- Email: [email protected]
Influence of tissue preparation on the high-frequency acoustic properties of normal kidney tissue
Copyright
Ultrasound in Med. & Biol., Vol. 22, No. 9, pp. 1261- 1265, 1996 0 1996 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/96 $15.00 + 00
PII: SO301-5629( 96)00150-O
ELSEVIER
@Original Contribution INFLUENCE OF TISSUE PREPARATION ON THE HIGH-FREQUENCY ACOUSTIC PROPERTIES OF NORMAL KIDNEY TISSUE HIDEHIKO
SASAKI, YOSHIFUMI SAIJO, MONTONAO TANAKA, HIROAKI OKAWAI, YOSHIO TERASAWA, TOMOYUKI YAMBE and SHIN-ICHI NITTA Department of Medical Engineering and Cardiology, Division of Organ Pathophysiology, Institute of Development, Aging and Cancer, Tohoku University, Aoba-ku, Sendai, Japan (Received 15 January 1996; in final form 1 July 1996)
Abstract-The influence of various tissuepreparations on the acoustic properties of normal kidney tissue at high frequencieswasinvestigated. Eight surgically excisednormal kidney tissuespecimenswere classified into three groups: (i) fresh, frozen section, (ii) formalin&@ frozen section and (iii) formalin-lixed, paraffin section.Scanning acousticmicroscopy operating in the frequency range of 100-200 MHz wasused to display the two-dimensionaldistribution of attenuation constant and sound speed.Our results indicate that (i) there is no sign&ant variation in both acoustic parameters between the three tissue groups, (ii) fixation by 10% formalin producesno significant changein the acoustic parameters, (iii) in fat-free tissue regions, the acoustic parameters are independent of preparation method and (iv) frozen sectionsmust be usedto assess the acoustic parametersin fat-rich tissues. Copyright 0 1996 World Federation for Ultrasound
in Medicine & Biology. Kev Words: Ultrasonic tissue characterization. Acoustic microscopy, Attenuation constant, Sound speed, Tissuepreparation.
duced large changes in the acoustic properties of liver tissue. These studies were not, however, undertaken at the frequencies commonly used for SAM studies. The purpose of this study was to evaluate the influence of preparation techniques on the acoustic properties of normal kidney tissue over the lOO- to 200-MHz range, using a SAM system.
INTRODUCTION Scanning acoustic microscopy (SAM) provides data on physical properties that cannot be obtained by optical microscopy. Very often, formalin-fixed, paraffinembedded specimens are used for SAM investigations because the specimens can be accurately cut into thin sections with smooth surfaces and remain stable during the measurement procedure. However, the manipulations involved in tissue preparation, e.g., fixation, freezing, dehydration, defatting, paraffin embedding, deparaffinization and other steps, may alter the acoustic properties of the tissues. Because fat is eluted during the preparation of paraffin-embedded specimens, frozen sections must be used when measuring the acoustic properties of fat-rich tissues. The influence of formalin fixation on the acoustic properties of the tissue in the low-frequency range (2.55 MHz) has been studied in detail (Bamber et al. 1979; Hachiya et al. 1993). The investigations of Steen et al. ( 1991), which also included looking at the effect of embedding tissues in paraffin and deparaffinizing, in-
METHODS SAM system Figure 1 is a block diagram of the SAM system (Okawai et al. 1988; Saijo et al. 1991). The ultrasonic frequency is variable over the range of 100-200 MHz. The transducer is equipped with a lens having dimensions of diameter, radius of curvature and aperture angle of 1.25 mm, 1.25 mm and 60”, respectively, and having a -3-dB beam width in water (20°C) between 5 pm (at 200 MHz) and 10 pm (at 100 MHz). The transducer is oscillated in the x axis, and the sample holder is scanned in the y axis. The mechanical scanner is arranged so that the ultrasonic beam is transmitted at 4-ym intervals over a 2mm scanned width. The number of sampling points is 480 in one scanning line; 480 X 480 points make one frame within 4 s.
Address correspondence to: Hidehiko Sasaki, M.D., Department of Medical Engineering and Cardiology, Institute of Development, Aging and Cancer, Tohoku University, 4-1 Seiryo-machi, Aoba-ku, Sendai, 980-77 Japan. 1261
1262
Letter to the Editor-in-Chief
five frames of phase images in lo-MHz steps in the range 100-140 MHz. The thickness of the specimen is determined from the frequency-dependent characteristics of the amplitude and the phase of the received signals obtained at the same position for each of the 16 frames (Okawai et al. 1988). The attenuation constant can be calculated by normalizing the attenuation at twice the sample thickness.
