Image-based vessel-by-vessel analysis for red blood cell and plasma dynamics with automatic segmentation

Manual and Automatic Image Analysis Segmentation Methods for Blood Flow Studies in Microchannels

In blood flow studies, image analysis plays an extremely important role to examine raw data obtained by high-speed video microscopy systems. This work shows different ways to process the images which contain various blood phenomena happening in microfluidic devices and in microcirculation. For this purpose, the current methods used for tracking red blood cells (RBCs) flowing through a glass capillary and techniques to measure the cell-free layer thickness in different kinds of microchannels will be presented. Most of the past blood flow experimental data have been collected and analyzed by means of manual methods, that can be extremely reliable, but they are highly time-consuming, user-intensive, repetitive, and the results can be subjective to user-induced errors. For this reason, it is crucial to develop image analysis methods able to obtain the data automatically. Concerning automatic image analysis methods for individual RBCs tracking and to measure the well known microfluidic phenomena cell-free layer, two developed methods are presented and discussed in order to demonstrate their feasibility to obtain accurate data acquisition in such studies. Additionally, a comparison analysis between manual and automatic methods was performed.

Mapping of flow velocity using spatiotemporal changes in time-intensity curves from indocyanine green videoangiography

OBJECTIVE

The present study developed an image-based analysis method that uses indocyanine green videoangiography (ICG-VA) to measure flow velocity in the arteries and veins of the cortical surface in patients undergoing neurosurgery.

METHODS

MATLAB-based code was used to correct motion artifacts in the ICG-VA and determine the time-intensity curve of the ICG. The slope of the initial increase in ICG intensity following the bolus injection was measured and normalized using the predicted input function in the imaging field. Flow velocity over a certain distance determined by the user was measured based on a time shift of the time-intensity curves along the centerline of the vessels.

RESULTS

The normalized slope of ICG intensity represented the expected differences in the flow velocity among the artery (0.67 ± 0.05 s^-1 ), parenchymal tissue (0.49 ± 0.10 s^-1 ), and vein (0.44 ± 0.11 s^-1 ). The flow velocities measured along the vessel centerline were 2.5 ± 1.1 cm/s and 1.1 ± 0.3 cm/s in the arteries (0.5 ± 0.2 mm in diameter) and veins (0.6 ± 0.2 mm in diameter), respectively.

CONCLUSIONS

An image-based analysis method for ICG-VA was developed to map the expected differences in the flow velocity based on the rising slope of ICG intensity and to measure the absolute flow velocities using the flexible zone and cross-correlation methods.

Neurosurgical intraoperative ultrasonography using contrast enhanced superb microvascular imaging -vessel density and appearance time of the contrast agent

Background: Ultrasonography (US) provides real-time information on structures within the skull during neurosurgical operations. Superb microvascular imaging (SMI) is the latest imaging technique for detecting very low-velocity flow with minimal motion artifacts, and we have reported on this technique for intraoperative US monitoring. We combined SMI with administration of contrast agent to obtain detailed information during neurosurgical operations.Materials and methods: Twenty patients diagnosed with brain tumor (10 meningiomas, 5 glioblastomas, 2 hemangioblastomas, 1 schwannoma, 1 malignant lymphoma, 1 brain abscess) underwent neurosurgery under US with SMI and contrast agent techniques. Vessel density and appearance time following contrast administration were analyzed.Results: Flow in numerous vessels was not visualized by SMI alone, but appeared following injection of contrast agent in all cases. Flow in tumors was drastically enhanced by contrast agent in schwannoma, hemangioblastoma and meningioma, compared to normal brain tissue. Flows in the dilated and bent vessels of glioblastoma were also enhanced, although flow in hypoechoic lymphoma remained inconspicuous. The characteristics of tumor vessels were clearly visualized and tumor borders were demonstrated by the difference between tumor flow and brain flow, by the increased tumor vessel density and decreased appearance time of contrast agent compared to normal brain vessels.Conclusions: The combination of SMI and contrast agent techniques for intraoperative US monitoring could provide innovative flow images of tumor and normal brain. The neurosurgeon obtains information about tumor flow and tumor borderline before tumor resection.

Bridging macroscopic and microscopic methods for the measurements of cerebral blood flow: Toward finding the determinants in maintaining the CBF homeostasis

Methods exist to evaluate the cerebral blood flow (CBF) at both the macroscopic and microscopic spatial scales. These methods provide complementary information for understanding the mechanism in maintaining an adequate blood supply in response to neural demand. The macroscopic CBF assesses perfusion flow, which is usually measured using radioactive tracers, such as diffusible, nondiffusible, or microsphere. Each of them determines CBF based on indicator dilution principle or particle fraction principle under the assumption that CBF is steady state during the measurement. Macroscopic CBF therefore represents averaged CBF over a certain space and time domains. On the other hand, the microscopic CBF assesses bulk flow, usually measures using real-time microscopy. The method assesses hemodynamics of microvessels, ie, vascular dimensions and flow velocities of fluorescently labeled or nonlabeled RBC and plasma markers. The microscopic CBF continuously fluctuates in time and space. Smoothing out this heterogeneity may lead to underestimation in the macroscopic CBF. To link the two measurements, it is needed to introduce a common parameter which is measurable for the both methods, such as mean transit time. Additionally, applying the defined physiological and/or pharmacological perturbation may provide a good exercise to determine how the specific perturbations interfere the quantitative relationships between the macroscopic and microscopic CBF. Finally, bridging these two-scale methods potentially gives a further indication how the absolute CBF is regulated with respect to a specific type of the cerebrovascular tones or capillary flow velocities in the brain.

