- Email: [email protected]
Grinding wheel motion, force, temperature, and material removal in rotational atherectomy of calcified plaque
CIRP Annals - Manufacturing Technology 65 (2016) 345–348
Contents lists available at ScienceDirect
CIRP Annals - Manufacturing Technology jou rnal homep age : ht t p: // ees .e lse vi er . com /ci r p/ def a ult . asp
Grinding wheel motion, force, temperature, and material removal in rotational atherectomy of calcified plaque Albert J. Shih (1)a,b, Yao Liu a, Yihao Zheng a,* a b
Mechanical Engineering, University of Michigan, Michigan, USA Biomedical Engineering, University of Michigan, Michigan, USA
A R T I C L E I N F O
A B S T R A C T
Keywords: Biomedical Grinding Mechanism
This study investigates the grinding wheel motion, force, material removal, and temperature in rotational atherectomy (RA). RA utilizes a metal-bond diamond wheel to remove plaque from arteries to treat cardiovascular diseases. As a plaque surrogate, a bone workpiece was placed in a vessel simulator and subjected to RA with a wheel rotational speed of 160,000 rpm. This grinding process was monitored by a high-speed camera, a dynamometer, and embedded thermocouples. The results show this process has a 108 Hz wheel orbital frequency, an oscillating grinding force of 0.23 N, 90% debris smaller than 31 mm, and a 4.1 8C temperature rise. ß 2016 CIRP.
1. Introduction Rotational atherectomy (RA), as shown in Fig. 1, is a grinding process to remove plaque from arterial walls to restore blood flow. RA utilizes a high-speed metal-bond diamond grinding wheel (1.25–2.5 mm diameter and up to 210,000 rpm [1]) for the treatment of several cardiovascular diseases, e.g. the calcification (hardening) of plaque, arterial bifurcation, ostial stenosis, and instent restenosis in which RA is preferred over traditional angioplasty [2]. The grinding wheel is driven by a long flexible drive shaft and rotates around a stationary guidewire (0.23 mm diameter). The shaft rotates within a sheath where saline flows (marked by arrows in Fig. 1) for lubrication and cooling. Blood flows outside the sheath. Clinically, RA has high complication rates [1]: restenosis – the regrowth of the plaque after RA – occurs in over 50% patients; and
Fig. 1. Rotational atherectomy – the plaque grinding process. * Corresponding author. E-mail address: [email protected] (Y. Zheng). http://dx.doi.org/10.1016/j.cirp.2016.04.012 0007-8506/ß 2016 CIRP.
other complications include myocardial infarction, dissection, perforation, slow-flow/no-reflow, vasospasm, and grinding wheel entrapment. Extensive clinical studies have been conducted to lower complication rates. A current universally accepted protocol suggests an initial grinding wheel size of 1.25–1.5 mm with a rotational speed ranging from 135,000 to 180,000 rpm [3]. Safian et al. [4] suggested avoiding excessive rotational speed decreases (>5000 rpm) as this indicates large grinding forces that may damage healthy tissue. Kini et al. [5] recommended a plaque grinding time between 20 and 30 s to prevent vessel dissection and thermal damage, with a grinding wheel diameter less than 70% of the treated artery diameter to avoid over-stretching the lesion. Lin et al. [6] found that advancing the grinding wheel across long, angulated, and heavily calcified lesions may cause a steep decrease of the rotational speed and entrap the wheel, suggesting a slow and steady advance of the wheel for less than 15 s. Engineering studies in RA have thus far been limited to investigation into grinding wheel design and rotational speed. For instance, Kim et al. [7] modified the grinding wheel surface by laser engraving to reduce microcavitation and tissue damage. Nakao et al. [8] created micro-blades on the wheel surface to replace diamond abrasives. Reisman et al. investigated the effects of rotational speed on platelet aggregation [9] and tissue thermal injury [10]. However, our review of the literature has found that there is a lack of knowledge in the grinding process and plaque removal mechanism in RA. Understanding the grinding wheel motion, force, debris size, ground surface, and temperature is critical to ultimately improving RA techniques and devices. Because of the high rotational speeds, the flexibility of the drive shaft, and poor visibility and accessibility of the grinding site within arteries, investigation of these grinding mechanisms is and has been challenging. To address these issues, this study investigated the RA grinding process within a semi-transparent soft tissue-mimicking phantom embedded with a
346
A.J. Shih et al. / CIRP Annals - Manufacturing Technology 65 (2016) 345–348
ring-shape bovine bone workpiece as a plaque surrogate. To measure the grinding wheel motion, force, and temperatures during the grinding process, a high-speed camera, piezoelectric dynamometer, and embedded thermocouples were utilized, respectively. RA device and experimental setup and design are first introduced. Results on grinding wheel motion, force, debris size, surface topography, and temperatures of the blood-mimicking fluid and plaque surrogate are presented and discussed. 2. Experimental setup The experimental setup, as shown in Fig. 2, consists of three modules: (1) a rotational atherectomy device, (2) tissue phantom, and (3) measurement system.
shaft, the guidewire, and the sheath is reduced by flowing saline inside the sheath (shown in Fig. 1). The grinding wheel, as shown in Fig. 3, is an ellipsoid with the distal half coated with diamond abrasives with an average size of 10 mm, as shown in the scanning electron microscope (SEM) image. A 2.5 mm diameter grinding wheel was used in this study. The air turbine drives the grinding wheel from 6000 to 210,000 rpm. An advance knob, as shown in Fig. 2, enables the manual movement of the grinding wheel in the axial direction during the procedure. In this experiment, the knob was driven by a linear stage to control the wheel axial motion. An air pressure regulator sets the rotational speed of the grinding wheel.
