Tsafrir Kolatt, PhD
3.3 Other solutions
4 The opportunity with small cameras
5 Summary and Conclusions
Ts.K. would like to thank Aaron Jaffe for his valuable comments on the manuscript.
Recent medical trends have highlighted the functionality and value of miniature cameras. The quality of healthcare can be significantly enhanced by affording direct vision of otherwise inaccessible organs. Miniature cameras, commonly denoting a total diameter of less than 3mm to 3.2mm, present a new set of challenges together with their inherent benefits. This paper reviews the main shifts in medical practice and equipment that paved the way for the acceptance of miniature cameras, followed by a discussion of the main elements of such changes and an illustration of the benefits of direct vision in several select applications.Key words: medical imaging, image detectors, image quality
Of the prevalent trends in medicine during the last two decades, some are purely science-driven, a few are rooted in patient attitudes and others are influenced by the advancement of modern technology. To name a few of these trends, medicine is becoming more and more evidence-based, whereby physicians are less inclined to rely on subjective interpretations and instead seek documented measurements using sensors and imaging devices, as well as population surveys and biological indicators.
Conversely, current medicine is drifting towards a personal orientation, not in terms of physicians' personal attitudes, but in the sense that each patient represents a unique biological system that shares many general characteristics with the human species, while concurrently constituting a unique manifestation. As a phenotype, each patient warrants personally-tailored therapy, protocols and medical treatment routines , from diagnosis through follow-up. Thus, scientific justification of modern medicine is gained through the careful analysis of the population at large, but treatments are derived and formulated via careful consideration of the individual patient. Developments in information technology facilitate the accumulation and accessibility of data that can be analyzed at a later date in order to gain insights with regards to maladies, pathologies and curing methods. Moreover, when combined with the consistently increasing utilization of information sources and improved analysis techniques, this surge of information will frequently provide convincing evidence that may be used to refute long-standing medical misconceptions, thereby improving patient management.
The information explosion prompts new modalities to promote better understanding of the causes and evolution of illness, with imaging modalities occupying a dominant position. Better imaging, as a term, encompasses quite a few notions: more realistic imaging such as three-dimensional stereoscopic imaging, higher resolution, images from previously unattainable locations or angles of view, and different wavelength ranges facilitating otherwise undetectable features of the human body. Good examples of the latter are X-ray, gamma-ray (PET), infrared, radio-frequency and more. Different wave-forms, like ultrasound, have also been helpful where visual, electro-magnetic techniques fail to obtain the necessary data. Parallel to the widening of the frequency and wave-form horizons, two other critical abilities have emerged: the combination of many still images or video streams into a single multi-faceted view of the human body, e.g. computerized tomography (CT), and the increasing computational power that has been harnessed to augment imaging information through image processing (see e.g., (Rabbetts 2007)).
The concept that computer based processing of accumulated digital "raw" data could portray a clearer view of the object's condition led to computer-aided-diagnosis or computer-aided-technology (CAD/CAT) aiming to help physicians achieve more accurate diagnoses, higher sensitivity levels and somewhat better specificity in many routine tasks (Henning Müller 2004). Examples range from nodule or polyp identifications within documented scans to automatic cell counting under the pathologist's microscope.
The multiple modalities that are now available have also contributed to the use of multiple viewing angles that may allow physicians to better discern the patient's condition. As an example, registrations of MRI and CT scans have resolved the shortcomings of each separate modality.
The introduction of the 'personalized medicine' notion and the availability of imaging modalities with computerized analysis soon led to accurate localization of pathological conditions (Bankman 2008). This localization, in turn, facilitated pinpoint treatment such as interventional chirurgical procedure. High-precision abilities have enabled these procedures to be conducted in a minimally invasive and well localized manner. When invasive treatment is avoided as much as possible, the general well-being of the patient increases.
