Microscopy is most commonly defined as any technique for producing visible images of structures that are otherwise too small to be seen by the human eye. However, microscopes do more than image; they are a family of instruments that provide information on the spatial relationships of different aspects of composition and structure.
Besides imaging structure, microscopy can also provide chemical, mechanical, physical, molecular, biomolecular, cellular and other features. With medical devices, the universal value of microscopy becomes apparent, since success depends on interactions with living systems operating at the scale of proteins, cells, fibrotic capsules, capillary networks and related structures.
At this scale, biology interacts with the roughness of surface finishes, the integrity of organic coatings, polymer microdomains, metal grains and phase boundaries, nanoscale encapsulated drugs and micro-particulates generated from the wear of orthopaedic bearings. Thus, a medical device’s success or failure fundamentally occurs at microscopic dimensional scales.
In medical device development microscopy is applied to R&D, quality assurance and control, and regulatory approval documentation. Microscopy is the principal tool to ‘sell’ a device. This data should then be presented to physicians who would use the device, the reimbursement and purchasing gatekeepers who may choose to buy it, and perhaps even to the patients themselves.
Early stage firms will also need images to sell the concept or idea and obtain the necessary funds for development and marketing. The safety and effectiveness of the technology must also be sold to regulating bodies and reimbursement authorities so it can be approved and paid for.
Finally, the data must convince judges, juries and the press that the technology works. Poorly conducted microscopy-based analyses, or an obvious conflict of interest (such as analyses performed by the internal development team), can sink a device in court and greatly harm the company.
Why does selling technology require microscopy, more so than other analytical tools? Simply put, images are universally understood. We are hard-wired for image processing. In contrast, analysing quantitative data is a learned skill: try explaining XPS or FTIR spectroscopic data to a lay audience. But show them a scanning electron microscope (SEM) image and with explanation it will be understood.
Physicians, regulatory reviewers, investors, judges, juries, the lay public, as well as scientists and engineers can all interpret pictures. Of course, the proper type of microscopy with the correct preparation of the specimen is required for the image to be useful for any purpose.
The world of microscopy techniques, including instrumentation and sample preparation, is extremely broad and is growing rapidly. Choosing the most appropriate instrumentation and methodology can be overwhelmingly complex. It is easy to misinterpret a microscope image since we so readily interpret pictures, but we forget that microscope imaging physics is not the same as how our eyes and brain process light.
Microscopy is so much more complicated and diverse than it was a few decades ago that it is common to refer to a ‘microscopy renaissance’. Even in the 1970s there were only a few types of light microscopy in common use (brightfield, darkfield, fluorescence, phase contrast), plus transmission electron microscopy (TEM) and SEM.
There are now dozens of types of light microscopy, including multiple types of confocal, plus multiphoton, second harmonic, near field, optical coherence and Raman. There are multiple modalities known by their acronyms (for example DIC, Irm, IR, FLIM, AIC, FRET), several varieties of SEM, such as low voltage, variable pressure, and dual beam with focused ion beam, newer TEM types, such as energy filtration, and varieties of scanning probe microscopies, such as atomic force (AFM).
With the growth of imaging modalities and methods, where should one begin in determining the instrument to be used? There are several questions to consider in the formulation of a medical device microscopy plan. What size scale or scales are the structures to be examined – nanometres, microns or millimetres? Is the question compositional, structural, biological or a combination? Is there a chronological aspect, and if so, what is the time frame?
What types of materials: polymers, hydrogels, biologicals, metals, ceramics or organics? Will functional assessment be determined from the device during clinical use in vivo, with experimental animals in vivo, as explants ex vivo, or in vitro analysis? Is the question related to a living structure or an inanimate device?
Is complicated specimen preparation impossible, or can the specimen be processed for best preservation and imaging? Are we looking for a specific component or structure in relation to other structures (synthetic or biological), such as specific bioactive molecules? Can features of interest be chemically or biochemically labelled?
There are three instrument types that will address most medical device analyses. These are optical or light microscopy (LM), imaging spectroscopy (IS) and SEM. 3D information is an oftenoverlooked aspect of microscopical analysis since we are so accustomed to viewing 2D representations.
