VALIDATION OF AN ANALYSIS SOFTWARE FOR INTRAVITAL MICROSCOPY WITH OPS IMAGING APPLIED TO THE HAMSTER SKINFOLD CHAMBER 1B. Dadasch , 3A. G. Harris , 2K. R. Kellam & 1F. Christ Clinic for 1Anaesthesiology, Ludwig-Maximilians University Munich, Germany and 2KK Technologies, UK, 3Cytometrics Inc., US

BACKGROUND: Intravital microscopy has been used for many years to make quantitative observations of the microcirculation. The recently developed technique of OPS imaging allows intravital microscopy of the microcirculation in humans and animal models without the use of fluorescent dyes ^[1]. Since the system lacked an adequate software for analysing microvascular diameter and velocity of the images, we developed a new software (Capiscope^(R)) for an semi-automated measurement of these parameters.

GOAL: The purpose of this study was to validate the new Capiscope^(R) software against the previously used software package CapImage^{(R) [2]} .

METHODS: As a basis for our study we used an experiment carried out by Dr. Anthony Harris in the Institute of Surgical Research. Harris used the highly standardised model of the dorsal skinfold chamber of Syrian hamsters. In 10 conscious Syrian hamsters 303 vessels in 7 identical regions were investigated and diameter and velocity were measured at three points in time: under baseline conditions, 0.5 and 2 hours after a 4 hour pressure induced ischemia. Harris et al. ^[2] carried out the experiment using the Cytoscan model E-II and recorded the images on S-VHS videotapes ^[3]. Venular diameter (DIA [µm]) and red blood cell velocity (VEL [mm/s]) were analysed with CapImage^(R).

For our measurements we selected exactly the same vessels at the same points in time and analysed them and analysed them with the new Capiscope^(R) software.

DIA was given by calculating its mean from a gray level scan performed along the length of the vessel (see figure 1) and VEL by conducting an autocorrelation routine (see figure 2).

Statistics: The correlation was assessed using a Pearson-Product- Moment and a Bland-Altman-Plot ^[4].

Figure 1 shows a typical image after measurement and calculation of DIA and VEL with Capiscope^(R).

For the measurement a line is drawn down the length of the vessel and Dia and Vel are displayed along the side of the measured vessel. Simultaneously as a summary of the calculated data a dimension list is provided.

On the right a grey level profile, which reflects the diameter measurement, is displayed.

Figure 2 shows the line scan allowing the assessment of velocity.

RESULTS: Measurement of vessel diameter showed a good correlation giving a value of r² =0.89. When plotting the difference of the vessel diameters obtained with Cap-image^(R) and Capiscope^(R) against the average results of both softwares the Bland Altman plots showed an acceptable concensus ^[3] (see figure 2).

The correlation of the velocity measurements however, was quite poor (r² =0.14). Especially at higher values the velocities calculated with the Capiscope^(R) software were significantly lower (see figure 3).

Figure 2 : Correlation of diameter measured with Capiscope and Cap-image and the corresponding Bland-Altman-plot.

Figure 3 :Correlation of velocity measured with Cap-image^(R) and Capiscope^(R) and the corresponding Bland-Altman-plot

CONCLUSION: Since the previously used Cap-image^(R) software does not represent a Gold Standard it cannot be concluded that the velocity measurements with Capiscope^(R) are inaccurate. Since both methods have been validated sufficiently with the flying wheel, we suggest that these differences should be considered when interpreting results from intravital microscopy obtained with OPS-imaging.

REFERENCES:

[1] Groner W et al. Nat Med 1999;5:1209-12

[2] Harris AG et al. J Vasc Res 2000; 37(6):469-76

[3] Harris AG et al. Prog Appl Microcirc 2000; 24: 21-31

[4] Bland JM et al. Lancet 1995; 346: 1085-7