Characterizing tissue stiffness at the tip of a rigid needle using an opto-mechanical force sensor
- First Online:
- Cite this article as:
- Beekmans, S.V. & Iannuzzi, D. Biomed Microdevices (2016) 18: 15. doi:10.1007/s10544-016-0039-1
- 808 Downloads
We present a novel device that allows the user to measure the Young Modulus of a material at the opening of a 5 mm diameter needle. The device relies on a miniaturized cantilever spring mounted at the end of the needle and interrogated via Fabry-Pérot optical fiber interferometry. The probe is repetitively brought in and out of contact with the sample at the end of the needle by means of a steel cable that is controlled via a piezoelectric actuator located at the proximal end. We demonstrate the ability of our device to detect and quantify layers of varying stiffness during needle insertion in a gelatin phantom and to successfully locate tissue boundaries in bovine liver tissue embedded in gelatin.
KeywordsFerrule-top technology In situ indentation Remote actuation Interferometry Tissue Stiffness Micromechanics Minimally invasive instrument
The mechanical properties of biological networks are often overseen in functional research of healthy tissue as well as in the diagnosis of potentially diseased tissue and in treatment monitoring. For example, in the classification of skin conditions (such as scars or burn wounds) and the following control of disease progression or healing, physicians in most cases prefer visual assessment and subjective scaling above quantitative mechanical information (Draaijers et al. 2004; Durani et al. 2009; Huang et al. 2004; Gurtner et al. 2008). Moreover, tissue mechanics can be linked to a wide array of physiological processes (Cowin and Doty 2007; Cowin and Humphrey 2007; Butler et al. 2000). Cells have very sophisticated methods to sense and adapt to their mechanical environment. Stem cells, for instance, have been proven to adapt their differentiation to the stiffness of their extracellular matrix (Swift et al. 2013; Fu et al. 2010) and white blood cells, as well as tumor cells, are able to manipulate their own stiffness and shape to migrate in and out of blood vessels (Friedl and Wolf K. 2003). At the tissue level, the interplay between the cellular mechanics and the extracellular network determines the stiffness at the micron scale, changes of which have been associated with Alzheimer’s disease (Murphy et al. 2011) and multiple sclerosis (Wuerfel et al. 2010; Streitberger et al. 2012) (in the brain), cancerous growth (Plodinec et al. 2012; Li et al. 2008) (breast) and osteoarthritis (Stolz et al. 2009; Desrochers et al. 2010) (cartilage). Hence, a method to quantify local mechanical properties of tissue, preferably in situ, is of high interest.
Classically, the biomechanical response of complex networks is assessed by means of Atomic Force Microscope (AFM) nanoindentation (Hengsberger et al. 2001; Franze 2011; Zhu et al. 2011; Mathur et al. 2001; Li et al. 2015; Gautier et al. 2015). The utilization of an AFM for classification of biological tissues has, however, some principle limitations that cannot be easily overcome, such as size, stability and flexibility. To mitigate those limitations, we have recently introduced a new probe, called ferrule-top cantilever, that provides a good alternative for indentation of biological samples in harsh environments (Chavan et al. 2012; Kahn et al. 2015; Neufurth et al. 2015).
Both AFM and ferrule-top indentation are restricted to probing the surface of a sample, while the advantages of probing in depth (i.e., underneath the surface) would clearly be multiple. By integrating an indenter at the tip of a needle one could not only quantify between layers of different stiffness, but also navigate to the target location and perform a minimally invasive measurement based on the tissue mechanical properties. An example of an in situ AFM indenter for arthroscopic knee cartilage inspection was presented by Imer et al. (2006). The indenter consists of an extensive stabilization stage connected to a piezoelectric scanning module, both of which are inserted into the sample, resulting in a large footprint. Moreover, the lack of calibration of the piezoelectric tube hampered a quantitative analysis.
Here, we demonstrate a ferrule-top indenter on the distal end of a rigid needle and we show its ability to quantify local mechanical properties of tissue in situ. Thanks to the remote actuation of the sensor by a piezoelectric translator the size of the needle is limited to the dimensions of the indentation probe at the tip. The performance of our indenter is tested on an engineered layered sample as well as on biological tissue.