However, it is critical to remember that medical devices and their interactions with biology occur in 3D: while it is commonly stated that the surface of a device is where the interaction occurs, the surface may be very complicated and no device surface is truly flat at some scale. Finally, the effect of a device always extends into the tissue.
SEM has the broadest range of resolution and field of view of any microscope modality with the capability to ‘get the big picture’ and then to zoom in to (near) molecular details. SEM images are compelling since they uniquely provide a pleasing photo-like appearance.
While the appearance of depth is compelling, it is not possible to measure or even estimate the magnitude of depth within these images. Depth is measured with the SEM using stereo imaging. SEM can also be used to image atomic density and determine the position of certain polymer structures.
It is useful to compare SEM with AFM, as both provide image topography. Due to the differences in imaging physics, an AFM and SEM image of, for example, a commercial pyrolytic carbon heart valve prosthetic will appear dramatically different. SEM images surfaces by scanning with a highly focused electron beam and then measuring the emission of secondary (lower energy) electrons that are emitted from a ‘small’ near-surface volume of the specimen.
AFM ‘feels’ the structure by tracing the surface with a very sharp tip at the apex of an inverted pyramid. This gives us an important caveat to microscopical analysis: understanding how the image is generated is critical to understanding its meaning.
While the AFM tip that scans the specimen may be atomically sharp, the supporting inverted pyramid interacts with the sides of specimen features and thus prevents the tip from reaching sufficiently deep into the ~1um deep roughness of the valve material; thus it is unable to provide an accurate high resolution picture of the structure.
SEM analysis of biological materials requires careful specimen preparation, since SEM imaging occurs in a vacuum. It is also important to consider that many routine methods of biological specimen preparation can alter the structure of certain polymers used in medical devices.
MAKING THE RIGHT CHOICE
LM can address many questions of medical device structure and even functional activity and/or associated biological systems. Perhaps the areas of LM that are least known to engineers in medical device development are its capabilities for live biological imaging, and the variety of specific labels and probes that can determine the location and/or activity of cells, cellular structures, enzymes, other biological and some non-biological molecules and structures.
The development of these fluorescent and particulate labels (typically gold nanoparticles and more recently semiconductor quantum dots) is a core aspect of the post-1970s renaissance in microscopy. The fluorescent labels are used in light microscopy, while the nano-particulate labels also enable detectability with SEM and TEM, and allow imaging of the same specimen with LM and SEM.
One of the most common problems associated with medical devices is infection. LM images can provide colour-based coding, enabling computer-based image analysis of the ‘zone of killing’ of an antibacterial coating on a device such as a catheter.
Imaging spectroscopy (IS) can provide the same rapid analysis as light microscopy where no stain or label is available, or where the structure of interest has a specific molecular construction that can be detected with infrared (IR) or Raman spectroscopy. IS is essentially light microscopy with the addition of spectroscopic analysis to the thousands of individual pixels that form the image.
IS comes in many types, including IR and Raman, and may be coupled with different types of light microscopy including widefield confocal to enable 3D or depth imaging, and near-field.
Scanning Near-field Optical Microscopy (SNOM) uses an AFM-like probe with a submicron ~50nm aperture that rastors over the specimen to obtain spectroscopic data with much higher spatial resolution. These three major classes of microscope instruments address the majority of medical device imaging questions. The table (right) provides a general guide to choosing the most suitable microscope within these three classes.
These are general guidelines for most biomaterial and medical device questions. However, every problem is unique and complicated, since most medical devices include multiple types of synthetic and biological materials, each requiring different types of specimen preparation and analyses. This can be problematic since few engineers possess the necessary breadth of knowledge of methods.
Many investigations benefit from using ‘correlative microscopy’ wherein multiple microscope types are used to image different structural or functional aspects, either by examining multiple samples of the same type of specimen with different instruments or by examining the same sample sequentially.
Microscopy is a central instrument for R&D, quality control, regulatory submission and legal defence. No other family of instrumentation provides spatial relationship information, and such images are accessible for all concerned with the development of medical devices, including their use, regulation and, potentially, litigation.