Advanced technologies for cost-effective endomicroscopy

Endomicroscopy enables real-time optical biopsies to be performed at the cellular level by using miniature optical systems integrated into the tip of a small imaging probe. The resulting high-resolution (0.5–3.0μm) images can help clinicians diagnose cancer and detect the margins of tumors, and the technique has potential in areas such as targeted therapy and the development of genetic drugs.

In recent years, endomicroscopy has used various imaging modalities, including wide-field, confocal, and two-photon microscopy. But regardless of the underlying imaging principle, one of the most crucial parts of the endomicroscope is a high-performance miniature microscope objective.

This discussion focuses on technology for building cost-effective miniature objectives. Two of the most important aspects of optics design for the endomicroscope include a small probe size to access confined locations in the body, and a camera with a large field of view (FOV) that allows screening a wide area of tissue, which also simplifies correlation between high resolution images and the macroscopic region of tissue in question, as well as comparison with standard histopathology images. Also key is a high numerical aperture (NA) to provide fine spatial resolution and maximum light collection. Finally, low cost is essential to allow for use in routine disease-screening procedures.

Unfortunately, small dimensions and low cost are difficult to achieve with large FOV and a high NA, which are in principle inversely proportional to each other for fixed system dimensions. The overall objective design has to be a compromise between the costs and benefits of each, while fabrication and assembly methods have to adhere to strict design specifications.

Recently demonstrated objectives^1–7 usually provide a FOV of between 250–500μm and a NA of 0.3–1.0, with an outer diameter of 2–10mm. These systems consist of several lenses and demand fine precision during fabrication and assembly. For example, lens centration tolerances can be as tight as ±5μm. Researchers usually do not publish the cost of their systems, but the fabrication price tag for a prototype objective made with miniature glass lenses and traditional assembly methods can be $5000–25,000. However, if an objective system reaches production, these figures can be reduced by a factor of 5–20.

To further lower costs it is necessary to exploit new technologies in fabrication and assembly, such as plastic-injection molding,³ grayscale lithography,⁸ deep x-ray lithography and electrodeposition,⁴ as well as LIGA (an acronym from the German ‘lithographie, galvanoformung und abformung’). Techniques such as injection molding and grayscale lithography enable the construction of optical components with arbitrary shapes, and allow direct incorporation of assembly features into the optical component. Consequently, it is possible to reduce the number of individual lens elements while simultaneously increasing accuracy.

Figure 1. A: diagram of a lens that incorporates features of a kinematic mount; B: diagram of a self-centering mount; C: self-entering mount fabricated for a miniature multi-modal microscope. (Image courtesy of Todd Christenson, HT MicroAnalytical Inc.)

High-throughput fabrication techniques can reduce the cost of individual high-precision components.^{3, 8} The most significant price contribution is then system assembly. One solution to this problem is the zero alignment concept,⁹ which relies on an opto-electromechanical design to produce a unit that can be assembled in a plug-and-snap fashion, requiring no further adjustments because system tolerances are accommodated within the fabrication and assembly steps. Examples of this assembly principle include kinematic mounts or self-centering springs incorporated into optical and opto-mechanical components. Conceptual drawings of a kinematic mount, self-centering mount, and a fabricated self-centering unit⁴ are presented in Figure 1. The kinematic approach is based on embedding alignment features onto two neighboring assembled parts. The self-centering principle relies on the geometry of parts and physical forces to move each lens element into place.

Examples of endomicroscope objectives built using this design philosophy are presented in Figure 2. Part A (left) shows two NA=1.0, FOV=250μm microscope objectives. The objective on the left (outer diameter: 3.9mm)⁵ consists of three individual lenses: one spherical glass lens and two aspheric plastic lenses. Opto-mechanical components were made with LIGA technology and incorporated self-centering mounts. The objective on the right (outer diameter: 8mm)^{3, 10} was built using five aspheric injection-molded plastic lenses, each incorporating V-groves and assembled using precision spheres. Parts B and C (center and right) of Figure 2 show different views of a NA=0.4, FOV=250μm objective (outer diameter: 1.9mm) that consists of four individual elements: one spherical glass lens and three aspheric sol-gel glass lenses fabricated using grayscale lithography.

Figure 2. Miniature endomicroscope objectives. A: NA=1.0, FOV=250μm objective, featuring a 3.9mm objective at left and an 8mm plastic-injection molded objective at right; B: components used for building a NA=0.4, FOV=250μm objective with an outer diameter of 1.9mm; C: side view of an assembled NA=0.4 objective as seen in B.

Thanks to significant technological progress it is now possible to build low-cost, diffraction-limited high-performance miniature objectives. Remaining challenges in our research include the research and development of new optical materials to increase the range of dispersion, as expressed by the Abbe number, and to reach higher refractive indices. The development of new materials must be matched by new optical designs and the fabrication of multi-lens systems that deliver accurate correction of chromatic aberrations. The ultimate goal, however, is effective mass production to bring these sophisticated new devices into routine diagnostic use.