The Influence of Mechanical Rubbing on the Dissolution of Blood Clots

Dalia Mahdy, Nabila Hamdi, Sarah Hesham, Ahmed El Sharkawy, and Islam S. M. Khalil

Abstract— Mechanical rubbing of blood clots is a potentialminimally-invasive method for clearing clogged blood vessels.In this work, we investigate the influence of the interactionof the tip of a helical robot with blood clots. This interactionenables the dissolution of the blood clot and the release ofthe entrapped red blood cells and platelets from its three-dimensional fibrin fiber network. We analyze the pre- and post-conditions of the blood clots following 40 minutes of mechanicalrubbing, under the influence of a rotating magnetic field in thefrequency range of 20 Hz to 45 Hz. Our measurements showthat the weight of the blood clots is decreased by 22.5±11.1% atfrequency of 25 Hz. We also validate the influence of mechanicalrubbing using cell count and spectrophotometric analysis onphosphate buffered saline samples past the robot and the clot.The maximum cell count is measured as 654 ± 108 × 104

cells/ml and 54 ± 12 × 104 cells/ml, whereas the absorbanceis measured as 4.35 × 10−6 mol and 1.05 × 10−6 mol underthe influence of mechanical rubbing and without mechanicalrubbing, respectively.

I. INTRODUCTION

The early concepts of the formation of thrombus havebeen described by the German Physician Rudolph Virchowin 1856. The major categories of risk factors of thrombosishave been grouped later on under the name of Virchow’striad which includes injury to the endothelium, abnormalblood flow, and abnormal blood consistency (hypercoagula-bility) [1]. Thrombosis can form in arteries and veins, leadingto their obstruction which in many cases results in seriouscardiovascular diseases. According to the American HeartAssociation, cardiovascular diseases are listed as the leadingcause of death in USA, with coronary heart diseases beingthe leading cause followed by stroke [2]. The gold standardtherapy of thrombosis is the administration of anti-coagulantand thrombolytic drugs. However, this systemic approachmight be associated with risks of serious hemorrhagic com-plications. Targeted drug delivery of anticoagulant drugs tothe blood clot using micro and nano robots could preventthe occurrence of such complications [3]. Researchers havedemonstrated various methods to fabricate and control microand nano robots towards potential biological in vivo ap-plications. For instance, Qiu et al. have characterized theswimming performance of non-cytotoxic artificial bacterialflagella (ABF) to create highly biocompatible helical swim-mers [4]. Servant et al. have also demonstrated optical imag-ing for in vivo tracking of magnetically controlled swarm ofABF in the peritoneal cavity of a mouse [5]. Power et al.have used biocompatible hydrogels in the fabrication ofhybrid microrobots for tumor targeting and treatment [6].

The authors are affiliated with the German University in Cairo, Egypt.This work was supported by the Science and Technology Development Fundin Egypt (No. 23016) and the DAAD-BMBF funding project.

Fig. 1. Mechanical rubbing of blood clots is achieved using a helical robotin vitro. (a) A schematic representation shows the interaction of the tip ofthe helical robot with the fibrin network of the blood clot. (b) A blood clotis prepared and inserted inside a catheter segment and mechanical rubbingwith a helical robot is achieved experimentally. The helical robot rotatesunder the influence of a rotating magnetic field. The diameter of the robotis 346 µm and its length is 4 mm.

Recently, Adams et al. have also achieved wireless controlof microrobots in a rabbit bladder for endoscopic applicationsin urology [7].

In this work, we focus on the removal of blood clots usingmagnetically actuated helical robots, as shown in Fig. 1(a).Mechanical removal of blood clots represents an attractivetherapeutic modality that could prevent the complicationsrelated to systemic therapy as presented in our previouswork [8], [9]. Thus, we resume to investigate the influenceof mechanical rubbing using helical robots on the dissolutionof blood clots and achieve the following:

• measure the influence of rubbing frequency on theweight of the blood clots;

• determine the influence of the rubbing frequency on thered blood cell (RBCs) and platelets count;

• spectrophotometric analysis on samples past the clotto measure the concentration of released RBCs andplatelets from the fibrin network of the clot.

