Technical Report on Semiautomatic Segmentation Using the Adobe Photoshop

Park, Jin Seo; Chung, Min Suk; Hwang, Sung Ho; Lee, Yong Sook; Har, Dong-Hwan

doi:10.1007/s10278-005-6704-1

Cited by 87 publications

(77 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The cryomacrotome used (width, 3 m x height, 4 m x length 5 m; 15 ton; HanwonTM of Republic of Korea) with a moving error of 0.001 mm enabled the whole body to be serially ground at a constant thickness, as previously described (Park et al, 2005b). Using the cryomacrotome, the frozen embedding box containing the cadaver was ground at 0.2 mm (from the vertex of the head and perineum; upper body) or at 1.0 mm intervals (from under the perineum to the toes; lower body) to produce sectioned surfaces.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Improved Sectioned Images and Surface Models of the Whole Female Body

2015

Self Cite

View full text Add to dashboard Cite

SUMMARY:The objective of this research was to present high-quality sectioned images of a whole female body which would be helpful in creating an atlas, virtual dissection, and various applications for medical education and clinical trial. In addition, the authors sought to demonstrate the applicabilities of sectioned images. A female cadaver was ground serially using the cryomacrotome and photographed to make the sectioned images. Structures in the images were segmented to produce segmented images in Photoshop. In the self-developed browsing software, the sectioned and segmented images were stored. Based on the segmented images, surface models were built on commercial software and saved as PDF file. High-quality sectioned images of the female body were taken (intervals, 0.2 mm or 1 mm; pixel size, 0.1 mm; color depth, 48 bit color). In the images obtained, very small and complicated structures could be identified in color of living body. In order to ascertain the applicability of the images, the browsing software including sectioned and segmented images and the PDF file including surface models were produced. The sectioned images and surface models produced during this research will prove to be a useful source for medical software. All data generated is available free of charge.

show abstract

Section: Methodsmentioning

confidence: 99%

“…After selecting sectioned images at 1 mm intervals, the boundaries of selected structures were outlined using the lasso or magnetic lasso tools in Photoshop, and the outlined structures were then filled with specific colors (Park et al, 2005b;Park et al, 2010). Structure name, Fig.…”

Section: Park H S; Choi D H and Park J Smentioning

confidence: 99%

Improved Sectioned Images and Surface Models of the Whole Female Body

2015

Self Cite

View full text Add to dashboard Cite

show abstract

“…The outlines of anatomic structures can be quickly and accurately drawn (Park et al, 2005a). Also, the bitmap property of the segmented images can readily be converted into the vector property (Fig.…”

Section: Discussionmentioning

confidence: 99%

“…., 976.psd; intervals, 1.0 mm; resolution, 1,200 3 1,100; pixel size, 0.2 mm) were prepared ( Fig. 1) (Table 1) (Park et al, 2005a).…”

Section: Anatomic Structures Were Segmented Using Photoshopmentioning

confidence: 99%

Technique of semiautomatic surface reconstruction of the visible Korean human data using commercial software

et al. 2007

Self Cite

View full text Add to dashboard Cite

This article describes the technique of semiautomatic surface reconstruction of anatomic structures using widely available commercial software. This technique would enable researchers to promptly and objectively perform surface reconstruction, creating three-dimensional anatomic images without any assistance from computer engineers. To develop the technique, we used data from the Visible Korean Human project, which produced digitalized photographic serial images of an entire cadaver. We selected 114 anatomic structures (skin [1], bones [32], knee joint structures [7], muscles [60], arteries [7], and nerves [7]) from the 976 anatomic images which were generated from the left lower limb of the cadaver. Using Adobe Photoshop, the selected anatomic structures in each serial image were outlined, creating a segmented image. The Photoshop files were then converted into Adobe Illustrator files to prepare isolated segmented images, so that the contours of the structure could be viewed independent of the surrounding anatomy. Using Alias Maya, these isolated segmented images were then stacked to construct a contour image. Gaps between the contour lines were filled with surfaces, and three-dimensional surface reconstruction could be visualized with Rhinoceros. Surface imperfections were then corrected to complete the three-dimensional images in Alias Maya. We believe that the three-dimensional anatomic images created by these methods will have widespread application in both medical education and research. Clin. Anat. 20:871-879, 2007. V V C 2007 Wiley-Liss, Inc.

show abstract

“…This resulted in the enhanced contrast between target structures and the background. The lumbosacral structures were then outlined automatically, semi-automatically, or manually in sequence (Park et al, 2005a;Shin et al, 2009b). Table 2 Sequential procedures of outlining and surface reconstruction.…”

Section: Systemsmentioning

confidence: 99%