Fig. 2. Experimental setup.
2.1. Rotational atherectomy device
2.2. Tissue phantom
The RA device, RotablatorTM by Boston Scientific, includes a catheter, an air turbine, a grinding wheel advancing unit, and an air pressure regulator with an information display panel. The catheter, as shown in Fig. 3, is inserted into the artery during RA. The stationary guidewire extends from the air turbine through the drive shaft and grinding wheel to beyond the plaque (Fig. 1). The guidewire guides the rotation and translation of the grinding wheel and drive shaft. As illustrated in the cross-section A-A in Fig. 3, the catheter consists of a stainless steel guidewire (0.23 mm diameter), a drive shaft (0.65 mm outer diameter (OD)), and a sheath (1.43 mm OD, 0.2 mm thick). The drive shaft connects the grinding wheel and air turbine and is made of three helically wound 0.18 mm diameter stainless steel coils. The shaft rotates inside a stationary Teflon sheath. The friction between the drive
A tissue phantom, shown in Fig. 4, was fabricated to simulate a lesion in the popliteal artery [11]. The phantom consisted of a ringshape plaque surrogate made from a bovine femoral bone, the vessel phantom (PVC with 45 kPa Young’s modulus), a rigid acrylic shell, and a muscle phantom (PVC with 8 kPa Young’s modulus) in between the vessel phantom and acrylic shell. This phantom was connected to a blood mimicking water source via a PVC tube (9.53 mm OD and 1.59 mm thickness). The RA catheter was inserted into the PVC tube to give the grinding wheel access to the plaque surrogate, as shown in Fig. 2. Water at 37 8C flowed through the PVC tube and the tissue phantom at 30 mL/min to simulate blood flow.
Fig. 4. Tissue phantom design, material, and dimensions (Unit: mm).
2.3. Measurement system
Fig. 3. Catheter and grinding wheel dimensions, catheter cross-section, and SEM image of the metal bond diamond wheel surface (Unit: mm).
As shown in Fig. 2, a high-speed camera (Model FASTCAM1024PCI by Photron) was used to record the grinding wheel motion through the semi-transparent portion of the phantom at a rate of
A.J. Shih et al. / CIRP Annals - Manufacturing Technology 65 (2016) 345–348
18,000 frame/s.Under this high frame rate, bright, consistent lighting conditions were necessary to obtain quality images and as such a fiber optic light source (Model 8375 by Fostec) delivered a concentrated light on the grinding wheel. A piezo-electric dynamometer (Model 9256-C by Kistler) was mounted under the tissue phantom block to measure the force in the vessel radial direction at 5000 Hz sampling frequency. Two thermocouples, denoted as TC1 and TC2, as shown in Figs. 2 and 3, were embedded in the bovine bone to record the temperature at 500 Hz sampling rate. As illustrated in Fig. 4, tips of TC1 and TC2 were 0.2 and 0.3 mm away from the inner surface of the plaque surrogate (bovine bone), respectively. 2.4. Experiment design In this experiment, the grinding wheel rotated at 160,000 rpm, as suggested by Barboto et al. [3]. The grinding wheel axially moved back and forth at 10 mm/s though the plaque surrogate three times. During grinding, the water running though the tissue phantom was collected for debris size analysis. Motion, force, and temperature were recorded. After the test, the plaque surrogate was cut in half for observation of ground surfaces using SEM. 3. Grinding wheel motion An orbital motion of the grinding wheel around the arterial wall was observed as illustrated in Fig. 5. To analyze the grinding wheel motion, an image processing technique based on pixel intensity was adopted. As shown in Fig. 5(a), pixels in a sample frame image from the video were divided into four regions based on intensity in MATLAB (R2014b by MathWorks). Pixels with the highest intensity, marked in red in Fig. 5(a), were the grinding wheel and drive shaft, due to the focused illumination. A pixel (in the black circle) at the center of the high intensity region on the grinding wheel was selected to indicate the grinding wheel position. The number of pixels from this point to the bottom edge of the image was counted in each frame to track the grinding wheel motion.
