Medical display application for degraded image sharpness restoration based on the modulation transfer function: Initial assessment for a five-megapixel mammography display monitor

1.

Kagadis GC, Walz-Flannigan A, Krupinski EA, Nagy PG, Katsanos K, Diamantopoulos A, Langer SG (2019)

Medical

imaging

displays

and

their

use

in

image

interpretation.

Radiographics

33(1):275-291.

https://doi.org/10.1148/rg.331125096

2.

Samei E, Badano A, Chakraborty D, Compton K, Cornelius C, Corrigan K, Flynn MJ, Hemminger B, Hangiandreou

N, Johnson J, Moxley M, Pavlicek P, Roehrig H, Rutz L, Shepard J, Uzenoff R, Wang J, Willis C (2014) Assessment

of display performance for medical imaging systems: Executive summary of AAPM TG18 report. Med Phys

32(4):1205-1225. https://doi.org/10.1118/1.1861159

3.

Fan J, Roehrig H, Sundareshan MK, Krupinski EA (2006) Noise estimation and reduction on five medical liquidcrystal displays. J Soc Inf Disp 14(10):861-866. https://doi.org/10.1889/1.2372419

4.

Ichikawa K, Kodera Y, Nishimura A, Hasegawa M, Kimura N, Takemura A, Matsubara K (2008) Analysis method of

noise power spectrum for medical monochrome liquid crystal displays. Radiol Phys Technol 1(2):201-207.

https://doi.org/10.1007/s12194-008-0029-y

5.

Ichikawa K, Hasegawa M, Kimura N, Kawashima H, Kodera Y (2008) A new resolution enhancement technology

using the independent sub-pixel driving for the medical liquid crystal displays. IEEE J Display Technol 4(4):377-382.

https://doi.org/10.1109/jdt.2008.922414

6.

Tokurei S, Morishita J (2015) A method for evaluating image quality of monochrome and color displays based on

luminance by use of a commercially available color digital camera. Med Phys 42(8):4773–4782.

https://doi.org/10.1118/1.4926850

7.

Giger ML, Doi K (1984) Investigation of basic imaging properties in digital radiography. I. Modulation transfer

function. Med Phys 11(3):287-295. https://doi.org/10.1118/1.595629

8.

Sivaramakrishna R, Obuchowski NA, Chilcote WA, Cardenosa G, Powell KA (2000) Comparing the performance

of mammographic enhancement algorithms. AJR Am J Roentgenol 175(1):45-51.

https://doi.org/10.2214/ajr.175.1.1750045

9.

Pisano ED, Cole EB, Hemminger BM, Yaffe MJ, Aylward SR, Maidment AD, Johnston RE, Williams MB,

Niklason LT, Conant EF, Fajardo LL, Kopans DB, Brown ME, Pizer SM (2000) Image processing algorithms for

digital mammography: a pictorial essay. Radiographics 20(5):1479-1491.

https://doi.org/10.1148/radiographics.20.5.g00se311479

10. Reiker GG, Blume HR, Slone RM, Woodard PK, Gierada DS, Sagel SS, Jost RG, Blaine GJ (1997) Filmless digital

chest radiography within the radiology department. Proc SPIE 3035:355-368. https://doi.org/10.1117/12.274591

11. Roehrig H, Dallas WJ, Krupinski E, Fan J, Chawla A, Gandhi K (2003) Display of mammograms on a CRT. In:

Peitgen HO (ed) Digital mammography, Springer, Berlin, 452-454. https://doi.org/10.1007/978-3-642-59327-7_107

12. Krupinski EA, Roehrig H, Engstrom M, Johnson J, Lubin J (2003) MTF correction for optimizing softcopy display

of digital mammograms: Use of a vision model for predicting observer performance. Proc SPIE 5034:323-327.

https://doi.org/10.1117/12.479980

13. Krupinski EA, Johnson J, Roehrig H, Engstrom M, Fan J, Nafziger J, Lubin J, Dallas WJ (2003) Using a human

visual system model to optimize soft-copy mammography display: influence of MTF compensation. Acad Radiol

10(9):1030-1035. https://doi.org/10.1016/s1076-6332(03)00293-9

14. EIZO Corporation. Study to demonstrate efficiency of sharpness recovery. White Paper.

https://www.eizo.co.jp/support/db/files/technical_information/2016/Q16B014-AS-07001.pdf. Accessed 7 Aug 2020.