JLTRASONICTRANSOUCERS
SIGNAL
PROCESSOR
A = Ll2fd
(3)
where A is the attenuation constant, fis the frequency and d is the thickness of the specimen. The sound speed (C) is calculated by the following equation: Fig. 1. Block diagram of the SAM
system.
C = (l/C,,, - 8/2rfd)-’
Figure 2 is an illustration of the relationship between the ultrasonic beam components and the tissue sample. The detected wave amplitude and phase deviation are described by the following equations: L = -20 log(y,/y,)
(1)
8 = arg(YllY3)
(2)
where L is the amplitude, 0 is the phase deviation and y1 and y3 are the ultrasonic beam components. In the analogue signal processor, the detector of both amplitude and phase is fixed, and an A/D converter is used, in which the error of quantity is approximately 2 m/s for a lo-pm thickness at an ultrasonic frequency of 130 MHz. The original signals for imaging produced in the analogue signal processor can be displayed on the display unit directly. However, these original signals do not have accurate values of attenuation constant and sound speed. The image processor stores the original image signals in 16 frames: 11 frames of amplitude images in lo-MHz steps in the range 100-200 MHz, and
(4)
where C is the sound speed in the specimen, C, is the sound speed in the coupling medium and 8 is the phase deviation. A linear frequency dependence of the attenuation, in the ultrasonic frequency range of 100-200 MHz, was obtained from consideration of the interference of sound wave components (Okawai et al. 1988). The values of the attenuation constant and sound speed thus obtained were converted into colour signals, which corresponded to the values of the amplitude and phase, and were displayed on a colour monitor as a twodimensional distribution. Table 1 shows the relationship between colour-coded scales and values of attenuation constant and sound speed. Distilled water was used for the coupling medium, which was maintained between 20-22°C during the measurement procedure. Tissuepreparation procedures
Eight surgically excised normal human renal cortex and medulla specimens were classified into the following three groups.
Table 1. Relationship between colour-coded scales and values for attenuation constant and sound speed. Color
GLASS SURFACE
Fig. 2. Relationship
between the ultrasonic the tissue sample.
components
and
Red Magenta Grange Brown Yellow Green Olive green cyan Royal blue Blue Black
Attenuation (dB/mm/MHz) 1.91.7 - 1.9 1.5 - 1.7 1.3 - 1.5 1.1 - 1.3 0.9 - 1.1 0.7 - 0.9 0.5 - 0.7 0.3 - 0.5 0.1 - 0.3 -0.1
Sound speed (m/s) 16901670 - 1690 1650 - 1670 1630 - 1650 1610 - 1630 1590 - 1610 1570 - 1590 1550 - 1570 1530 - 1550 1510 - 1530 -1510
Attenuation Sound speed 5ooJ;n ACOUSTIC IMAGES Fig. 3. Acoustic images of a sample of normal renal cortex in fresh, frozen section.
Sound speed Attenuation c ACOUSTIC IMAGES Fig. 4. Acoustic images of a sample of normal renal cortex in formalin-fixed, frozen set :tion.
Soop;;l
Attenuation
Sound speed
Fig. 5. PLcoustic images of a sample of normal renal cortex in formalin-fixed, paraffin section
1263
1264
Letter to the Editor-in-Chief
Table 2. Attenuation
data in kidney tissue (dB/mm/MHz). Fresh, Frozen
Renal tubule Renal corpuscle Cortex Medulla
1.15 k 0.29 1.08 + 0.28
Formalin, Frozen 0.58 1.22 1.18 1.11
+ + ? ”
0.09 0.22 0.23 0.30
Formalin, Paraffin 0.60 1.27 1.18 1.12
5 2 ? 2
0.11 0.28 0.30 0.28
Group A. Fresh, frozen section. Group B. Ten percent formalin-fixed, frozen section. Kidney tissues were fixed by 10% formalin and frozen in acetone. The frozen specimens, cut 10 pm thick on a cryostat, were mounted onto glass slides. Group C. Ten percent formalin-fixed, paraffin section. Tissues were fixed by formalin, embedded in paraffin and cut in sections 10 pm thick with a microtome. The specimens were mounted onto glass slides but coverslips were not used. The paraffin was removed by the graded alcohol method, just before the ultrasonic measurement. A neighbouring section of the SAM specimen was stained with Elastica-Masson and used for optical microscopy. The region of interest (ROI) for acoustic microscopy was determined from the optical microscopic observations. The size of the ROI was 200 X 200 pm in optical microscopy and 50 X 50 sampling points in acoustic microscopy. The attenuation constant and sound speed were obtained for each tissue element by averaging the values in the ROI. Results are given as mean + SD. Statistical analysis utilized Student’s t-test. A level of p < 0.05 was considered statistically significant.