347
Fig. 5(b) shows the relationship between grinding wheel displacement over time. Five video images with 2.9 ms time step in the radial and axial directions are shown in Fig. 5(c) and (d), respectively. The grinding wheel orbits around the vessel at 5170 rpm, much lower than the 160,000 rpm rotational speed. The grinding wheel motion in RA is illustrated schematically in Fig. 5(e). The wheel rotational and orbital directions are the same. The wheel rotates about its axis (at 160,000 rpm) and orbits about the vessel axis (at 5170 rpm). To our knowledge, this observation of wheel motion in RA has not yet been reported in the literature. The grinding zone orbits around the vessel. The wheel orbiting motion enables a small diameter wheel grinding a large diameter vessel and enhances the heat dissipation from the grinding zone, which will be discussed in Section 6. Similar to a hydrodynamic bearing, we believe the wheel orbital motion is created by a fluid pressure film between the wheel and vessel. A computational fluid dynamics (CFD) simulation was conducted in ANSYS Fluent 15.0. The rotation of the grinding wheel induces a flow that lifts and orbits the wheel. The pressure distribution and streamline in five positions of an orbital period are shown in Fig. 5(f). The orbital direction matches the motion observed in Fig. 5(c) and (d). The orbital motion in RA has been largely ignored in clinical practice. A grinding wheel with a diameter larger than that of the lesion lumen was typically chosen to ‘‘drill’’ through the lesion, which may lead to excessive grinding forces and heat generation. This study suggests a grinding wheel much smaller than the lumen size is capable of orbiting and still grinding the plaque. 4. Grinding force Fig. 6(a) shows the measured force during RA. Six peaks indicate the elevated force during plaque grinding when the grinding wheel axially passed though the plaque surrogate. The average magnitude of the plaque grinding force, illustrated by the pink marker in Fig. 6(a), was 0.23 N. The coefficient of variation is small (3%) among the six peaks. The force magnitude in vessel phantom (orange line) was 0.07 N smaller than that in the bone, as shown in Fig. 6(a). The force in grinding the vessel phantom is mainly due to the elastohydrodynamic film generated between the grinding wheel and the soft PVC. A close-up view of 0.03 s of the plaque grinding force is shown in Fig. 6(b). A low pass filter with a 200 Hz cut-off frequency is
Fig. 6. (a) Force measurement, (b) plaque grinding force, and (c) frequency analysis.
applied. The filtered data, marked in red in Fig. 6(b), shows a sinusoidal wave. A frequency analysis via fast Fourier transform is shown in Fig. 6(c). A dominant frequency of 107.9 Hz matches the grinding wheel orbital frequency of the grinding wheel motion observed by the high-speed camera. 5. Material removal 5.1. Debris size Fig. 5. Grinding wheel motion: (a) image processing to track the wheel position, (b) wheel displacement and corresponding video frames in (c) radial- and (d) axialdirections, (e) schematic of the wheel motion in RA, and (f) CFD simulation results.
The bone debris mixed in the water was collected during RA and analyzed by a laser particle size analyser (Mastersizer S2000 by
A.J. Shih et al. / CIRP Annals - Manufacturing Technology 65 (2016) 345–348
348
Malvern). The average debris size distribution and variance for 10 measurements are shown in Fig. 7. This curve fits well with the Rayleigh distribution, a typical distribution of undeformed chip thickness found in grinding [12]. About 50% of debris is smaller than 10 mm and 90% of debris is smaller than 31 mm. This debris size is reasonable for RA considering human monocytes in the blood that range from 10 to 30 mm. Compared to previous studies, this result is larger than Hansen et al.’s measurement in a rabbit iliac model (98% smaller than 10 mm) [13] and Ahn et al.’s observation in cadaveric arteries (90% smaller than 20 mm) [14]. This may be due to using bovine bone as a surrogate for plaque.
Fig. 7. The debris size and distribution.
FðT w ; hÞ ¼
J I X X jT ij T ij FEM j i¼1 j¼1
where I is the number of thermocouples and J is the total time steps. Tij is the experimentally measured temperature Ti by TCi at time step j, and TijFEM is calculated by FEM with Tw and h. The Ti in Phases I and II are utilized independently to find Tw and h, denoted as Tw-I, Tw-II, hI, and hII. In Phase I, Tw-I = 41.1 8C and hI = 103.2 W/(m2 8C). In Phase II, Tw-II = 41.1 8C and hII = 111.0 W/ (m2 8C). Using these four parameters (Tw-I, Tw-II, hI, and hII) in FEM, the calculated temperatures at TC1 and TC2 in Phases I and II are shown in Fig. 9 and matched well with the experimentally measured results (T1 and T2). This analysis gives a 4.1 8C total temperature rise in the plaque surrogate. The inner surface temperature of the plaque surrogate raised from 37.0 8C (blood mimicking water temperature) to 41.1 8C in Phase I due to the saline flow heated by the friction within the catheter. In Phase II, the surface temperature is essentially the same with Phase I even the plaque been grinded. 7. Conclusions
5.2. Surface morphology Surface topography of the pre- and post-RA surfaces was examined by SEM and is presented in Fig. 8. Compared to the reamed pre-RA surface in Fig. 8(b), grooves due to grinding can be observed on the post-RA surface (Fig. 8(c)). The groove direction is perpendicular to the vessel axial direction (red arrow) and conforms to the grinding wheel motion. Some debris stick to the ground surface, marked by yellow arrows, can be observed. A single chip about 20 mm in size is shown in Fig. 8(d). The debris size generally matches well with the debris size distribution in Fig. 7.