Regulation of the microvascular circulation in the leg muscles, pancreas and small intestine in rats

To study the microvascular circulation, we examined the proportion of open and functioning capillaries in the leg muscles, pancreas and small intestine of anesthetized rats. Fluorescein isothiocyanate (FITC)-labeled Lycopersicon esculentum lectin was injected into the heart and allowed to circulate for 3 min to label open and functioning capillaries. Specimens were removed, frozen, sectioned and double-immunostained. Using one section, open and functioning capillaries were detected by immunostaining for this lectin bound to endothelial cells, while all capillaries were visualized by immunostaining for platelet endothelial cell adhesion molecule-1 (PECAM-1 or CD31). These capillaries were semi-automatically detected and counted by fluorescence microscopy. The percentages of open and functioning capillaries were as follows: the soleus muscle, 93.0 ± 5.5%; superficial zone of the gastrocnemius muscle, 90.8 ± 6.2%; deep zone of the gastrocnemius muscle, 95.6 ± 4.0%; the plantaris muscle, 94.1 ± 2.7%; the pancreas, 86.3 ± 11.7%; and the small intestine, 91.1 ± 4.9% (n = 8, each). There was no significant difference among these data by the Kruskal–Wallis test. This study clearly demonstrated that the proportions of open and functioning capillaries are high and similar among the leg muscles, pancreas and small intestine in spite of their structural and functional differences. This finding agrees with previous studies and supports the notion that the microvascular circulation is mainly controlled by changing of the blood flow in each capillary rather than changing the proportion of open and functioning capillaries.

Pial arteries respond earlier than penetrating arterioles to neural activation in the somatosensory cortex in awake mice exposed to chronic hypoxia: an additional mechanism to proximal integration signaling?

The pial and penetrating arteries have a crucial role in regulating cerebral blood flow (CBF) to meet neural demand in the cortex. Here, we examined the longitudinal effects of chronic hypoxia on the arterial diameter responses to single whisker stimulation in the awake mouse cortex, where activity-induced responses of CBF were gradually attenuated. The vasodilation responses to whisker stimulation under prehypoxia normal conditions were 8.1% and 12% relative to their baselines in the pial arteries and penetrating arterioles, respectively. After 3 weeks of hypoxia, however, these responses were significantly reduced to 5.5% and 4.1%, respectively. The CBF response, measured using laser-Doppler flowmetry (LDF), induced by the same whisker stimulation was also attenuated (14% to 2.6%). A close linear correlation was found for the responses between the penetrating arteriolar diameter and LDF, and their temporal dynamics. After 3 weeks of chronic hypoxia, the initiation of vasodilation in the penetrating arterioles was significantly extended, but the pial artery responses remained unchanged. These results show that vasodilation of the penetrating arterioles followed the pial artery responses, which are not explainable in terms of proximal integration signaling. The findings therefore indicate an additional mechanism for triggering pial artery dilation in the neurovascular coupling.

Measuring the vascular diameter of brain surface and parenchymal arteries in awake mouse

The present study reports a semiautomatic image analysis method for measuring the spatiotemporal dynamics of the vessel dilation that was fluorescently imaged with either confocal or two-photon microscope. With this method, arterial dilation induced by whisker stimulation was compared between cortical surface and parenchymal tissue in the vibrissae area of somatosensory cortex in awake Tie2-GFP mice in which the vascular endothelium had genetically expressed green fluorescent protein. We observed that a mean arterial diameter during a pre-stimulus baseline state was 39 ± 7, 19 ± 1, 16 ± 4, 17 ± 4, and 14 ± 3 μm at depths of 0, 100, 200, 300, and 400 μm, respectively. The stimulation-evoked dilation induced by mechanical whisker deflection (10 Hz for 5 s) was 3.4 ± 0.8, 1.8 ± 0.8, 1.8 ± 0.9, 1.6 ± 0.9, and 1.5 ± 0.6 μm at each depth, respectively. Consequently, no significant differences were observed for the vessel dilation rate between the cortical surface and parenchymal arteries: 8.8 %, 9.9 %, 10.9 %, 9.2 %, and 10.3 % relative to their baseline diameters, respectively. These preliminary results demonstrate that the present method is useful to further investigate the quantitative relationships between the spatiotemporally varying arterial tone and the associated blood flow changes in the parenchymal microcirculation to reveal the regulatory mechanism of the cerebral blood flow.