RESULTS Group A (fresh, frozen section)
Figure 3 shows the acoustic images of a sample of the normal renal cortex. The renal corpuscle and renal tubules were not separately identified in the frozen section. The average values for attenuation constant and sound speed of whole renal cortex were 1.15 dB /mm/MHz and 1611 m/s, respectively. frozen section) Figure 4 shows the acoustic images of a sample of normal renal cortex. In formalin-fixed sections, renal corpuscles and renal tubules were identified. The values for attenuation constant and sound speed were approximately 1.22 dB/mrn/MHz and 1607 m/s, respectively, in renal corpuscle, and 0.58 dB/mm/MHz and 1554 m/s, respectively, in renal tubule. The average values for attenuation constant and sound speed of
whole renal cortex were 1.2 dB /mm/MHz m/s, respectively.
and 1610
Group C (formalin-jixed, parafJin section)
Figure 5 shows the acoustic images of a sample of renal cortex. The values for attenuation constant and sound speed were 1.3 dB/mm/MHz and 1610 m/s, respectively, in renal corpuscle, and 0.6 dB/mm/MHz and 1558 m/s, respectively, in renal tubule. The average values for attenuation constant and sound speed of whole renal cortex were 1.2 dB /mm/MHz and 16 12 m/s, respectively. The mean and SDS of the acoustic parameters of the kidney tissues studied are shown in Tables 2 and 3. Blank entries signify that the tissue elements were not identified. The attenuation constant of the whole renal cortex was 1.15 ? 0.29 dB/mm/MHz in fresh, frozen sections, 1.18 + 0.23 dB /mm/MHz in formalinfixed, frozen sections and 1.18 t 0.30 dB/mm/MHz in formalin-fixed, paraffin sections. Differences were not significant. The sound speed of renal cortex in groups A, B and C was 1611 + 24, 1608 ? 29 and 1612 226 m/s, respectively. There was no significant difference between any two groups. The attenuation constant and sound speed of renal medulla showed no significant difference between any two groups, viz, 1.08 +- 0.28, 1.11 -t 0.30 and 1.12 2 0.28 dB/mm/MHz, respectively, and 1592 2 23, 1590 t 27, and 1592 + 29m/s, respectively. DISCUSSION Acoustic microscopy has been used extensively for quantitative evaluation of the physical properties of soft tissues. The tissue preparation techniques used are generally identical to those used for optical microscopy, and the effects that these preparations have on the acoustic properties are usually not considered. For this reason, it was desirable to determine exactly the influence of the preparation procedures on the measured tissue properties. Previous studies, such as that of Steen et al. ( 1991) , considered the same problems, although their investigations were performed at 5 MHz and the implications for acoustic microscopy inferred. We are not aware of any previous work that has studied
Group B (formalin-fied,
Table 3. Sound speed data in kidney tissue (m/s). Fresh, Frozen Renal tubule Renal corpuscle Cortex Medulla
1611 + 24 1592 2 23
Formalin, Frozen
Formalin, Paraffin
1554 1607 1608 1590
1558 1610 1612 1592
k 2 k 2
10 18 29 27
5 ? t ?