This paper presents a comprehensive study of the RA grinding process with the following key findings: (1) the rotating grinding wheel orbited around the lumen due to hydrodynamics; (2) the bone (plaque surrogate) experienced an oscillating grinding force with the orbital frequency; (3) friction inside the catheter (particularly between the drive shaft and the guidewire) was the major heat source and raised the saline by over 4 8C; and (4) the ground surfaces had parallel grooves and showed the grinding action. Effects of grinding wheel rotational speed, advance speed, and wheel size on the RA grinding process will require further investigation. The wheel motion and material removal mechanism in RA are critical to future RA device development and surgical planning.
References
Fig. 8. (a) Specimen for SEM, (b) pre-RA surface, (c) post-RA surface, and (d) a single chip.
6. Temperature Temperatures measured by TC1 and TC2, denoted as T1 and T2, respectively, are shown in Fig. 9. The grinding and temperature measurement have three phases. In Phase I (0–16 s), the grinding wheel rotated at 160,000 rpm outside the plaque surrogate without axial motion. In Phase II (16–41 s), the rotating wheel moved in the axial direction and ground the plaque surrogate six times. In Phase III (after 41 s), the grinding wheel rotation stopped.
Fig. 9. Plaque surrogate temperature measurement and analysis.
The temperature of water, Tw, is assumed to be a constant. An inverse heat transfer method and finite element model (FEM) (in Abaqus V6.11) were applied to find Tw and the heat transfer coefficient, h, that fit best to the measurement in Phases I–II. The objective function to find the optimized Tw and h is
[1] Tomey MI, Kini AS, Sharma SK (2014) Current Status of Rotational Atherectomy. JACC Cardiovasular Interventions 7(4):345–353. [2] Vom Dahl J, Dietz U, Haager PK, Silber S, Niccoli L, Buettner HJ, et al (2002) Rotational Atherectomy Does Not Reduce Recurrent in-Stent Restenosis: Results of the Angioplasty Versus Rotational Atherectomy for Treatment of Diffuse in-Stent Restenosis Trial (ARTIST). Circulation 105:583–588. [3] Barbato E, Carrie´ D, Dardas P, Fajadet J, Gaul G, Haude M, et al (2015) European Expert Consensus on Rotational Atherectomy. EuroIntervention 11(1):30–36. [4] Safian RD, Feldman T, Muller DW, Mason D, Schreiber T, Haik B, Mooney M, O’Neill WW (2001) Coronary Angioplasty and Rotablator Atherectomy Trial (CARAT): Immediate and Late Results of a Prospective Multicenter Randomized Trial. Catheterization and Cardiovascular Interventions 53:213–220. [5] Kini AS, Marmur JD, Duvvuri S, Dangas G, Choudhary S, Sharma SK (1999) Rotational Atherectomy: Improved Procedural Outcome with Evolution of Technique and Equipment: Single-Center Results of First 1000 Patients. Catheterization and Cardiovascular Interventions 46(3):305–311. [6] Lin CP, Wang JH, Lee WL, Ku PM, Yin WH, Tsao TP, Chang CJ (2013) Mechanism and Management of Burr Entrapment: A Nightmare of Interventional Cardiologists. Journal of Geriatric Cardiology 10:230–234. [7] Kim MH, Kim HJ, Kim NN, Yoon HS, Ahn SH (2011) A Rotational Ablation Tool for Calcified Atherosclerotic Plaque Removal. Biomedical Microdevices 13(6): 963–971. [8] Nakao M, Tsuchiya K, Maeda W, Iijima D (2005) A Rotating Cutting Tool to Remove Hard Cemented Deposits in Heart Blood Vessels without Damaging Soft Vessel Walls. CIRP Annals – Manufacturing Technology 54(1):37–40. [9] Reisman M, Shuman BJ, Harms V (1998) Analysis of Heat Generation During Rotational Atherectomy Using Different Operational Techniques. Catheterization and Cardiovascular Diagnosis 44(4):453–455. [10] Reisman M, Shuman BJ, Dillar D, Fei R, Misser KH, Gordon LS, Harms V (1998) Analysis of Low-Speed Rotational Atherectomy for the Reduction of Platelet Aggregation. Catheterization and Cardiovascular Diagnosis 45(2):208–214. [11] Zheng Y, Belmont B, Shih A (2015) Experimental Investigation of the Grinding Wheel Dynamics in Atherectomy. Procedia Manufacturing 1:879–891. [12] Younis MA, Alawi H (1984) Probabilistic Analysis of the Surface Grinding Process. Transactions of the Canadian Society for Mechanical Engineering 8(4):208–213. [13] Hansen DD, Auth DC, Vracko R, Ritchie JL (1988) Rotational Atherectomy in Atherosclerotic Rabbit Iliac Arteries. American Heart Journal 115(1):160–165. [14] Ahn SS, Auth D, Marcus DR, Moore WS (1988) Removal of Focal Atheromatous Lesions by Angioscopically Guided High-Speed Rotary Atherectomy: Preliminary Experimental Observations. Journal of Vascular Surgery 7(2):292–300.