15. Miller TR, Sampathkumaran KS (1982) Design and application of finite impulse response digital filters. Eur J

Nucl Med 7(1):22-27. https://doi.org/10.1007/bf00275240

16. Ichikawa K, Kodera Y, Fujita H (2006) MTF measurement method for medical displays by using a bar-pattern

image. J Soc Inf Disp 14(10):831-837. https://doi.org/10.1889/1.2372416

17. Ichikawa K, Nishi Y, Hayashi S, Hasegawa M, Kodera Y (2008) Noise reduction effect in super-high resolution

LCDs using independent sub-pixel driving technology. Proc SPIE 6917: 69171F. https://doi.org/10.1117/12.770848

18. Shiraishi J, Abe H, Ichikawa K, Schmidt RA, Doi K (2010) Observer study for evaluating potential utility of a

super-high-resolution LCD in the detection of clustered microcalcifications on digital mammograms. J Digit

Imaging 23(2):161-169. https://doi.org/10.1007/s10278-009-9192-x

19. Ihori A, Fujita N, Sugiura A, Yasuda N, Kodera Y (2013) Phantom-based comparison of conventional versus phasecontrast mammography for LCD soft-copy diagnosis. Int J Comput Assist Radiol Surg 8(4):621-633. http://

doi.org/10.1007/s11548-012-0805-3

20. Muramatsu C, Li Q, Schmidt RA, Shiraishi J, Suzuki K, Newstead GM, Doi K (2007) Determination of subjective

similarity for pairs of masses and pairs of clustered microcalcifications on mammograms: comparison of similarity

ranking scores and absolute similarity ratings. Med Phys 34(7):2890-2895. http:// doi.org/10.1118/1.2745937

Table Caption

Table 1

Inverse filter coefficients with a restricted size of 5 × 5, calculated from the inverse of the measured MTFs in

the horizontal (non-subpixel) and vertical (subpixel) directions of the 5-MP mammography LCD monitor

Figure Captions

Fig. 1

Pixel structure of two 5-MP monochrome LCD panels (pixel size: 0.165 mm) with different pixel aperture

ratios: (a) A conventional panel with a pixel aperture ratio of 33.0% and (b) a higher-luminance panel with a pixel

aperture ratio of 41.7% used in this study. Each pixel in a monochrome LCD monitor is composed of three subpixel

elements (separated by dashed lines). Note that the “black matrix area” means the inactive area created in the

manufacturing process to prevent light leakage from the backlight of the LCD panels.

Fig. 2

(a) MTFs and (b) NPS in the horizontal (non-subpixel) and vertical (subpixel) directions of the 5-MP

monochrome LCD monitor

Fig. 3

Two types of frequency responses of the finite impulse response (FIR) inverse filters in the horizontal (H) and

vertical (V) directions of the LCD monitor; solid lines indicate the desired responses calculated from the inverse of the

monitor MTFs and the dashed lines indicate the ones implemented with 5 × 5 filter coefficients

Fig. 4

(a) A partial image of the captured horizontal bar-pattern image including the five bars with 1-, 2-, 3-, 4-, and

6-pixel widths displayed on the 5-MP monochrome LCD monitor. Square-wave profiles of the five bars with three-pixel

width corresponding to approximately 1.0 cycle/mm in the (b) horizontal and (c) vertical directions, obtained from the

area bounded by the dashed line of the SR-processed and unprocessed bar-pattern images.

Fig. 5

MTFs in the (a) horizontal and (b) vertical directions measured from the square-wave profiles of the SR-

processed and unprocessed bar-pattern images displayed on the 5-MP monochrome LCD monitor

Fig. 6

The NPS in the (a) horizontal and (b) vertical directions measured from the X-ray exposed image of the ACR

accreditation phantom (for overall NPS) and uniform pattern image (for inherent NPS), displayed on the 5-MP

monochrome LCD monitor, respectively. The overall NPS value of the SR-processed X-ray phantom image was slightly

greater than that of the unprocessed one, especially at frequencies higher than about 1 cycle/mm.