12 17 26 29
Influence
of tissue preparation
the effects of tissue preparation at the actual frequencies used for SAM investigations. Fixation by 10% formalin preserved the acoustic properties of fresh, unfixed kidney tissue. This finding is consistent with previous reports from studies in the low-frequency range (Bamber et al. 1979; Steen et al. 1991) . Fixation is intended to prevent the putrefaction that can occur immediately after tissue excision and to keep the protein so that a sample can be maintained structurally close to the living condition. Ultrasonic identification of tissue elements is possible in formalinfixed tissue sections, but not in fresh, frozen sections; therefore, formalin fixation is preferable for acoustic microscopy. Contrary to the findings of Steen et al. ( 1991)) the present study showed that the acoustic properties of tissue were not influenced by freezing, paraffin embedding and deparaffinizing. The most important difference between the frozen and the paraffin-embedded sections is the presence or absence of fat tissue in the specimen. Thus, a comparison of these two kinds of sections requires use of fat-free areas of tissue. If fatrich tissue. such as liver tissue, is selected, the acoustic properties of the fat cells induce large differences between these two kinds of sections. Our findings demonstrate that, if paraffin is completely removed from the fat-free areas of mounted sections, the acoustic properties of the tissue are not
0 H. SASAKI
I265
et at.
altered by different methods of tissue preparation. The results also suggest that, for fat-rich areas of tissue, acoustic parameters should be assessed only in frozen sections. CONCLUSIONS We investigated the effect of tissue preparation procedures on the measured acoustic properties of normal, human, kidney samples. When acoustic microscopy is used to study soft tissues, formalin-fixed specimens are preferable. Paraffin-embedded sections should be used for acoustic microscopy in fat-free areas of tissue, but frozen sections should be used in fat-rich tissues. REFERENCES Bamber JC, Hill CR. King JA. Dunn F. Ultrasonic propagation through fixed and unfixed tissues. Ultrasound Med Biol 1979; 5:159-165. Hachiya H, Otuki S. Andoh H. Tanaka M. Relation between the sound speed change with time and the physical change in preserved tissues. Jpn J Med Ultrason 1993;20:579-580. Okawai H, Tanaka M, Dunn F, Chubachi N, Honda K. Quantitative display of acoustic properties of the biological tissue elements. Acoustical Imaging 1988: 17: 193-201. Saijo Y. Tanaka M, Okawai H, Dunn F. The ultrasonic properties of gastric cancer tissues obtained with a scanning acoustic microscope system. Ultrasound Med Biol 1991; 17:709-714. Steen AFW, Cuypers MHM, Thijssen JM, Wilde PCM. Influence of histochemical preparation on acoustic parameters of liver tissue: A ~-MHZ study. Ultrasound Med Biol 1991; 17:879-89 I.
Ultrasound in Med. & Biol., Vol. 22, No. 9, pp. 1261- 1265, 1996 0 1996 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/96 $15.00 + 00
PII: SO301-5629( 96)00150-O
ELSEVIER
@Original Contribution INFLUENCE OF TISSUE PREPARATION ON THE HIGH-FREQUENCY ACOUSTIC PROPERTIES OF NORMAL KIDNEY TISSUE HIDEHIKO
SASAKI, YOSHIFUMI SAIJO, MONTONAO TANAKA, HIROAKI OKAWAI, YOSHIO TERASAWA, TOMOYUKI YAMBE and SHIN-ICHI NITTA Department of Medical Engineering and Cardiology, Division of Organ Pathophysiology, Institute of Development, Aging and Cancer, Tohoku University, Aoba-ku, Sendai, Japan (Received 15 January 1996; in final form 1 July 1996)
Abstract-The influence of various tissuepreparations on the acoustic properties of normal kidney tissue at high frequencieswasinvestigated. Eight surgically excisednormal kidney tissuespecimenswere classified into three groups: (i) fresh, frozen section, (ii) formalin&@ frozen section and (iii) formalin-lixed, paraffin section.Scanning acousticmicroscopy operating in the frequency range of 100-200 MHz wasused to display the two-dimensionaldistribution of attenuation constant and sound speed.Our results indicate that (i) there is no sign&ant variation in both acoustic parameters between the three tissue groups, (ii) fixation by 10% formalin producesno significant changein the acoustic parameters, (iii) in fat-free tissue regions, the acoustic parameters are independent of preparation method and (iv) frozen sectionsmust be usedto assess the acoustic parametersin fat-rich tissues. Copyright 0 1996 World Federation for Ultrasound
in Medicine & Biology. Kev Words: Ultrasonic tissue characterization. Acoustic microscopy, Attenuation constant, Sound speed, Tissuepreparation.
duced large changes in the acoustic properties of liver tissue. These studies were not, however, undertaken at the frequencies commonly used for SAM studies. The purpose of this study was to evaluate the influence of preparation techniques on the acoustic properties of normal kidney tissue over the lOO- to 200-MHz range, using a SAM system.