2 Experimental section
2.1 The ferrule-top force transducer
2.2 Indentation module
To reduce the dimensions of the indenter, we have developed an indentation module that enables remote actuation of the force transducer. A schematic view of the indentation module is shown in Fig. 1b. The optical force transducer is housed in a square borosilicate capillary with an inner lumen of 3.05 mm x 3.05 mm, which restricts the movement of the probe to the axial direction. The probe is mechanically connected to a calibrated piezoelectric translator via a steel cable (diameter = 120 μm) similar to those that are commonly used to actuate the tip of surgical steerable needles (Breedveld et al. 2005; van de Berg et al. 2015). A small compression spring is used to load the probe against a backplane in the capillary. If the spring is initially compressed with pretension, in fact, the movement of the piezoelectric translator can be smoothly transferred, from remote position, to the probe. The translator (P-602.5L8, Physike Instrumente GmbH) has a 500 μm stroke and a 325 N blocking force. It is important to note that, due to hysteresis in the steel cable and friction between the capillary and the probe, in our case, it is not possible to assume that the movement of the probe is exactly equal to that indicated by the strain gauge feedback system of the driving piezoelectric device. To solve this issue, the movement of the probe is monitored by a second single mode optical fiber, anchored to the backplane of the square capillary and aligned with the back of the ferrule (see Fig. 1b).
2.3 Experimental setup
The indentation module is housed at the distal end of a 20cm long custom designed needle, that is used to insert the sensor into the specimen. Fig. 1c shows a schematic view of the ferrule-top indenter at the tip of the rigid needle. The needle (diameter = 5mm) is fixed to a motorized linear translation stage (LTS300, Thorlabs GmbH) that is used for insertion in the sample. At the proximal end of the needle the steel cable is fixed to the piezoelectric translator, which in turn is mounted on a coarse position stage, allowing for adjustment of the pretension in the cable-spring system. The two optical fibers (i.e., the cantilever deflection readout and the sensor movement readout), are fed through the lumen of the needle and connected to two interferometric readout systems (OP1550, Optics11)1. Both interferometers are equipped with a tunable infrared laser (35 nm tunability), the wavelength of which is internally locked with 10 pm accuracy by a feedback system, and can be swept by driving the injection current sinusoidally.
The sample is placed on an xy-translation stage (MAX312D, Thorlabs GmbH), which is used to select a position for needle insertion. To reduce vibrations, the setup is built on a passive anti-vibration stage.
2.4 Working principle
Wdc can then be recorded by means of a low-pass filter with a cut-off frequency below the modulation frequency. To record Wω, the unfiltered amplitude response of the photodiode is sent to a lock-in amplifier, which is locked at frequency ω via a square wave reference signal. We note that, thanks to the high bandwidth of our measurement, acquisition of Wdc and Wω and the following linearization of the signal is performed in real time.
2.5 Specimen preparation
To demonstrate the working principle of our device, we performed a series of insertions consisting of multiple indentation measurements at varying depth on two gelatin-based phantoms (Gelatin from bovine skin, Sigma-Aldrich). In the first specimen a stiffness gradient was created by compiling layers (10 mm in height) with decreasing mass gelatin to water ratio, hence creating a sample with layers of decreasing stiffness from bottom to top. Starting from a layer of 15 % mass gelatin at the bottom, each following layer contained 2.5 % less mass gelatin, ultimately creating a specimen with 6 layers of decreasing stiffness: 15 %, 12.5 %, 10 %, 7.5 %, 5 % and 2.5 % mass gelatin. For each layer, the gelatin was dissolved in demineralised water at 60 ° C, poured in the container and kept at 4 ° C for 30 minutes to allow the layer to stiffen. After the final layer was poured, the sample was stored 4 ° C overnight and measured the next day.
