Peapack, New Jersey) eyedrops, some patients experience darkening of the iris. We used such a patient as the subject in the evaluation of our software. The software compared the color of the irises before treatment and after 9 months of treatment. We used a Kowa photo slit SC-1200 (Kowa Optimed, Nagoya, Japan) to take the pictures and then made 35-mm slides. The 35-mm slides were scanned (600 dpi [dots/inch] is usually adequate) under the same conditions with the Microtek ScanMaker 35t plus (Microtek Laboratory Inc., Redondo Beach, California) and Adobe Photoshop in the Windows 98 operating system Japanese version. Then, we used our new software. The process is very simple. Users open the picture to be analyzed, click on the edge of the area where they would like to analyze the color value, and click the “clip” button. Then they click the “start” button (Figure 1). The picture color analyzer automatically calculates total and average RGB value. For example, total red (R) value means (red value of each pixel [1 to 255]) ⫻ (number of pixels in the area). Average (R) value is (total R value)/(255 ⫻ number of pixels in the area). The color
Isao Otaka, MD, Kenjiro Kumagai, MD, Yoko Inagaki, MD, Masaru Shimoyama, MD, Katsuyuki Saegusa, and Takeshi Hara, MD PURPOSE:
To report on new software that was specially designed to evaluate color. METHODS: Software development and observational case report. Each pixel on a computer screen is composed of three colors: red, green, and blue (RGB). Our software analyzes the intensity of each RGB component in a specific area chosen by the user. To test our software, we evaluated the color level of the irises of a subject, which became darker as a side effect of Xalatan (latanoprost; Pharmacia Corporation, Peapack, New Jersey) eyedrops. RESULT: We successfully expressed the level of the color of the iris by number. CONCLUSION: This software measures the color of a lesion and thereby provides an objective evaluation of color. The software we developed is downloadable, without cost, from http://www.isao.com. (Am J Ophthalmol 2002;133:140 –142. © 2002 by Elsevier Science Inc. All rights reserved.)
A
CASE REPORT OF SOFTWARE DEVELOPMENT AND OB-
servation. Although the accurate and objective evaluation of color is important in medicine, it is difficult to achieve. Recently, we developed software that is specially designed for this purpose. Here, we describe the simple software that enables us to evaluate color objectively and digitally. In the Windows operating system (Microsoft Corporation, Redmond, Washington), colors are recorded as a combination of three parameters: hue, saturation, and luminosity. The basic quality of a color, its redness, blueness, and so on, is defined by its hue. The purity of a color is defined by its saturation; a low saturation value means that more gray is mixed in. The brightness or dullness of a color is defined by its luminosity. Hue, saturation, and luminosity are the parameters that the Windows operating system uses internally, but videodisplay hardware uses a different set of numbers. Images on a color monitor are formed by a combination of dots, or pixels. To make each pixel visible, a beam of electrons is fired at three tiny spots of phosphor, one red, one green, and one blue. The result is three points of distinctly Accepted for publication Aug 7, 2001. From the Department of Ophthalmology, Shizuoka Red Cross Hospital, Shizuoka, Japan (I.O., Y.I., M.S.), the Department of Ophthalmology, Saiseikai Central Hospital, Tokyo, Japan (K.K.), Saegusa Precision Software, Fukui, Japan (K.S.), and the Department of Ophthalmology, Omiya Red Cross Hospital, Saitama, Japan (T.H.). Inquiries to Isao Otaka, MD, Department of Ophthalmology, Shizuoka Red Cross Hospital, 8-2 Outemachi, Shizuoka city, Shizuoka 420-0853, Japan; fax: 81-54-252-8816; e-mail: [email protected]
140
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
JANUARY 2002
FIGURE 1. Users open the picture to be analyzed, click on the edge of the area where they would like to analyze the color value (arrow), and then click the “clip” button and the “start” button (arrowhead).
FIGURE 2. The picture color analyzer automatically calculates total and average RGB value (arrow). The color that matches the average RGB value is also shown on the screen (arrowhead).