Experimentally, mechanical rubbing is achieved using ahelical robot [Fig. 1(b)]. As blood clots are formed of afibrin network where blood cells are entrapped, mechanicalrubbing of the fibrin network results in the release of RBCsand platelets. Samples past the robot and the blood clot are

978-1-5386-3646-6/18/$31.00 ©2018 IEEE 1660

collected at different time intervals during the experimentsfor analysis. We calculate the change in the weight of theblood clot after mechanical rubbing, as well as cell countand absorbance of the samples. The remainder of this paperis organized as follows: Section II provides descriptions per-taining to the interaction between the tip of the helical robotwith the fibrin network of the blood clot. Characterizationof the influence of rubbing frequency on the dissolution ofblood clots is discussed in Section III. Finally, Section IVconcludes and provides directions for future work.

II. CHARACTERIZATION OF THE INFLUENCE OFMECHANICAL RUBBING ON THE PROPERTIES OF CLOTS

Following our previous work [9], the helical robot isdirected inside a catheter segment towards a blood clotagainst the flowing streams of phosphate buffered saline(PBS). Mechanical rubbing is achieved once the robot comesinto contact with the clot.

A. Helical Propulsion and System Description

The helical robot consists of a magnetic head and a screwlike tail. The tail is fabricated using an aluminum spring(length, diameter, and pitch of 2.2 mm, 346 mm, and 580mm, respectively). This spring is rigidly attached to a perma-nent magnet (Neodymium grade N40) with an edge lengthof 500 µm and a magnetization vector perpendicular to thespring axis. Helical propulsion is achieved using two rotatingpermanent magnets [10]. Each permanent magnet generatesmagnetic field of 0.552 T on its surface. The permanentmagnets are placed parallel to the catheter segment [11].Blood clots are prepared [12], with length of 7.5 mm anddiameter of 4 mm, and inserted inside the segment. Wecalculate Reynolds number as Re = ρ|U|L

µ = 0.004, whereρ is the density of the PBS (995 kg.m−3), U is the velocityof the helical robot before rubbing, L is its length (4×10−3

m), and µ is the dynamic viscosity of the PBS (0.8882 Pa.s).The hydrodynamic drag experienced by the helical robotin low Reynolds numbers can be simplified by employingthe resistive-force theory [13], in which the drag force andtorque are related with the linear and angular velocities viaan anisotropic operator as follows:(

Fd

Td

)= Zr

(U+Uch

Ω

), (1)

where Fd and Td denote the viscous drag force and torquevectors, respectively. Further, Ω is the rigid-body angularvelocity of the helical robot. Uch represents the flow-fieldsinside the channel, and Zr is a matrix of the resistive coef-ficients of the head and tail [9]. The drag torque Td exertedon the helical robot is linearly proportional to the rigid-bodyangular velocity of the helical robot (Ω). Hence, we deducethat the exerted drag torque is linearly proportional to theactuation frequency of mechanical rubbing below the stepout frequency of the helical robot. Therefore, we achievemechanical rubbing at different actuation frequencies andanalyze the pre- and post-conditions (weight, cell count, andabsorbance) of the clots.

TABLE ITHE DECREASE IN WEIGHT OF THE BLOOD CLOTS IS MEASURED AFTER

40 MINUTES OF MECHANICAL RUBBING, UNDER THE INFLUENCE OF A

ROTATING MAGNETIC FIELD WITH VARYING FREQUENCY IN THE RANGE

OF 20 HZ TO 45 HZ. AVERAGES AND STANDARD DEVIATIONS ARE

CALCULATED FROM 3 TRIALS.

f [Hz] wd [%] f [Hz] wd [%]20 19.5± 5.1 25 22.5± 11.1

30 15.9± 4.5 35 20.7± 5

40 21.6± 8.8 45 13.8± 2.6

B. Analysis of the Blood Clot Weight

Mechanical rubbing of blood clots results in dissolutionand a volume decrease, which implies a loss in weight.Hence, we measure the weight of blood clots before andafter 40 minutes of mechanical rubbing using an electronicbalance (ABS 220-4 Analytical Balance, KERN & SOHNGmbH, Balingen, Germany). The percentage of decrease inthe weight of the blood clot (wd) is calculated using

wd =w(tf)− w(t0)

w(t0)× 100. (2)