It has become clear that patient comfort and well-being, as trivial as this may sound, contributes tremendously to the success of the treatment. Thus, minimally invasive methods emerged as the state-of-the-art in patient diagnosis and treatment. When invasiveness cannot be altogether avoided, it is only gradually introduced. Attempts are made to first utilize natural orifices, either as a direct portal to the region of interest or as an enabling passage to the final destination using miniature incisions. Thus, of the notion that a single incision (and a small one at that!) should be regarded as a port for laparoscopic procedures or NOTES (Natural orifice Translumenal Endoscopic Surgery (Chamberlain RS 2009, Abu Gazala 2012)), for example, brought a significant advantage to the standard of care for many conditions that otherwise would have been treated with the vigorous use of a scalpel.
These advancements acted as a call for technology development. The ability to refrain from making incisions in order to view internal organs, or alternatively, to rely on smaller and smaller incisions and operating tools, suggested that vision, either direct or indirect, may also be useful and necessary in order to take full advantage of these minimally-invasive techniques (Rosen M 2011). Continuous view (i.e., video) lends another aspect to the treatment – the capability to "close the loop", namely to monitor the diagnosis (e.g., biopsy) or treatment in real time, change the tools, adapt the maneuvers, and improve the protocol according to the visual information, while directly observing the effect of these changes.
It is precisely at this point that miniature cameras enter the picture. The combination of evidence-based (i.e., image) medicine, with the consideration of the patient's individual anatomy and condition ("personalized medicine") together with the newly-gained information and the extraction of the valuable parameters from it, demands the use of miniature video cameras. When technology enabled low-cost manufacturing, the disposable nature of the cameras has also played an important role in their desirability for use in medical devices which are more commonly intended for single use. Reservation of the term "miniature camera" for outer diameters of less than 3mm to 3.2mm, as specified earlier, is mainly derived from the inner diameter of standard endoscope working channels. However, today's cameras may be far smaller than these dimensions and are not necessarily introduced through an endoscope's working channel.
In summary, recent trends in medicine support each other and spur yet more advancement that in turn provokes science and technology to devise new methods and to improve upon existing technology to catch up with the exacting requirements of modern medicine.
2 Imaging modalities – limitations and challenges
As much as medicine attempts to be an advanced and cutting-edge quasi-science, it is still, to the benefit of us all, a rather conservative field. The progress in medicine and the paradigm shifts occur at a relatively slow pace, both for the sake of patient safety and due to economic factors.
Since the biggest body of evidence to date for the diagnosis of many pathologies is based on visual inspection by the human eye (colors, shapes, sizes, etc.), whether in-vivo or in-vitro, physicians prefer to continue relying on such routines when they diagnose and evaluate pathological conditions.
In the frequency spectrum visible to the unaided eye, and in abutting frequencies (near infrared and ultraviolet), it is common to assess and evaluate the tissue or organ condition through visual inspection and its technological counterpart – photography. The fact that photography (or filmography) does not exactly manifest the same impression as the one obtained by the naked eye should be taken into account, but should not be a deterring factor for the use of the former as a diagnostic means.
The twentieth century semi-conductor technology produced the CCD and CMOS sensors as affordable - and more importantly, accurate and scalable - devices for visual imaging needs. Medicine soon adopted these solutions for its own needs, which paved the road to visual imaging in many common clinical scenarios. The marriage between fiber-optics technology and visual imagers allowed, for the first time, in-vivo visual exploration inside the human body, without the need for highly invasive surgery.
However, the motivation for miniature cameras goes beyond the mere ability to penetrate narrow, small locations within the human body. The benefits of smaller incisions are clear: lower risk, faster healing and recovery, smaller chance for related infections, reduced sedation and pain and aesthetic appearance (scars). All of these may lead to an additional advantage which can be superficially seen as purely economic – the transition of procedures from the operating room to office-based clinics.
Better yet are the zero-incision procedures that have become feasible by the introduction of miniature cameras. Trans-oral procedures that may involve intubation can now be performed transnasally, even in children. No harm is caused by the examination, and this alone may cause the procedures to be so appealing as to be considered screening routines rather than symptom-triggered tests.