Contents lists available at ScienceDirect
CIRP Annals - Manufacturing Technology jou rnal homep age : ht t p: // ees .e lse vi er . com /ci r p/ def a ult . asp
Grinding wheel motion, force, temperature, and material removal in rotational atherectomy of calcified plaque Albert J. Shih (1)a,b, Yao Liu a, Yihao Zheng a,* a b
Mechanical Engineering, University of Michigan, Michigan, USA Biomedical Engineering, University of Michigan, Michigan, USA
A R T I C L E I N F O
A B S T R A C T
Keywords: Biomedical Grinding Mechanism
This study investigates the grinding wheel motion, force, material removal, and temperature in rotational atherectomy (RA). RA utilizes a metal-bond diamond wheel to remove plaque from arteries to treat cardiovascular diseases. As a plaque surrogate, a bone workpiece was placed in a vessel simulator and subjected to RA with a wheel rotational speed of 160,000 rpm. This grinding process was monitored by a high-speed camera, a dynamometer, and embedded thermocouples. The results show this process has a 108 Hz wheel orbital frequency, an oscillating grinding force of 0.23 N, 90% debris smaller than 31 mm, and a 4.1 8C temperature rise. ß 2016 CIRP.
1. Introduction Rotational atherectomy (RA), as shown in Fig. 1, is a grinding process to remove plaque from arterial walls to restore blood flow. RA utilizes a high-speed metal-bond diamond grinding wheel (1.25–2.5 mm diameter and up to 210,000 rpm [1]) for the treatment of several cardiovascular diseases, e.g. the calcification (hardening) of plaque, arterial bifurcation, ostial stenosis, and instent restenosis in which RA is preferred over traditional angioplasty [2]. The grinding wheel is driven by a long flexible drive shaft and rotates around a stationary guidewire (0.23 mm diameter). The shaft rotates within a sheath where saline flows (marked by arrows in Fig. 1) for lubrication and cooling. Blood flows outside the sheath. Clinically, RA has high complication rates [1]: restenosis – the regrowth of the plaque after RA – occurs in over 50% patients; and
Fig. 1. Rotational atherectomy – the plaque grinding process. * Corresponding author. E-mail address: [email protected] (Y. Zheng). http://dx.doi.org/10.1016/j.cirp.2016.04.012 0007-8506/ß 2016 CIRP.
other complications include myocardial infarction, dissection, perforation, slow-flow/no-reflow, vasospasm, and grinding wheel entrapment. Extensive clinical studies have been conducted to lower complication rates. A current universally accepted protocol suggests an initial grinding wheel size of 1.25–1.5 mm with a rotational speed ranging from 135,000 to 180,000 rpm [3]. Safian et al. [4] suggested avoiding excessive rotational speed decreases (>5000 rpm) as this indicates large grinding forces that may damage healthy tissue. Kini et al. [5] recommended a plaque grinding time between 20 and 30 s to prevent vessel dissection and thermal damage, with a grinding wheel diameter less than 70% of the treated artery diameter to avoid over-stretching the lesion. Lin et al. [6] found that advancing the grinding wheel across long, angulated, and heavily calcified lesions may cause a steep decrease of the rotational speed and entrap the wheel, suggesting a slow and steady advance of the wheel for less than 15 s. Engineering studies in RA have thus far been limited to investigation into grinding wheel design and rotational speed. For instance, Kim et al. [7] modified the grinding wheel surface by laser engraving to reduce microcavitation and tissue damage. Nakao et al. [8] created micro-blades on the wheel surface to replace diamond abrasives. Reisman et al. investigated the effects of rotational speed on platelet aggregation [9] and tissue thermal injury [10]. However, our review of the literature has found that there is a lack of knowledge in the grinding process and plaque removal mechanism in RA. Understanding the grinding wheel motion, force, debris size, ground surface, and temperature is critical to ultimately improving RA techniques and devices. Because of the high rotational speeds, the flexibility of the drive shaft, and poor visibility and accessibility of the grinding site within arteries, investigation of these grinding mechanisms is and has been challenging. To address these issues, this study investigated the RA grinding process within a semi-transparent soft tissue-mimicking phantom embedded with a
346
A.J. Shih et al. / CIRP Annals - Manufacturing Technology 65 (2016) 345–348
ring-shape bovine bone workpiece as a plaque surrogate. To measure the grinding wheel motion, force, and temperatures during the grinding process, a high-speed camera, piezoelectric dynamometer, and embedded thermocouples were utilized, respectively. RA device and experimental setup and design are first introduced. Results on grinding wheel motion, force, debris size, surface topography, and temperatures of the blood-mimicking fluid and plaque surrogate are presented and discussed. 2. Experimental setup The experimental setup, as shown in Fig. 2, consists of three modules: (1) a rotational atherectomy device, (2) tissue phantom, and (3) measurement system.
shaft, the guidewire, and the sheath is reduced by flowing saline inside the sheath (shown in Fig. 1). The grinding wheel, as shown in Fig. 3, is an ellipsoid with the distal half coated with diamond abrasives with an average size of 10 mm, as shown in the scanning electron microscope (SEM) image. A 2.5 mm diameter grinding wheel was used in this study. The air turbine drives the grinding wheel from 6000 to 210,000 rpm. An advance knob, as shown in Fig. 2, enables the manual movement of the grinding wheel in the axial direction during the procedure. In this experiment, the knob was driven by a linear stage to control the wheel axial motion. An air pressure regulator sets the rotational speed of the grinding wheel.
