A comparison of the quality of image acquisition between two different sidestream dark field video-microscopes

Sidestream dark field (SDF) imaging enables direct visualisation of the microvasculature from which quantification of key variables is possible. The new MicroScan USB3 (MS-U) video-microscope is a hand-held SDF device that has undergone significant technical upgrades from its predecessor, the MicroScan Analogue (MS-A). The MS-U claims superior quality of sublingual microcirculatory image acquisition over the MS-A, however, this has yet to be robustly confirmed. In this manuscript, we therefore compare the quality of image acquisition between these two devices. The microcirculation of healthy volunteers was visualised to generate thirty video images for each device. Two independent raters, blinded to the device type, graded the quality of the images according to the six different traits in the Microcirculation Image Quality Score (MIQS) system. Chi-squared tests and Kappa statistics were used to compare not only the distribution of scores between the devices, but also agreement between raters. MS-U showed superior image quality over MS-A in three of out six MIQS traits; MS-U had significantly more optimal images by illumination (MS-U 95% optimal images, MS-A 70% optimal images (p-value 0.003)), by focus (MS-U 70% optimal images, MS-A 35% optimal images (p-value 0.002)) and by pressure (MS-U 72.5% optimal images, MS-A 47.5% optimal images (p-value 0.02)). For each trait, there was at least 85% agreement between the raters, and all the scores for each trait were independent of the rater (all p-values > 0.05). These results show that the new MS-U provides a superior quality of sublingual microcirculatory image acquisition when compared to old MS-A


Background
Sublingual video-microscopy is becoming an increasingly important clinical technique used for real-time assessment of the in-vivo microcirculation [1]. The technology permits evaluation of several variables including vessel density, perfusion indices (such as the proportion of perfused vessels and microvascular flow index), and the heterogeneity of the blood flow throughout the capillary bed. Through measuring these variables, sublingual video-microscopy directly quantifies the microcirculation, and this is essential given that it can bear no resemblance to common 'macro-circulation'-variables such as blood pressure which we usually quantify and then make microcirculatory inferences from [2]. Additionally, studies have shown it is possible to measure variables related to leucocytes, including quantity and kinetics [3][4][5]. In light of this, video-microscopy therefore offers the potential to optimize treatment of the microvasculature, particularly fluid management and inotropic support in critically ill patients [6].
Since the advent of orthogonal polarisation spectroscopy in 1971 [7] numerous methods have been developed to illuminate the microcirculation including sidestream-(SDF) [8] and incident-dark field imaging (IDF) [9]. The technique exploits the process of incident dark field illumination, whereby blood vessels < 100 µm in diameter, and < 1000 µm below the surface of the organ, are illuminated and visualised in a two-dimensional plane. Both SDF and IDF illuminate the microcirculation using a series of concentrically placed light emitting diodes (LEDs) surrounding a central light guide that contains the lens system. This structure optically isolates the lens from the illuminating outer ring of LEDs, thus preventing contamination of the image with tissue surface reflections [7]. Pulsed green light (wavelength 540 ± 10 nm) that is in synchrony with the video camera frame rate, performs intra-vital stroboscopy, with short illumination times used to help to prevent the smearing of moving objects such as flowing red cells, and the motion-induced blurring of capillaries [10].
The first SDF camera, the MicroScan Analogue (MS-A), was released by Microvision Medical, (Amsterdam, The Netherlands) in 2007. In 2012, Braedius Medical (Huizen, The Netherlands) introduced a new sublingual videomicroscope-the Cytocam IDF, and this demonstrated significantly superior image acquisition when compared to the MS-A [11]. In 2018, Microvision Medical revealed their new and updated version, the MicroScan USB3 (MS-U), claiming an improved quality of the data acquisition compared to their earlier model-the MS-A. The updated camera has a number of objective improvements compared to its predecessor (see Table 1; Fig. 1), including a higher camera resolution, an increased frame rate, a much lower weight (predominantly due to its custom built camera as opposed to its predecessors use of a bulkier third-party camera), and a conversion from analogue to digital image capture. Although these improvements would imply that the MS-U should demonstrate significant superiority in terms of the quality of image acquisition over its predecessor, this has not been validated and requires confirmation. This study therefore directly compares the upgraded 2018 Microvision MS-U camera with the previous 2007 analogue MS-A model.