INTRODUCTION Scanning acoustic microscopy (SAM) provides data on physical properties that cannot be obtained by optical microscopy. Very often, formalin-fixed, paraffinembedded specimens are used for SAM investigations because the specimens can be accurately cut into thin sections with smooth surfaces and remain stable during the measurement procedure. However, the manipulations involved in tissue preparation, e.g., fixation, freezing, dehydration, defatting, paraffin embedding, deparaffinization and other steps, may alter the acoustic properties of the tissues. Because fat is eluted during the preparation of paraffin-embedded specimens, frozen sections must be used when measuring the acoustic properties of fat-rich tissues. The influence of formalin fixation on the acoustic properties of the tissue in the low-frequency range (2.55 MHz) has been studied in detail (Bamber et al. 1979; Hachiya et al. 1993). The investigations of Steen et al. ( 1991), which also included looking at the effect of embedding tissues in paraffin and deparaffinizing, in-
METHODS SAM system Figure 1 is a block diagram of the SAM system (Okawai et al. 1988; Saijo et al. 1991). The ultrasonic frequency is variable over the range of 100-200 MHz. The transducer is equipped with a lens having dimensions of diameter, radius of curvature and aperture angle of 1.25 mm, 1.25 mm and 60”, respectively, and having a -3-dB beam width in water (20°C) between 5 pm (at 200 MHz) and 10 pm (at 100 MHz). The transducer is oscillated in the x axis, and the sample holder is scanned in the y axis. The mechanical scanner is arranged so that the ultrasonic beam is transmitted at 4-ym intervals over a 2mm scanned width. The number of sampling points is 480 in one scanning line; 480 X 480 points make one frame within 4 s.
Address correspondence to: Hidehiko Sasaki, M.D., Department of Medical Engineering and Cardiology, Institute of Development, Aging and Cancer, Tohoku University, 4-1 Seiryo-machi, Aoba-ku, Sendai, 980-77 Japan. 1261
1262
Letter to the Editor-in-Chief
five frames of phase images in lo-MHz steps in the range 100-140 MHz. The thickness of the specimen is determined from the frequency-dependent characteristics of the amplitude and the phase of the received signals obtained at the same position for each of the 16 frames (Okawai et al. 1988). The attenuation constant can be calculated by normalizing the attenuation at twice the sample thickness.
JLTRASONICTRANSOUCERS
SIGNAL
PROCESSOR
A = Ll2fd
(3)
where A is the attenuation constant, fis the frequency and d is the thickness of the specimen. The sound speed (C) is calculated by the following equation: Fig. 1. Block diagram of the SAM
system.
C = (l/C,,, - 8/2rfd)-’
Figure 2 is an illustration of the relationship between the ultrasonic beam components and the tissue sample. The detected wave amplitude and phase deviation are described by the following equations: L = -20 log(y,/y,)
(1)
8 = arg(YllY3)
(2)
where L is the amplitude, 0 is the phase deviation and y1 and y3 are the ultrasonic beam components. In the analogue signal processor, the detector of both amplitude and phase is fixed, and an A/D converter is used, in which the error of quantity is approximately 2 m/s for a lo-pm thickness at an ultrasonic frequency of 130 MHz. The original signals for imaging produced in the analogue signal processor can be displayed on the display unit directly. However, these original signals do not have accurate values of attenuation constant and sound speed. The image processor stores the original image signals in 16 frames: 11 frames of amplitude images in lo-MHz steps in the range 100-200 MHz, and
(4)
where C is the sound speed in the specimen, C, is the sound speed in the coupling medium and 8 is the phase deviation. A linear frequency dependence of the attenuation, in the ultrasonic frequency range of 100-200 MHz, was obtained from consideration of the interference of sound wave components (Okawai et al. 1988). The values of the attenuation constant and sound speed thus obtained were converted into colour signals, which corresponded to the values of the amplitude and phase, and were displayed on a colour monitor as a twodimensional distribution. Table 1 shows the relationship between colour-coded scales and values of attenuation constant and sound speed. Distilled water was used for the coupling medium, which was maintained between 20-22°C during the measurement procedure. Tissuepreparation procedures
Eight surgically excised normal human renal cortex and medulla specimens were classified into the following three groups.