In (2), w(tf ) is the weight of blood clot after 40 minutesof mechanical rubbing and w(t0) is the initial weight ofthe blood clot. Measurements of wd under the influence ofrotating magnetic fields with varying frequency in the rangeof 20 Hz to 45 Hz are shown in Table 1. An increase in wd

is observed for all actuation frequencies f , as opposed tof = 0 Hz (zero-input response). This increase validates thedissolution of blood clots under the influence of mechanicalrubbing. However, the weight decrease does not describeother changes related to the effect of the mechanical rubbingon the fibrin network of the blood clot. Hence, we calculatethe cell count of the PBS samples past the robot and bloodclot, which indicates the dissolution of the fibrin networkcausing the release of red blood cells (RBCs) and plateletspreviously entrapped within the blood clot.

C. Analysis of the Red Blood Cells and Platelets Count

The interaction between the rotating tip of the helical robotand the blood clot allows the RBCs and platelets to breakfree from the fibrin network. We experimentally study theinfluence of rubbing frequency on the cell count. Mixturepast the robot and the blood clot is collected every 5 minutesinside small tubes [Fig. 2(a)]. A hemocytometer (NeubauerImproved, Germany), a device originally designed for count-ing blood cells, is used to calculate the number of RBCs andplatelets for each sample [Fig. 2(b)] under microscope (MFSeries 176 Measuring Microscopes, Mitutoyo, Kawasaki,Japan). Averaged sum of cell count after 40 minutes ofrubbing provides a maximum value of 654 ± 108 × 104

cells/ml at f = 40 Hz, compared to 54± 12× 104 cells/mlin the absence of mechanical rubbing [Fig. 3(c)]. Averagesand standard deviations are calculated from 3 trials. The

1661

Fig. 2. Number of red blood cells (RBCs) and platelets is analyzed tostudy the influence of mechanical rubbing on blood clots. Mixture past therobot and the blood clot is collected every 5 minutes and analyzed usinga hemocytometer to calculate the cell count. (a) RBCs and platelets arecounted during mechanical rubbing with frequency in the range of 20 Hzto 45 Hz, and against flow rate of 10 ml.hr−1. (b) Total sum of RBCs andplatelets provides maximum of 646 ± 206 × 104 cells/ml at f = 40 Hzcompared to 54 ± 12 × 104 cells/ml at f = 0 Hz (zero-input response).Averages and standard deviations are calculated using 3 trials.

cell count indicates an increase in the number of releasedRBCs and platelets from the fibrin network of the bloodclot, validating the mechanical breakdown of fibrin by therotating tip of the helical robot.

D. Spectrophotometric Analysis of Blood Clots

Spectrophotometric analysis is conducted to measure theabsorbance to indicate the concentration of the blood clot and

validate the efficiency of mechanical rubbing. PBS samplespast the robot and blood clot are collected every 5 minutesduring the experiments [Fig. 3(a)]. The absorbance of thecollected samples is measured using a spectrophotometer (V-730 UV-Visible Spectrophotometer, Oklahoma city, USA), asshown in Fig. 3(b). The spectophotometer generates a beamof light that interacts with the tested sample and measures theamount of light absorbed by the sample over a certain rangeof wavelength. This measurement allows for the calculationof the amount of substance in the tested sample using Beer-Lambert law, which relates the attenuation of light to theproperties of the material which the light travels throughas follows:

A = ϵbc. (3)