VOL. 133, NO. 1
BRIEF REPORTS
141
that matches the average RGB value is also shown on the screen (Figure 2). Before treatment, the iris shows an average value of R 0.92, G 0.77, and B 0.56. After 9 months of treatment with Xalatan, the iris shows an average value of R 0.65, G 0.59, and B 0.48. All average RGB values are less after treatment than before, which indicates that this is an objective measure of the darkening of the iris. This study has a limitation. It is very important to have exactly the same lighting conditions during the photographic sessions; so we used only the flashlight of the camera under exactly the same settings, but this does not mean “exactly the same” lighting conditions. Despite this limitation, the result of this study is encouraging. We believe this software is useful for the evaluation of color in the medical field. We call this software “Picture Color Analyzer,” and it is downloadable, without cost, from http://www.isao.com. REFERENCES
1. Stinson C. Running Microsoft Windows 98. Microsoft Press: Redmond, WA, 1998. 2. Narayanan S. Adobe Photoshop version 3.0 “help”. Swirsky R, Corboy D (creators). Adobe Systems Incorporated, 1995. 3. Hanson DG, Jiang J, Chi W. Quantitative color analysis of laryngeal erythema in chronic posterior laryngitis. J Voice 1998;12(1):78 – 83.
treated with 8-methoxypsoralen with ultraviolet irradiation, and Taq treated with Sau 3A1. RESULTS: Using untreated Taq, false-positive results were obtained in nested polymerase chain reaction with all 10 control samples, which were not seen with the other two methods of nested polymerase chain reaction. However, the senstivity of nested polymerase chain reaction using Sau 3A1 was the same sensitivity as the conventional culture (34.4%), whereas the sensitivity of the nested polymerase chain reaction using 8-methoxypsoralen was 46.9% higher than in the conventional culture. CONCLUSION: To eliminate the problem of false positives in bacterial nested polymerase chain reaction, we recommend the routine utilization of Taq treated with 8methoxypsoralen and ultraviolet irradiation. (Am J Ophthalmol 2002;133:142–144. © 2002 by Elsevier Science Inc. All rights reserved.)
M
Reliability of Nested Polymerase Chain Reaction in the Diagnosis of Bacterial Endophthalmitis Savitri Sharma, MD, Debashish Das, MPhil, Raj Anand, MS, Taraprasad Das, MD, and Chitra Kannabiran, PHD PURPOSE:
To test the utility of DNA Taq polymerase (Taq) treated with 8-methoxypsoralen with ultraviolet irradiation and Taq treated with restriction endonuclease in a nested polymerase chain reaction test for the diagnosis of bacterial endophthalmitis. METHODS: Prospective, comparative study. Vitreous biopsy fluid from 32 cases of clinically diagnosed bacterial endophthalmitis and 10 noninfective controls were cultured for aerobic/anaerobic bacteria and tested by nested polymerase chain reaction using two sets of universal eubacterial primers in triplicate with untreated Taq, Taq Accepted for publication Aug 6, 2001. From the Jhaveri Microbiology Centre (S.S., D.D.), Kanuri Santhamma Centre for Vitreo Retinal Diseases (R.A., T.D.), and Molecular Genetics (C.K.), Hyderabad Eye Research Foundation, L.V. Prasad Eye Institute, L.V. Prasad Marg, Banjara Hills, Hyderabad, India. Inquiries to Savitri Sharma, MD, Jhaveri Microbiology Centre, Hyderabad Eye Research Foundation, L.V. Prasad Eye Institute, L.V. Prasad Marg, Banjara Hills, Hyderabad 500 034, AP, India; fax: ⫹91 40 3548271; e-mail: [email protected]
142
AMERICAN JOURNAL
OLECULAR DIAGNOSTIC TECHNIQUES, BASED ON
conserved gene sequences of 16S rDNA, have shown great potential as extremely sensitive diagnostic tools in the diagnosis of a wide range of bacterial diseases, including bacterial endophthalmitis.1,2 Published techniques have used nested polymerase chain reaction using two sets of universal eubacterial primers, with or without primers for specific bacterial species,1 or DNA sequencing with broad-range primers.2 Compared with the latter technique, which requires expensive setup for DNA sequencing, nested polymerase chain reaction technique has the advantage of being within the reach of a reasonably equipped diagnostic microbiology laboratory. Over the past years, despite the relative ease of performance and high sensitivity of nested polymerase chain reaction, it did not achieve the expected popularity, probably owing to the inherent problem of unknown bacterial DNA contamination of several commercial Taq polymerase (Taq) preparations leading to false-positive results.3 The problem of false positivity, as a result of contaminated Taq, has been reported with primers for Legionella 5S ribosomal RNA and Escherichia coli ribosomal RNA