Acknowledgments

The authors are grateful to Kenshi Shiotsuki, RT for useful discussions, Takafumi Nomura, MD, Keisuke Miyoshi,

MD, Shoko Ariyoshi, MD, and Masaki Kamiya, MD (Yamaguchi University) for their participation as observers, Sono

Kanoya, RT, Fumiko Yurino, RT, Ayumi Hashimoto, RT, and Eri Tokurei, RT for their assistance in the preparation of

a database containing clinical digital mammograms, Mamoru Ogaki, Yusuke Bamba, Masaki Kita, and Noriyuki

Hashimoto (EIZO Corporation) for the technical support on display monitors with image processing systems, and

Kazuyuki Watanabe, Kazushige Hatori, and Hideaki Mizobe (Canon Corporation) for providing a single-lens reflex

digital color camera.

Declarations

Funding This work was supported in part by grant from EIZO Corporation (Ishikawa, Japan).

Conflicts of interest/Competing interests J. Morishita received a research grant from EIZO Corporation (Ishikawa,

Japan).

Ethics approval All procedures in studies involving human participants were performed in accordance with the ethical

standards of the Institutional Review Board and with the 1964 Helsinki Declaration and its later amendments, or

comparable ethical standards.

Consent to participate Informed consent was waived for the all images used in this study by the Institutional Review

Board.

Consent for publication Not applicable

Availability of data and material Not applicable

Code availability Not applicable

Fig. 1

Horizontal (non-subpixel) direction

Light transmission area

0.165 mm

Vertical (subpixel) direction

Light transmission area

Black matrix area

(a)

Black matrix area

(b)

Fig. 2

1.0×10−1

(a)

Noise power spectral value (mm 2)

Modulation transfer factor

1.0

0.8

0.6

Horizontal direction

0.4

Vertical direction

0.2

(b)

1.0×10−2

1.0×10−3

1.0×10−4

Horizontal direction

1.0×10−5

1.0×10−6

Vertical direction

1.0×10−7

0.0

0.0

1.0

2.0

Spatial frequency (cycles/mm)

3.0

0.0

1.0

2.0

Spatial frequency (cycles/mm)

3.0

2.5

(a) Horizontal direction

3×3 filter

2.0

7×7 filter

1.5

1.0

5×5 filter

Inverse of MTF

0.5

0.0

0.0

1.0

2.0

Spatial frequency (cycles/mm)

3.0

Magnitude of frequency response of inverse filters

Magnitude of frequency response of inverse filters

Fig. 3

2.5

(b) Vertical direction

3×3 filter

2.0

7×7 filter

1.5

1.0

5×5 filter

Inverse of MTF

0.5

0.0

0.0

1.0

2.0

Spatial frequency (cycles/mm)

3.0

Fig. 4

(a) Bar-pattern image

1 pixel

3 pixel

4 pixel

6 pixel

Without sharpness recovery function

With sharpness recovery function

(b) Horizontal direction

80

Luminance (cd/m 2)

2 pixel

60

40

20

1 pixel

0.0

1.0

2.0

3.0

4.0

Distance (mm)

Luminance (cd/m2)

Without sharpness recovery function

With sharpness recovery function

(c) Vertical direction

60

40

1 pixel

20

0.0

1.0

2.0

3.0

Distance (mm)

4.0

Fig. 5

1.0

1.0

(b)

0.8

Modulation transfer factor

Modulation transfer factor

(a)

With sharpness recovery function

0.6

0.4

Horizontal

direction

Without sharpness recovery function

(Inherent MTF from bar-pattern image)

0.2

0.0

0.8

With sharpness recovery function

0.6

Without sharpness recovery function

(Inherent MTF from bar-pattern image)

0.4

Vertical

direction

0.2

0.0

0.0

1.0

2.0

Spatial frequency (cycles/mm)

3.0

0.0

1.0

2.0

Spatial frequency (cycles/mm)

3.0

Fig. 6

1.0×10−1

(a) Horizontal direction

1.0×10−2

(b) Vertical direction

Noise power spectral value (mm 2)

Noise power spectral value (mm 2)

1.0×10−1

X-ray phantom image with SR function

1.0×10−3

1.0×10−4

X-ray phantom image without SR function

1.0×10−5

1.0×10−6

1.0×10−2

X-ray phantom image with SR function

1.0×10−3

1.0×10−4

X-ray phantom image without SR function

1.0×10

−5

1.0×10−6

Inherent NPS (uniform pattern image)

Inherent NPS (uniform pattern image)

1.0×10−7

1.0×10

0.0

1.0

2.0

Spatial frequency (cycles/mm)

3.0

−7

0.0

1.0

2.0

Spatial frequency (cycles/mm)

3.0

...