Going beyond visualization per-se, the integration of mini-cameras within endoscopes of all sorts vacates precious space for other elements within the endoscope and allows more complicated procedures to be performed via surgical endoscopy. A useful embodiment of such utilization is found in the "mother - baby" endoscope combination, in which a small-diameter camera system is threaded through the working channel of the "mother" scope. This arrangement enables the removal of the visualization tool and the use of the channel for a different tool as required during the endoscopic session. The "mother" endoscope serves as a robust, fully steerable modality supporting the somewhat less controllable baby-scope. The baby-scope can nonetheless reach places that are beyond the reach of the mother scope due to its smaller size and lack of certain physical and anatomical constraints. One notable usage of this method can be found in the NOTES-type procedures that accesses the peritoneum through the stomach.
3 Visualization methods: Fibers Versus Cameras
3.1 Fiber-based cameras
Upon the introduction of miniature cameras, the natural tendency would have been to combine them with optic fiber technology. Optic fiber bundles containing thousands of individual fibers, each one acting as a mirroring channel, may be connected directly to the cameras. Optics are usually placed at the distal tip of the bundle, on top of its polished edge. Nevertheless, the distal tip optics tend to be one of the vulnerabilities of this technology. At the bundle proximal end, there is usually a coupler to the camera to facilitate the best image quality possible.
This arrangement entails several shortcomings. First, the common fibers have a circular cross section which causes "blank spaces" between the bundled fibers. Although hexagonal shape fibers have been proposed, their use is not widespread due inter alia to high costs and complicated manufacturing. This brings about another limitation of the fiber solution, namely the cost per performance factor. Usually, vision optic fibers, as opposed to illumination fibers, are rather expensive. Had they been robust, they could have been a cost-effective solution. Unfortunately, the fragility of the fibers prevents them from being a truly multiple-use device, especially if they are not placed in a rigid endoscope.
The flexible endoscopes/catheters pose an acute challenge for fiber-based devices. The limited curvature radius that the bundles can withstand without breaking is a severe constraint when it comes to small cameras whose main advantage is their ability to accommodate narrow, winding lumens. Sharp turns, even when generally within device specifications, end up causing fractures along several or even many of the fibers. Such fractures result in immediate image quality deterioration.
Another obstacle that may arise during the design and implementation of fiber-based medical devices is the somewhat limited field of view. Since fiber-based solutions necessarily dictate at least two coupling elements (at the tip and then at the proximal end), the tip coupling suffers from the optical matching of the lens (if any) with the polished fibers behind it. This usually restricts the field of view (namely the opening angle) of the front lens and thus constrains the device's function. The rear (proximal) end coupling must be performed in a very precise way in order to bring the bundle plane to the focal plane of the proximal camera, when the two planes are close to each other for full coverage of the bundle.
Another potential shortcoming of fiber-based devices may be expressed when the desired angle of view is different than 0º. There are two alternatives to meeting this requirement: the fiber tip maybe polished diagonally to accommodate the desired angle and acts like a periscope mirror or a prism, or else, the prism is coupled directly to the tip. Both solutions carry unwanted consequences: the diagonal polish limits the use within a specific medium (e.g. water) and precludes a full (round) image when introduced to a different medium (e.g. air). The prism solution makes the abovementioned coupling between the distal tip and the prism even more challenging and casts an additional cost burden on the device. Fiber-based scopes may go to diameters of ~1mm.
Most of the abovementioned limitations of fiber-based imaging solutions are alleviated by the use of a distal tip camera, instead of at the proximal end where the light must first pass through the fiber bundle. When the camera is situated at the front end of the device, it is the element that first encounters the collected light. Thereafter, an electric wire transfers the video signals detected by the camera.
3.2 Camera at tip solution
Even though the camera at tip solution is usually superior to the fiber-based solution, it bears its own challenges. Most of these are directly derived by the camera dimensions that dictate unconventional solutions for packing, processing and control of the camera.