gene sequences 1376 to 1395 and 1521 to 1540, in addition to universal eubacterial primers. We report the results of a prospective study, designed to compare two methods of elimination of DNA contamination in Taq, expecting that by eliminating the DNA contamination in Taq, a nested polymerase chain reaction with universal eubacterial primers can retain its utility in the rapid diagnosis of bacterial endophthalmitis. We tested vitreous biopsy fluid from 32 cases of clinically diagnosed bacterial endophthalmitis (posttraumatic-16, postoperative-7, endogenous-8, postinfectious keratitis-1) and 10 noninfective controls (vitreous hemorrhage-3, retinal detachment-7). All samples were cultured for aerobic/anaerobic bacteria and tested by nested polymerase chain
OF
OPHTHALMOLOGY
JANUARY 2002
colored light so close together that they are seen as a single light source. The apparent color of that light source is determined by the relative intensities of its red, green, and blue (RGB) components.1 These components are known as the primary colors.2 The Windows operating system translates each combination of hue, saturation, and luminosity into varying degrees of energy directed at the spots of red, green, and blue phosphor.1 In RGB mode, various brightness values of red, green, and blue light combine to form the colors on the screen. The range of colors in the visible spectrum is represented by controlling the intensities of the individual RGB components. In Adobe Photoshop (Adobe Systems Incorporated, San Jose, California), the excellent standard software for handling pictures, an intensity value for each of the RGB components (ranging from 0 to 255) is assigned to each pixel. For example, a bright red color may have an R value of 246, a G value of 20, and a B value of 50. When the values of all three components are equal, the result is a shade of gray; when the value of each component is 255, the result is pure white; and when all components have values of 0, the result is pure black. These numbers are called color values.2 Hanson and associates digitized video-documented laryngoscopies and used Adobe Photoshop to analyze the level of redness of chronic laryngitis lesions.3 Adobe Photoshop measures average RGB color values of up to 5 ⫻ 5 pixels simultaneously2; so Hanson and associates made 10 random measurements in each of 5 areas.3 We thought this was a rational approach; so we developed software that was specially designed to facilitate the analysis of color value. During treatment with Xalatan (latanoprost; Pharmacia Corporation, Peapack, New Jersey) eyedrops, some patients experience darkening of the iris. We used such a patient as the subject in the evaluation of our software. The software compared the color of the irises before treatment and after 9 months of treatment. We used a Kowa photo slit SC-1200 (Kowa Optimed, Nagoya, Japan) to take the pictures and then made 35-mm slides. The 35-mm slides were scanned (600 dpi [dots/inch] is usually adequate) under the same conditions with the Microtek ScanMaker 35t plus (Microtek Laboratory Inc., Redondo Beach, California) and Adobe Photoshop in the Windows 98 operating system Japanese version. Then, we used our new software. The process is very simple. Users open the picture to be analyzed, click on the edge of the area where they would like to analyze the color value, and click the “clip” button. Then they click the “start” button (Figure 1). The picture color analyzer automatically calculates total and average RGB value. For example, total red (R) value means (red value of each pixel [1 to 255]) ⫻ (number of pixels in the area). Average (R) value is (total R value)/(255 ⫻ number of pixels in the area). The color
Isao Otaka, MD, Kenjiro Kumagai, MD, Yoko Inagaki, MD, Masaru Shimoyama, MD, Katsuyuki Saegusa, and Takeshi Hara, MD PURPOSE:
To report on new software that was specially designed to evaluate color. METHODS: Software development and observational case report. Each pixel on a computer screen is composed of three colors: red, green, and blue (RGB). Our software analyzes the intensity of each RGB component in a specific area chosen by the user. To test our software, we evaluated the color level of the irises of a subject, which became darker as a side effect of Xalatan (latanoprost; Pharmacia Corporation, Peapack, New Jersey) eyedrops. RESULT: We successfully expressed the level of the color of the iris by number. CONCLUSION: This software measures the color of a lesion and thereby provides an objective evaluation of color. The software we developed is downloadable, without cost, from http://www.isao.com. (Am J Ophthalmol 2002;133:140 –142. © 2002 by Elsevier Science Inc. All rights reserved.)