Fig. 2. Experimental setup.
2.1. Rotational atherectomy device
2.2. Tissue phantom
The RA device, RotablatorTM by Boston Scientific, includes a catheter, an air turbine, a grinding wheel advancing unit, and an air pressure regulator with an information display panel. The catheter, as shown in Fig. 3, is inserted into the artery during RA. The stationary guidewire extends from the air turbine through the drive shaft and grinding wheel to beyond the plaque (Fig. 1). The guidewire guides the rotation and translation of the grinding wheel and drive shaft. As illustrated in the cross-section A-A in Fig. 3, the catheter consists of a stainless steel guidewire (0.23 mm diameter), a drive shaft (0.65 mm outer diameter (OD)), and a sheath (1.43 mm OD, 0.2 mm thick). The drive shaft connects the grinding wheel and air turbine and is made of three helically wound 0.18 mm diameter stainless steel coils. The shaft rotates inside a stationary Teflon sheath. The friction between the drive
A tissue phantom, shown in Fig. 4, was fabricated to simulate a lesion in the popliteal artery [11]. The phantom consisted of a ringshape plaque surrogate made from a bovine femoral bone, the vessel phantom (PVC with 45 kPa Young’s modulus), a rigid acrylic shell, and a muscle phantom (PVC with 8 kPa Young’s modulus) in between the vessel phantom and acrylic shell. This phantom was connected to a blood mimicking water source via a PVC tube (9.53 mm OD and 1.59 mm thickness). The RA catheter was inserted into the PVC tube to give the grinding wheel access to the plaque surrogate, as shown in Fig. 2. Water at 37 8C flowed through the PVC tube and the tissue phantom at 30 mL/min to simulate blood flow.
Fig. 4. Tissue phantom design, material, and dimensions (Unit: mm).
2.3. Measurement system
Fig. 3. Catheter and grinding wheel dimensions, catheter cross-section, and SEM image of the metal bond diamond wheel surface (Unit: mm).
As shown in Fig. 2, a high-speed camera (Model FASTCAM1024PCI by Photron) was used to record the grinding wheel motion through the semi-transparent portion of the phantom at a rate of
A.J. Shih et al. / CIRP Annals - Manufacturing Technology 65 (2016) 345–348
18,000 frame/s.Under this high frame rate, bright, consistent lighting conditions were necessary to obtain quality images and as such a fiber optic light source (Model 8375 by Fostec) delivered a concentrated light on the grinding wheel. A piezo-electric dynamometer (Model 9256-C by Kistler) was mounted under the tissue phantom block to measure the force in the vessel radial direction at 5000 Hz sampling frequency. Two thermocouples, denoted as TC1 and TC2, as shown in Figs. 2 and 3, were embedded in the bovine bone to record the temperature at 500 Hz sampling rate. As illustrated in Fig. 4, tips of TC1 and TC2 were 0.2 and 0.3 mm away from the inner surface of the plaque surrogate (bovine bone), respectively. 2.4. Experiment design In this experiment, the grinding wheel rotated at 160,000 rpm, as suggested by Barboto et al. [3]. The grinding wheel axially moved back and forth at 10 mm/s though the plaque surrogate three times. During grinding, the water running though the tissue phantom was collected for debris size analysis. Motion, force, and temperature were recorded. After the test, the plaque surrogate was cut in half for observation of ground surfaces using SEM. 3. Grinding wheel motion An orbital motion of the grinding wheel around the arterial wall was observed as illustrated in Fig. 5. To analyze the grinding wheel motion, an image processing technique based on pixel intensity was adopted. As shown in Fig. 5(a), pixels in a sample frame image from the video were divided into four regions based on intensity in MATLAB (R2014b by MathWorks). Pixels with the highest intensity, marked in red in Fig. 5(a), were the grinding wheel and drive shaft, due to the focused illumination. A pixel (in the black circle) at the center of the high intensity region on the grinding wheel was selected to indicate the grinding wheel position. The number of pixels from this point to the bottom edge of the image was counted in each frame to track the grinding wheel motion.