Table 1. Relationship between colour-coded scales and values for attenuation constant and sound speed. Color
GLASS SURFACE
Fig. 2. Relationship
between the ultrasonic the tissue sample.
components
and
Red Magenta Grange Brown Yellow Green Olive green cyan Royal blue Blue Black
Attenuation (dB/mm/MHz) 1.91.7 - 1.9 1.5 - 1.7 1.3 - 1.5 1.1 - 1.3 0.9 - 1.1 0.7 - 0.9 0.5 - 0.7 0.3 - 0.5 0.1 - 0.3 -0.1
Sound speed (m/s) 16901670 - 1690 1650 - 1670 1630 - 1650 1610 - 1630 1590 - 1610 1570 - 1590 1550 - 1570 1530 - 1550 1510 - 1530 -1510
Attenuation Sound speed 5ooJ;n ACOUSTIC IMAGES Fig. 3. Acoustic images of a sample of normal renal cortex in fresh, frozen section.
Sound speed Attenuation c ACOUSTIC IMAGES Fig. 4. Acoustic images of a sample of normal renal cortex in formalin-fixed, frozen set :tion.
Soop;;l
Attenuation
Sound speed
Fig. 5. PLcoustic images of a sample of normal renal cortex in formalin-fixed, paraffin section
1263
1264
Letter to the Editor-in-Chief
Table 2. Attenuation
data in kidney tissue (dB/mm/MHz). Fresh, Frozen
Renal tubule Renal corpuscle Cortex Medulla
1.15 k 0.29 1.08 + 0.28
Formalin, Frozen 0.58 1.22 1.18 1.11
+ + ? ”
0.09 0.22 0.23 0.30
Formalin, Paraffin 0.60 1.27 1.18 1.12
5 2 ? 2
0.11 0.28 0.30 0.28
Group A. Fresh, frozen section. Group B. Ten percent formalin-fixed, frozen section. Kidney tissues were fixed by 10% formalin and frozen in acetone. The frozen specimens, cut 10 pm thick on a cryostat, were mounted onto glass slides. Group C. Ten percent formalin-fixed, paraffin section. Tissues were fixed by formalin, embedded in paraffin and cut in sections 10 pm thick with a microtome. The specimens were mounted onto glass slides but coverslips were not used. The paraffin was removed by the graded alcohol method, just before the ultrasonic measurement. A neighbouring section of the SAM specimen was stained with Elastica-Masson and used for optical microscopy. The region of interest (ROI) for acoustic microscopy was determined from the optical microscopic observations. The size of the ROI was 200 X 200 pm in optical microscopy and 50 X 50 sampling points in acoustic microscopy. The attenuation constant and sound speed were obtained for each tissue element by averaging the values in the ROI. Results are given as mean + SD. Statistical analysis utilized Student’s t-test. A level of p < 0.05 was considered statistically significant.
RESULTS Group A (fresh, frozen section)
Figure 3 shows the acoustic images of a sample of the normal renal cortex. The renal corpuscle and renal tubules were not separately identified in the frozen section. The average values for attenuation constant and sound speed of whole renal cortex were 1.15 dB /mm/MHz and 1611 m/s, respectively. frozen section) Figure 4 shows the acoustic images of a sample of normal renal cortex. In formalin-fixed sections, renal corpuscles and renal tubules were identified. The values for attenuation constant and sound speed were approximately 1.22 dB/mrn/MHz and 1607 m/s, respectively, in renal corpuscle, and 0.58 dB/mm/MHz and 1554 m/s, respectively, in renal tubule. The average values for attenuation constant and sound speed of
whole renal cortex were 1.2 dB /mm/MHz m/s, respectively.