A
CASE REPORT OF SOFTWARE DEVELOPMENT AND OB-
servation. Although the accurate and objective evaluation of color is important in medicine, it is difficult to achieve. Recently, we developed software that is specially designed for this purpose. Here, we describe the simple software that enables us to evaluate color objectively and digitally. In the Windows operating system (Microsoft Corporation, Redmond, Washington), colors are recorded as a combination of three parameters: hue, saturation, and luminosity. The basic quality of a color, its redness, blueness, and so on, is defined by its hue. The purity of a color is defined by its saturation; a low saturation value means that more gray is mixed in. The brightness or dullness of a color is defined by its luminosity. Hue, saturation, and luminosity are the parameters that the Windows operating system uses internally, but videodisplay hardware uses a different set of numbers. Images on a color monitor are formed by a combination of dots, or pixels. To make each pixel visible, a beam of electrons is fired at three tiny spots of phosphor, one red, one green, and one blue. The result is three points of distinctly Accepted for publication Aug 7, 2001. From the Department of Ophthalmology, Shizuoka Red Cross Hospital, Shizuoka, Japan (I.O., Y.I., M.S.), the Department of Ophthalmology, Saiseikai Central Hospital, Tokyo, Japan (K.K.), Saegusa Precision Software, Fukui, Japan (K.S.), and the Department of Ophthalmology, Omiya Red Cross Hospital, Saitama, Japan (T.H.). Inquiries to Isao Otaka, MD, Department of Ophthalmology, Shizuoka Red Cross Hospital, 8-2 Outemachi, Shizuoka city, Shizuoka 420-0853, Japan; fax: 81-54-252-8816; e-mail: [email protected]
140
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
JANUARY 2002
FIGURE 1. Users open the picture to be analyzed, click on the edge of the area where they would like to analyze the color value (arrow), and then click the “clip” button and the “start” button (arrowhead).
FIGURE 2. The picture color analyzer automatically calculates total and average RGB value (arrow). The color that matches the average RGB value is also shown on the screen (arrowhead).
VOL. 133, NO. 1
BRIEF REPORTS
141
that matches the average RGB value is also shown on the screen (Figure 2). Before treatment, the iris shows an average value of R 0.92, G 0.77, and B 0.56. After 9 months of treatment with Xalatan, the iris shows an average value of R 0.65, G 0.59, and B 0.48. All average RGB values are less after treatment than before, which indicates that this is an objective measure of the darkening of the iris. This study has a limitation. It is very important to have exactly the same lighting conditions during the photographic sessions; so we used only the flashlight of the camera under exactly the same settings, but this does not mean “exactly the same” lighting conditions. Despite this limitation, the result of this study is encouraging. We believe this software is useful for the evaluation of color in the medical field. We call this software “Picture Color Analyzer,” and it is downloadable, without cost, from http://www.isao.com. REFERENCES
1. Stinson C. Running Microsoft Windows 98. Microsoft Press: Redmond, WA, 1998. 2. Narayanan S. Adobe Photoshop version 3.0 “help”. Swirsky R, Corboy D (creators). Adobe Systems Incorporated, 1995. 3. Hanson DG, Jiang J, Chi W. Quantitative color analysis of laryngeal erythema in chronic posterior laryngitis. J Voice 1998;12(1):78 – 83.