347
Fig. 5(b) shows the relationship between grinding wheel displacement over time. Five video images with 2.9 ms time step in the radial and axial directions are shown in Fig. 5(c) and (d), respectively. The grinding wheel orbits around the vessel at 5170 rpm, much lower than the 160,000 rpm rotational speed. The grinding wheel motion in RA is illustrated schematically in Fig. 5(e). The wheel rotational and orbital directions are the same. The wheel rotates about its axis (at 160,000 rpm) and orbits about the vessel axis (at 5170 rpm). To our knowledge, this observation of wheel motion in RA has not yet been reported in the literature. The grinding zone orbits around the vessel. The wheel orbiting motion enables a small diameter wheel grinding a large diameter vessel and enhances the heat dissipation from the grinding zone, which will be discussed in Section 6. Similar to a hydrodynamic bearing, we believe the wheel orbital motion is created by a fluid pressure film between the wheel and vessel. A computational fluid dynamics (CFD) simulation was conducted in ANSYS Fluent 15.0. The rotation of the grinding wheel induces a flow that lifts and orbits the wheel. The pressure distribution and streamline in five positions of an orbital period are shown in Fig. 5(f). The orbital direction matches the motion observed in Fig. 5(c) and (d). The orbital motion in RA has been largely ignored in clinical practice. A grinding wheel with a diameter larger than that of the lesion lumen was typically chosen to ‘‘drill’’ through the lesion, which may lead to excessive grinding forces and heat generation. This study suggests a grinding wheel much smaller than the lumen size is capable of orbiting and still grinding the plaque. 4. Grinding force Fig. 6(a) shows the measured force during RA. Six peaks indicate the elevated force during plaque grinding when the grinding wheel axially passed though the plaque surrogate. The average magnitude of the plaque grinding force, illustrated by the pink marker in Fig. 6(a), was 0.23 N. The coefficient of variation is small (3%) among the six peaks. The force magnitude in vessel phantom (orange line) was 0.07 N smaller than that in the bone, as shown in Fig. 6(a). The force in grinding the vessel phantom is mainly due to the elastohydrodynamic film generated between the grinding wheel and the soft PVC. A close-up view of 0.03 s of the plaque grinding force is shown in Fig. 6(b). A low pass filter with a 200 Hz cut-off frequency is
Fig. 6. (a) Force measurement, (b) plaque grinding force, and (c) frequency analysis.
applied. The filtered data, marked in red in Fig. 6(b), shows a sinusoidal wave. A frequency analysis via fast Fourier transform is shown in Fig. 6(c). A dominant frequency of 107.9 Hz matches the grinding wheel orbital frequency of the grinding wheel motion observed by the high-speed camera. 5. Material removal 5.1. Debris size Fig. 5. Grinding wheel motion: (a) image processing to track the wheel position, (b) wheel displacement and corresponding video frames in (c) radial- and (d) axialdirections, (e) schematic of the wheel motion in RA, and (f) CFD simulation results.
The bone debris mixed in the water was collected during RA and analyzed by a laser particle size analyser (Mastersizer S2000 by
A.J. Shih et al. / CIRP Annals - Manufacturing Technology 65 (2016) 345–348
348
Malvern). The average debris size distribution and variance for 10 measurements are shown in Fig. 7. This curve fits well with the Rayleigh distribution, a typical distribution of undeformed chip thickness found in grinding [12]. About 50% of debris is smaller than 10 mm and 90% of debris is smaller than 31 mm. This debris size is reasonable for RA considering human monocytes in the blood that range from 10 to 30 mm. Compared to previous studies, this result is larger than Hansen et al.’s measurement in a rabbit iliac model (98% smaller than 10 mm) [13] and Ahn et al.’s observation in cadaveric arteries (90% smaller than 20 mm) [14]. This may be due to using bovine bone as a surrogate for plaque.
Fig. 7. The debris size and distribution.
FðT w ; hÞ ¼
J I X X jT ij T ij FEM j i¼1 j¼1
where I is the number of thermocouples and J is the total time steps. Tij is the experimentally measured temperature Ti by TCi at time step j, and TijFEM is calculated by FEM with Tw and h. The Ti in Phases I and II are utilized independently to find Tw and h, denoted as Tw-I, Tw-II, hI, and hII. In Phase I, Tw-I = 41.1 8C and hI = 103.2 W/(m2 8C). In Phase II, Tw-II = 41.1 8C and hII = 111.0 W/ (m2 8C). Using these four parameters (Tw-I, Tw-II, hI, and hII) in FEM, the calculated temperatures at TC1 and TC2 in Phases I and II are shown in Fig. 9 and matched well with the experimentally measured results (T1 and T2). This analysis gives a 4.1 8C total temperature rise in the plaque surrogate. The inner surface temperature of the plaque surrogate raised from 37.0 8C (blood mimicking water temperature) to 41.1 8C in Phase I due to the saline flow heated by the friction within the catheter. In Phase II, the surface temperature is essentially the same with Phase I even the plaque been grinded. 7. Conclusions
5.2. Surface morphology Surface topography of the pre- and post-RA surfaces was examined by SEM and is presented in Fig. 8. Compared to the reamed pre-RA surface in Fig. 8(b), grooves due to grinding can be observed on the post-RA surface (Fig. 8(c)). The groove direction is perpendicular to the vessel axial direction (red arrow) and conforms to the grinding wheel motion. Some debris stick to the ground surface, marked by yellow arrows, can be observed. A single chip about 20 mm in size is shown in Fig. 8(d). The debris size generally matches well with the debris size distribution in Fig. 7.