In (3), A is the absorbance measured by spectrophotometry,ϵ is the wavelength-dependent molar absorptivity coefficient,b is the path length of the cuvette in which the sampleis contained, and c is the compound concentration. First,a base line measurement of the samples is done for theoptimum wavelength selection within the range of visiblelight (400 nm to 800 nm). Maximum absorbance is measuredat wavelength (λ = 416 nm), as shown in Fig. 3(c). Second,absorbance of each sample is measured at the selectedwavelength. Finally, Beer-Lambert Law is used to calculatethe concentration where ϵ = 521880 as reported in [14]and b = 1 cm. Absorbance is linearly proportional to theconcentration of the substance in the same sample. Thus,the increase in the absorbance implies an increase in theconcentration and correspondingly an increase in the numberof blood cells released from the fibrin fiber network of theblood clot [Fig. 1(a)]. Concentration of samples calculatedpast the robot and blood clot, under the influence of variousfrequencies of rotating magnetic fields in the range of 20 Hzto 45 Hz, is shown in Fig. 3(c). Maximum total concentrationis measured as 4.35× 10−6 mol at f = 35 Hz, compared to1.05× 10−6 mol in the absence of mechanical rubbing.

The pre- and post-conditions of the blood clots are an-alyzed following 40 minutes of mechanical rubbing. Threemethods are used to validate the influence of mechanicalrubbing using helical robots. While the weight of the bloodclot samples provides an essential measure (Table 1) to theinfluence of mechanical rubbing, the samples are affectedduring handling (extraction from the catheter segments).Therefore, cell count and concentration measurements areused to provide a comprehensive information pertaining tothe pre- and post-conditions of the clots. We also find qualita-tive agreements between the mentioned three techniques. Thereduction in weight (Table 1), number of released red bloodcells and platelets [Fig. 2(c)], and concentration [Fig. 3(c)]are in qualitative agreement and indicate the influence ofmechanical rubbing at relatively high actuation frequency ofthe helical robot, as opposed to zero-input response. Ourmeasurements do not provide an optimal rubbing frequencyowing to the clot-to-clot variability. Nevertheless, maximumremoval rate is observed within a frequency range of 30 Hzto 40 Hz.

1662

Fig. 3. Spectrophotometric analysis is performed to study the influence ofmechanical rubbing on the concentration of blood clots. (a) Mixture past therobot and the blood clot is collected every 5 minutes (b) and analyzed usinga spectophotometer (V-730 UV-Visible Spectrophotometer, Oklahomacity,USA). (c) First, base line is selected as λ = 416 nm, then the absorbanceis measured under the influence of a rotating magnetic field with varyingfrequency in the range of 20 Hz to 45 Hz, and against flow rate of 10ml.hr−1. Total concentration provides maximum of 4.35 × 10−5 mol atf = 35 Hz compared to 1.05× 10−5 at f = 0 Hz.

III. CONCLUSIONS AND FUTURE WORK

We investigate the influence of mechanical rubbing onthe weight of blood clot, RBCs and platelets count, andabsorbance of collected samples past the helical robot andthe blood clot. The maximum decrease in weight is measuredas 22.5 ± 11.1% at f = 25 Hz. The maximum cell countin the mixture past the clot and the robot is measured as654 ± 108 × 104 cells/ml (n = 3) at f = 40 Hz and54 ± 12 × 104 cells/ml (n = 3) under the influence ofmechanical rubbing and without mechanical rubbing, re-spectively. In addition, maximum absorbance is measured as4.35×10−6 mol (n = 2) at f = 35 Hz and 1.05×10−6 mol(n = 2) in the presence and absence of mechanical rubbing,respectively. Therefore, our measurements show a significanteffect for the helical robot on the removal of the blood clotsduring mechanical rubbing.

As part of future studies, we will achieve mechanicalrubbing of blood clots in combination with chemical lysis atdifferent doses of fibrinolytics and anticoagulant drugs. Thiscomparative study between mechanical rubbing, rubbingin combination with different percentages of thrombolyticagents, and chemical lysis is essential to optimize the inte-gration between mechanical rubbing and chemical lysis ofblood clots. In addition, we will characterize the locomotion

behavior of helical robot inside animal blood vessels tomimic the physiological environment and our system willalso incorporate a medical imaging modality to track thehelical robot in vivo [15].

IV. ACKNOWLEDGMENTS

We thank Dr. R. Hanafi for assistance with the spectropho-tometric analysis. We would also like to thank Mr. I. Baslafor assistance with the experimental work.

REFERENCES

[1] B. C. Dickson, “Venous Thrombosis: On the History of Virchow’sTriad,” University of Toronto Medical Journal, vol 81, no. 3, pp. 166–171, May 2004.