treated with 8-methoxypsoralen with ultraviolet irradiation, and Taq treated with Sau 3A1. RESULTS: Using untreated Taq, false-positive results were obtained in nested polymerase chain reaction with all 10 control samples, which were not seen with the other two methods of nested polymerase chain reaction. However, the senstivity of nested polymerase chain reaction using Sau 3A1 was the same sensitivity as the conventional culture (34.4%), whereas the sensitivity of the nested polymerase chain reaction using 8-methoxypsoralen was 46.9% higher than in the conventional culture. CONCLUSION: To eliminate the problem of false positives in bacterial nested polymerase chain reaction, we recommend the routine utilization of Taq treated with 8methoxypsoralen and ultraviolet irradiation. (Am J Ophthalmol 2002;133:142–144. © 2002 by Elsevier Science Inc. All rights reserved.)
M
Reliability of Nested Polymerase Chain Reaction in the Diagnosis of Bacterial Endophthalmitis Savitri Sharma, MD, Debashish Das, MPhil, Raj Anand, MS, Taraprasad Das, MD, and Chitra Kannabiran, PHD PURPOSE:
To test the utility of DNA Taq polymerase (Taq) treated with 8-methoxypsoralen with ultraviolet irradiation and Taq treated with restriction endonuclease in a nested polymerase chain reaction test for the diagnosis of bacterial endophthalmitis. METHODS: Prospective, comparative study. Vitreous biopsy fluid from 32 cases of clinically diagnosed bacterial endophthalmitis and 10 noninfective controls were cultured for aerobic/anaerobic bacteria and tested by nested polymerase chain reaction using two sets of universal eubacterial primers in triplicate with untreated Taq, Taq Accepted for publication Aug 6, 2001. From the Jhaveri Microbiology Centre (S.S., D.D.), Kanuri Santhamma Centre for Vitreo Retinal Diseases (R.A., T.D.), and Molecular Genetics (C.K.), Hyderabad Eye Research Foundation, L.V. Prasad Eye Institute, L.V. Prasad Marg, Banjara Hills, Hyderabad, India. Inquiries to Savitri Sharma, MD, Jhaveri Microbiology Centre, Hyderabad Eye Research Foundation, L.V. Prasad Eye Institute, L.V. Prasad Marg, Banjara Hills, Hyderabad 500 034, AP, India; fax: ⫹91 40 3548271; e-mail: [email protected]
142
AMERICAN JOURNAL
OLECULAR DIAGNOSTIC TECHNIQUES, BASED ON
conserved gene sequences of 16S rDNA, have shown great potential as extremely sensitive diagnostic tools in the diagnosis of a wide range of bacterial diseases, including bacterial endophthalmitis.1,2 Published techniques have used nested polymerase chain reaction using two sets of universal eubacterial primers, with or without primers for specific bacterial species,1 or DNA sequencing with broad-range primers.2 Compared with the latter technique, which requires expensive setup for DNA sequencing, nested polymerase chain reaction technique has the advantage of being within the reach of a reasonably equipped diagnostic microbiology laboratory. Over the past years, despite the relative ease of performance and high sensitivity of nested polymerase chain reaction, it did not achieve the expected popularity, probably owing to the inherent problem of unknown bacterial DNA contamination of several commercial Taq polymerase (Taq) preparations leading to false-positive results.3 The problem of false positivity, as a result of contaminated Taq, has been reported with primers for Legionella 5S ribosomal RNA and Escherichia coli ribosomal RNA gene sequences 1376 to 1395 and 1521 to 1540, in addition to universal eubacterial primers. We report the results of a prospective study, designed to compare two methods of elimination of DNA contamination in Taq, expecting that by eliminating the DNA contamination in Taq, a nested polymerase chain reaction with universal eubacterial primers can retain its utility in the rapid diagnosis of bacterial endophthalmitis. We tested vitreous biopsy fluid from 32 cases of clinically diagnosed bacterial endophthalmitis (posttraumatic-16, postoperative-7, endogenous-8, postinfectious keratitis-1) and 10 noninfective controls (vitreous hemorrhage-3, retinal detachment-7). All samples were cultured for aerobic/anaerobic bacteria and tested by nested polymerase chain
OF
OPHTHALMOLOGY
JANUARY 2002