The second specimen consisted of a square piece of bovine liver (6x6x3 mm 3) fixated in a gelatin solution. To create a base layer (20 mm in height), gelatin was dissolved at a 12 % mass to water ratio at 60 °C, poured in a container and stored overnight at 4 °C. Subsequently, the liver sample was placed on top of the first gelatin layer and a second solution of gelatin (12%) was poured over the sample until it was fully submerged. The gelatin solution was cooled down to 40 °C prior to pouring to prevent thermal damage of the tissue. After the container was filled with gelatin (± 20 mm above the sample), it was again stored overnight at 4 °C to ensure proper stiffening.
2.6 Experimental details and indentation protocol
For this experiment, the cantilever was equipped with a spherical borosilicate bead of radius equal to 85 μm. For optimal sensitivity, the cleaved end of the single mode fiber was positioned directly underneath the center of the microbead, positioning the Fabry-Pérot cavity directly underneath the point of contact (Fig. 1a). A central wavelength of 1551 nm was used in both interferometers in combination with a 90 kHz sinusoidal wavelength modulation of approximately 50-200 pm modulation depth. The spring constant of the cantilever was measured to be 12.97 ± 0.06 N/m via the method reported in Beekmans and Iannuzzi (2015).
To validate our measurements, before the first insertion of the needle, we cut a cross-sectional slice of the specimen, in which the different layers were clearly identifiable. The slice was then indented on the surface at locations corresponding to the different layers. Each reference measurement was obtained as an average of 15 indentations spread over 3 locations.
3 Results and discussions
In the following section we discuss some of the limitations of our indenter in its current form.
The design of the cantilever for this experiment was optimized to measure the stiffness of materials with Young Modulus between 1 kPa and 100 kPa. To cover the wide range of biological tissue stiffness, which varies over at least three orders of magnitude (McKee et al. 2011), one may need to use cantilevers with different spring constants – a major complication for future applications. Furthermore, throughout the entire analysis, we have implicitly neglected the viscous and plastic components of the sample mechanics. Both these issues have been recently addressed by another paper of our group (van Hoorn et al. 2016), where we showed that, applying a dynamic modulation analysis, it is possible to measure both the loss and storage modulus of a largely heterogenous material. Still, the method adds some limitations, as it is not as fast and as straightforward as the one presented here.
A further obvious limitation of our needle indenter here is its large diameter. Work is under way to reduce the dimensions of all its components to embed the device in a 3 mm diameter needle.
Another possible drawback is the risk that, during perforation, some debris of the sample enters the Fabry-Pérot cavity between the fiber and the cantilever. Although we have not observed any nuisance when indenting inside the liver specimen, one could circumvent this problem altogether by designing a membrane based sensor.
Finally, for further research on biological samples, sterilization of the device may become necessary. We designed the device such that the sensitive part, which will not survive a repetitive sterilization procedure, can be disposed without discarding the main working elements of the indenter.
We have successfully developed a cantilever based, all-optical indenter at the tip of a rigid needle. The indentation measurement is enabled by a sensor that probes the mechanical properties of the underlying specimen by indentation using a microsphere. The sensor is remotely actuated by a strain gauge controlled piezoelectric translator driving a microscopic cable and spring system. The movement of the sensor as well as the movement of the cantilever is interrogated by Fabry-Pérot interferometry. We have performed stiffness measurements at fixed depth positions during needle insertion in gelatin phantoms and animal liver specimens. The measurements showed that we are able to quantify a stiffness gradient in depth and that we can successfully identify stiffer layers in a uniform sample. Measurements in a gelatin embedded animal liver confirmed that we can localize the liver based on mechanical contrast. Moreover, as the needle protrudes further inside the liver tissue, a quantitative analysis of the liver can be made based on the mechanical properties alone. Despite some limitations, our needle may, on the long term, ultimately evolve into a minimally invasive tool for the analysis of the mechanical properties of tissues, with potential applications in needle navigation or tissue diagnostics.
The authors would like to thank T. de Jong and J.J. van den Dobbelsteen of the Delft Institute of Technology and J. Scheltes of DEAM corporation for their input and fruitful discussions. The research leading to these results is supported by the Dutch Technology Foundation (STW) under the iMIT program (P11-13) and has received funding from LASERLAB-EUROPE under the EC’s Seventh Framework Programme (grant agreement no. 284464) and the European Research Counsil (615170).
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.