and 1610
Group C (formalin-jixed, parafJin section)
Figure 5 shows the acoustic images of a sample of renal cortex. The values for attenuation constant and sound speed were 1.3 dB/mm/MHz and 1610 m/s, respectively, in renal corpuscle, and 0.6 dB/mm/MHz and 1558 m/s, respectively, in renal tubule. The average values for attenuation constant and sound speed of whole renal cortex were 1.2 dB /mm/MHz and 16 12 m/s, respectively. The mean and SDS of the acoustic parameters of the kidney tissues studied are shown in Tables 2 and 3. Blank entries signify that the tissue elements were not identified. The attenuation constant of the whole renal cortex was 1.15 ? 0.29 dB/mm/MHz in fresh, frozen sections, 1.18 + 0.23 dB /mm/MHz in formalinfixed, frozen sections and 1.18 t 0.30 dB/mm/MHz in formalin-fixed, paraffin sections. Differences were not significant. The sound speed of renal cortex in groups A, B and C was 1611 + 24, 1608 ? 29 and 1612 226 m/s, respectively. There was no significant difference between any two groups. The attenuation constant and sound speed of renal medulla showed no significant difference between any two groups, viz, 1.08 +- 0.28, 1.11 -t 0.30 and 1.12 2 0.28 dB/mm/MHz, respectively, and 1592 2 23, 1590 t 27, and 1592 + 29m/s, respectively. DISCUSSION Acoustic microscopy has been used extensively for quantitative evaluation of the physical properties of soft tissues. The tissue preparation techniques used are generally identical to those used for optical microscopy, and the effects that these preparations have on the acoustic properties are usually not considered. For this reason, it was desirable to determine exactly the influence of the preparation procedures on the measured tissue properties. Previous studies, such as that of Steen et al. ( 1991) , considered the same problems, although their investigations were performed at 5 MHz and the implications for acoustic microscopy inferred. We are not aware of any previous work that has studied
Group B (formalin-fied,
Table 3. Sound speed data in kidney tissue (m/s). Fresh, Frozen Renal tubule Renal corpuscle Cortex Medulla
1611 + 24 1592 2 23
Formalin, Frozen
Formalin, Paraffin
1554 1607 1608 1590
1558 1610 1612 1592
k 2 k 2
10 18 29 27
5 ? t ?
12 17 26 29
Influence
of tissue preparation
the effects of tissue preparation at the actual frequencies used for SAM investigations. Fixation by 10% formalin preserved the acoustic properties of fresh, unfixed kidney tissue. This finding is consistent with previous reports from studies in the low-frequency range (Bamber et al. 1979; Steen et al. 1991) . Fixation is intended to prevent the putrefaction that can occur immediately after tissue excision and to keep the protein so that a sample can be maintained structurally close to the living condition. Ultrasonic identification of tissue elements is possible in formalinfixed tissue sections, but not in fresh, frozen sections; therefore, formalin fixation is preferable for acoustic microscopy. Contrary to the findings of Steen et al. ( 1991)) the present study showed that the acoustic properties of tissue were not influenced by freezing, paraffin embedding and deparaffinizing. The most important difference between the frozen and the paraffin-embedded sections is the presence or absence of fat tissue in the specimen. Thus, a comparison of these two kinds of sections requires use of fat-free areas of tissue. If fatrich tissue. such as liver tissue, is selected, the acoustic properties of the fat cells induce large differences between these two kinds of sections. Our findings demonstrate that, if paraffin is completely removed from the fat-free areas of mounted sections, the acoustic properties of the tissue are not
0 H. SASAKI
I265
et at.
altered by different methods of tissue preparation. The results also suggest that, for fat-rich areas of tissue, acoustic parameters should be assessed only in frozen sections. CONCLUSIONS We investigated the effect of tissue preparation procedures on the measured acoustic properties of normal, human, kidney samples. When acoustic microscopy is used to study soft tissues, formalin-fixed specimens are preferable. Paraffin-embedded sections should be used for acoustic microscopy in fat-free areas of tissue, but frozen sections should be used in fat-rich tissues. REFERENCES Bamber JC, Hill CR. King JA. Dunn F. Ultrasonic propagation through fixed and unfixed tissues. Ultrasound Med Biol 1979; 5:159-165. Hachiya H, Otuki S. Andoh H. Tanaka M. Relation between the sound speed change with time and the physical change in preserved tissues. Jpn J Med Ultrason 1993;20:579-580. Okawai H, Tanaka M, Dunn F, Chubachi N, Honda K. Quantitative display of acoustic properties of the biological tissue elements. Acoustical Imaging 1988: 17: 193-201. Saijo Y. Tanaka M, Okawai H, Dunn F. The ultrasonic properties of gastric cancer tissues obtained with a scanning acoustic microscope system. Ultrasound Med Biol 1991; 17:709-714. Steen AFW, Cuypers MHM, Thijssen JM, Wilde PCM. Influence of histochemical preparation on acoustic parameters of liver tissue: A ~-MHZ study. Ultrasound Med Biol 1991; 17:879-89 I.