This paper presents a comprehensive study of the RA grinding process with the following key findings: (1) the rotating grinding wheel orbited around the lumen due to hydrodynamics; (2) the bone (plaque surrogate) experienced an oscillating grinding force with the orbital frequency; (3) friction inside the catheter (particularly between the drive shaft and the guidewire) was the major heat source and raised the saline by over 4 8C; and (4) the ground surfaces had parallel grooves and showed the grinding action. Effects of grinding wheel rotational speed, advance speed, and wheel size on the RA grinding process will require further investigation. The wheel motion and material removal mechanism in RA are critical to future RA device development and surgical planning.
References
Fig. 8. (a) Specimen for SEM, (b) pre-RA surface, (c) post-RA surface, and (d) a single chip.
6. Temperature Temperatures measured by TC1 and TC2, denoted as T1 and T2, respectively, are shown in Fig. 9. The grinding and temperature measurement have three phases. In Phase I (0–16 s), the grinding wheel rotated at 160,000 rpm outside the plaque surrogate without axial motion. In Phase II (16–41 s), the rotating wheel moved in the axial direction and ground the plaque surrogate six times. In Phase III (after 41 s), the grinding wheel rotation stopped.
Fig. 9. Plaque surrogate temperature measurement and analysis.
The temperature of water, Tw, is assumed to be a constant. An inverse heat transfer method and finite element model (FEM) (in Abaqus V6.11) were applied to find Tw and the heat transfer coefficient, h, that fit best to the measurement in Phases I–II. The objective function to find the optimized Tw and h is
[1] Tomey MI, Kini AS, Sharma SK (2014) Current Status of Rotational Atherectomy. JACC Cardiovasular Interventions 7(4):345–353. [2] Vom Dahl J, Dietz U, Haager PK, Silber S, Niccoli L, Buettner HJ, et al (2002) Rotational Atherectomy Does Not Reduce Recurrent in-Stent Restenosis: Results of the Angioplasty Versus Rotational Atherectomy for Treatment of Diffuse in-Stent Restenosis Trial (ARTIST). Circulation 105:583–588. [3] Barbato E, Carrie´ D, Dardas P, Fajadet J, Gaul G, Haude M, et al (2015) European Expert Consensus on Rotational Atherectomy. EuroIntervention 11(1):30–36. [4] Safian RD, Feldman T, Muller DW, Mason D, Schreiber T, Haik B, Mooney M, O’Neill WW (2001) Coronary Angioplasty and Rotablator Atherectomy Trial (CARAT): Immediate and Late Results of a Prospective Multicenter Randomized Trial. Catheterization and Cardiovascular Interventions 53:213–220. [5] Kini AS, Marmur JD, Duvvuri S, Dangas G, Choudhary S, Sharma SK (1999) Rotational Atherectomy: Improved Procedural Outcome with Evolution of Technique and Equipment: Single-Center Results of First 1000 Patients. Catheterization and Cardiovascular Interventions 46(3):305–311. [6] Lin CP, Wang JH, Lee WL, Ku PM, Yin WH, Tsao TP, Chang CJ (2013) Mechanism and Management of Burr Entrapment: A Nightmare of Interventional Cardiologists. Journal of Geriatric Cardiology 10:230–234. [7] Kim MH, Kim HJ, Kim NN, Yoon HS, Ahn SH (2011) A Rotational Ablation Tool for Calcified Atherosclerotic Plaque Removal. Biomedical Microdevices 13(6): 963–971. [8] Nakao M, Tsuchiya K, Maeda W, Iijima D (2005) A Rotating Cutting Tool to Remove Hard Cemented Deposits in Heart Blood Vessels without Damaging Soft Vessel Walls. CIRP Annals – Manufacturing Technology 54(1):37–40. [9] Reisman M, Shuman BJ, Harms V (1998) Analysis of Heat Generation During Rotational Atherectomy Using Different Operational Techniques. Catheterization and Cardiovascular Diagnosis 44(4):453–455. [10] Reisman M, Shuman BJ, Dillar D, Fei R, Misser KH, Gordon LS, Harms V (1998) Analysis of Low-Speed Rotational Atherectomy for the Reduction of Platelet Aggregation. Catheterization and Cardiovascular Diagnosis 45(2):208–214. [11] Zheng Y, Belmont B, Shih A (2015) Experimental Investigation of the Grinding Wheel Dynamics in Atherectomy. Procedia Manufacturing 1:879–891. [12] Younis MA, Alawi H (1984) Probabilistic Analysis of the Surface Grinding Process. Transactions of the Canadian Society for Mechanical Engineering 8(4):208–213. [13] Hansen DD, Auth DC, Vracko R, Ritchie JL (1988) Rotational Atherectomy in Atherosclerotic Rabbit Iliac Arteries. American Heart Journal 115(1):160–165. [14] Ahn SS, Auth D, Marcus DR, Moore WS (1988) Removal of Focal Atheromatous Lesions by Angioscopically Guided High-Speed Rotary Atherectomy: Preliminary Experimental Observations. Journal of Vascular Surgery 7(2):292–300.