[2] E. J. Benjamin et al., “Heart disease and stroke statistics-2018 update:a report from the American Heart Association,” circulation journal,January 2018.

[3] R. A. Freitas Jr., “Nanotechnology, nanomedicine and nanosurgery,”International Journal of Surgery, vol. 3, no. 4, pp. 243-246,April 2005.

[4] F. Qiu, L. Zhang, K. E. Peyer, M. Casarosa, A. Franco-Obregn, H.Choi, and B. J. Nelson, “Noncytotoxic artificial bacterial flagella fab-ricated from biocompatible ORMOCOMP and iron coating,” Journalof Materials Chemistry B, vol. 2, no. 4, pp. 357–362, November 2013.

[5] A. Servant, F. Qiu , M. Mazza, K. Kostarelos, and B. J. Nelson,“Controlled in vivo swimming of a swarm of bacteria-like micro-robotic flagella,” Advanced Materials, vol. 27, no. 19, pp.2981–2988,April 2015.

[6] M. Power, S. Anastasova, S. Shanel, and G. Yang, “Towards hybridmicrorobots using pH-and photo-responsive hydrogels for cancer tar-geting and drug delivery,” in Proceedings of the IEEE InternationalConference on Robotics and Automation(ICRA), pp. 6002-6007, Sin-gapore, May 2017.

[7] F. Adams, T. Qiu, A. Mark, k. Melde, S. Palagi A. Miernik, andP. Fischer, “Wireless micro-robots for endoscopic applications inurology,” European Urology supplements, vol. 16,no. 3, PP. 1914,March 2017.

[8] A. Hosney, J. Abdalla, I. S. Amin, N. Hamdi, and I. S. M. Khalil,“In vitro validation of clearing clogged vessels using microrobots,”in Proceedings of the IEEE RAS/EMBS International Conference onBiomedical Robotics and Biomechatronics (BioRob), pp. 272–277,Singapore, June 2016.

[9] I. S. M. Khalil, A. F. Tabak, K. Sadek, D. Mahdy, N. Hamdi, andM. Sitti, “Rubbing against blood clots using helical robots: modelingand in vitro experimental validation,” IEEE Robotics and AutomationLetters, vol. 2, no. 2, pp. 927–934, April 2017.

[10] M. E. Alshafeei, A. Hosney, A. Klingner, S. Misra, and I. S. M. Khalil,“Magnetic-Based motion control of a helical robot using two synchro-nized rotating dipole fields,” in Proceedings of the IEEE RAS/EMBSInternational Conference on Biomedical Robotics and Biomechatron-ics (BioRob), pp. 151-156, Sao Paulo, Brazil, August 2014.

[11] A. Hosney, A. Klingner, S. Misra, and I. S. M. Khalil, “Propulsionand steering of helical magnetic microrobots using two synchronizedrotating magentic field in three-dimensional space,” in Proceedingsof the IEEE/RSJ International Conference of Robotics and Systems(IROS), pp. 1988-1993, Hamburg, Germany, November 2015.

[12] A. Hoffmann and H. Gill, “Diastolic timed vibro-percussion at 50Hz delivered across a chest wall sized meat barrier enhances clotdissolution and remotely administered streptokinase effectiveness inan in-vitro model of acute coronary thrombosis,” Thrombosis Journal,vol. 10, no. 23, pp. 116, November 2012.

[13] J. Gray and G. J. Hancock, “The propulsion of sea-urchin sper-matozoa,” Journal of Experimental Biology, vol. 32, pp. 802814,December 1955.

[14] S. A. Prahl, “Optical absorption of hemoglobin,” Oregon MedicalLaser Center, December 1999.

[15] I. S. M. Khalil, D. Mahdy, A. El Sharkawy, R. R. Moustafa, A. F.Tabak, M. Elwi, S. Hesham, N. Hamdi, A. Klingner, A. Mohamed,and M. Sitti, “Mechanical rubbing of blood clots using helical robotsunder ultrasound guidance,” IEEE Robotics and Automation Letters,vol. 3, no. 2, pp. 1112–1119, April 2018.

1663