Methods
Ethical approval for the study was obtained from University College London Research and Ethics Committee. A total of sixty videos (30 for each device), were obtained from healthy volunteers who had given informed consent. The data capture was carried out in a single laboratory (London, UK). Volunteers rested for ten minutes in the supine position before images were obtained whereby the investigator positioned and focused the cameras under the participants' tongue. Ten seconds of video footage were digitally recorded onto the computer, where images were stored for later analysis. This process was repeated on each participant until six good quality recordings, three from each device, had been acquired from separate areas of the sublingual region. The order of use of the device was randomly generated. All images were obtained by one of two researchers, both of whom were experienced in using the video microscopes. The videos were taken according to the new video-microscopy consensus guidelines [12].
After video acquisition, the videos were saved onto a hard drive and were then reviewed on the same computer using Windows Media Player (Microsoft Corporation, Washington, US) without any pre-processing. Two raters (JC, EGK) blinded to the device on which the video file was recorded, independently graded the films according to the Microcirculation Image Quality Score (MIQS) system [13] (Table 2). With this semi-objective approach to grading the quality of image acquisition prior to analysis, each of the six categories is graded as 0 (optimal), 1 (acceptable) or 10 (unacceptable). If the total of the six categories is > 10, then the video is unsuitable for analysis. Chi-squared tests were used to determine whether scores (optimal and acceptable) for each trait were independent of the Rater and of the Video-microscope. Agreement between Raters and agreement between Video-microscopes were assessed using Kappa statistic. Agreement was not due to chance for values of Kappa statistic > 0.60. The two-tailed significance level was set at 0.05, and R(version 3.4.3) was used for the analyses.

Results
All 60 videos were analysed by both raters, and no problems were encountered. The distribution of scores by rater is shown in Fig. 2.
MS-U was rated as having superior image quality over MS-A in three of out six MIQS traits (Table 2). MS-U captured significantly more optimal images in terms of; (i) illumination (MS-U 95% optimal images, MS-A 70% optimal images (p-value 0.003)); (ii) focus (MS-U 70% optimal images, MS-A 35% optimal images (p-value 0.002)); and (iii) pressure (MS-U 72.5% optimal images, MS-A 47.5% optimal images (p-value 0.02)). There was no significant difference between the content capture of the two video-microscopes (MS-U 77.5% optimal images, MS-A 80% optimal images (p-value 0.79)), and both techniques demonstrated 100% optimal images acquisition in terms of duration and stability (Table 3). Please see Fig. 3 for example screenshots of higher and lower quality videos taken using MS-U and MS-A.
Agreement between the two raters was good, as evidenced by being 85% or over for each trait tested, and all kappa values were over 0.60 demonstrating these results were not due to chance (Table 4). Additionally the scores for each trait were independent of the rater (all p-values > 0.05) ( Table 4).

Discussion
These results demonstrate for the first time, that the MS-U video microscope is superior to MS-A video microscope in terms of the quality of image acquisition. The agreement between the raters on each MIQS trait was at least 85% with Kappa statistics of over 0.63, a positive indicator of the reliability of the study. Using the total score value to determine if an image was deemed suitable or unsuitable for analysis, there was 100% agreement between the two raters. The categories of the MIQS that showed the greatest difference between the two cameras were illumination, focus and pressure. The former two may be as a result of the new illumination management system, and also the improved optical resolution of the MS-U. The improvement in pressure scoring may be because the MS-U device is lighter and therefore less prone to a pressure artifact.
No difference was seen in the duration, stability and content of image capture, however, this is unsurprising given that these traits are generally independent of the device used. Duration and stability, were 100% optimal across both devices. This is likely to have been because these two traits in particular are less dependent on the device being used, but more dependent on the person capturing the images, and the subject's anatomy and degree tongue movement. Additionally, in the updated MS-U, the software stops filming after a specific time frame, and this can be preset prior to image capture. Whilst this study found significant differences between MS-U and MS-A, the Cytocam IDF video-microscope has also been shown to be superior to the MS-A [11]. Unfortunately it is not possible to make comparisons between the Cytocam IDF and MS-U using these two independent studies, however, one contrasting feature of this study compared to the Cytocam IDF vs. MS-A study, is that no videos in this study were scored as unacceptable [11]. A future study directly comparing the Cytocam IDF and MS-U is therefore warranted, as results obtained from video-microscopy  [13].
Strengths of this paper include the agreement witnessed between the two raters and the size of the p-values demonstrated in the results, with all significant p-values being at least 0.02 or below. Limitations are also evident, perhaps the foremost being that the MIQS still relies on subjective rater assessment of the videos. This is however, still the gold-standard approach for grading images of the microcirculation prior to variable analysis. Another limitation is that we have only compared these video-microscopes, on one capillary bed location in the body. Although the sublingual microvasculature is currently the most widely investigated, further work involving other capillary beds should use the results of this study with caution.
Notably further studies should be considered regarding video-microscopy image acquisition and analysis. Whilst this study has solely measured and compared the quality of image acquisition between two devices, it has not considered the recently developed automated analysis software that has been validated using IDF [14], enabling automated processing of the images, thus providing objective figures such as microcirculatory flow index. Of note, however, this software is reliant on high image quality to work [14]. As manual image analysis is both subjective in nature, and a very time consuming process, automated analysis is the key to enabling sublingual video microscopy to be used at the

Conclusions
In this study we have established that the latest MicroVision SDF video-microscope demonstrates superior image acquisition when compared to its predecessor. In three out of six MIQS categories -illumination, image focus and avoidance of pressure artifacts, the MS-U out-performed the MS-A. The findings therefore support the claims made by the manufacturers claiming superior image acquisition over the MS-A. With its optimal degree of image capture, the MS-U better portrays the underlying sublingual microcirculation, and should therefore be used for its real-time assessment.
Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.