The basic element of all cameras is the sensor that is situated behind the lens, its main characteristics being pixel size and pixel density (namely the number of pixels and the size of the sensor). For some applications the sensor sensitivity and its dynamic range are also of concern. Generally speaking, a CCD sensor will be more sensitive (see e.g. (P. Coan 2006)) and of a higher dynamic range in comparison to a CMOS sensor, it will also be generally more expensive.
In this solution, the optics of the camera is the element which should satisfy the field of view and the angle of view (the main optical axis direction with respect to the endoscope main axis) as deemed by the application.
The optics and the sensor are embedded together within the "house" or the "barrel", with or without adjusting the optic-sensor distance to provide the desired focus, focal depth, field of view, etc.
The sensor can be wired by various techniques, but usually direct placement onto a PCB or wire-bonding is used. In the production process, special attention is paid to accurate alignment between the optics, the sensor and its placement. This not only contributes to improved image quality, but also enables the minimization of the camera's overall dimension.
The electric wires cannot exceed the camera diameter, without defeating the raison d'être of the miniature camera. This requirement is not a trivial one, due to the requirement for shielding to reduce electro-magnetic interferences.
Custom filters, coating materials, connectors, etc. are also usually present within each camera system and should be coordinated with the sensor micro-lenses, Bayer format as well as outer parameters such as the expected illumination, environment and more.
In the camera in tip solution, the angle of view can be changed by a prism, like in the fiber-based case, with a prism-lens perfect coupler. Better yet, due the small length dimension of these cameras, some of them can be placed at an angle to the main mechanical axis of the endoscope, thereby resolving this issue altogether. Camera at scopes tip may go today down to outer diameters of ~1mm (e.g., Medigus Ltd.).The camera per se does not transmit a visible image, but is connected to an image processing unit that must also be synchronized and built according to the camera specifications.
In short (no pun intended), the main challenges in miniature camera implementation and production are (i) Small dimensions, not only the overall diameter but also the length dimension - a long rigid camera is not practical when inserted into a narrow meandering lumen; (ii) Accurate assembly and - in particular - precise optical alignment; (iii) Application-oriented design and implementation; (iv) Full match between the camera specifications and the accompanying image processing equipment; (v) Tight linkage between the camera assembly and its surroundings (e.g. illumination) (vi) Cost-effective assembly. Keep in mind that being a medical camera, all of the camera elements to which the patient may be exposed must be strictly bio-compatible. Disinfection or sterilization in-between patients can be avoided altogether by the virtue of the single-use, cheap camera.
In section 4 we will refer to different degrees of bio-compatibility with regards to the applications that the camera should serve.
3.3 Other solutions
One type celebrated miniature camera is embedded within medical capsules ("pill-cams"). Three main differences exist between the video cameras discussed here and the pill cameras. First, the size of the pill camera does not strictly fit in the "miniature camera" category. The second difference is the pill's independence in terms of its motion and power. This independence is technically restrictive (battery lifetime, strict power management, inability to steer or maneuver) but clinically advantageous for specific applications. For instance, the patient's freedom while under pill inspection cycle and the prolonged duration of the inspection are two positive aspects of the pill camera independence. The third difference is the fact that the pill does not have to show real time images. When the pill travels at a slow enough pace, its power management calls for a slowdown of the frame rate without an increase in latency due to the "nonreal-time" mode of operation. Needless to mention is the fact that the pill is also disposable.
4 The opportunity with small cameras
One of the main areas in which miniature cameras can demonstrate a real advantage is in the field of modern medicine. As described in section 1, modern medicine is becoming increasingly inclined toward minimally invasive procedures, whether diagnostic tasks or true treatments.
Miniature cameras provide a natural element within this trend, with which the physicians can gain valuable, direct information that formerly was only possible to obtain after surgical openings or frequencies outside the visible spectrum (X-ray, infrared, etc.).
We have named several fields of medicine in which miniature cameras may, and in some cases already do, provide indications and data to improve healthcare. Lumens of all sorts are the obvious candidate for miniature camera diagnoses. In the cardiovascular field, the main challenge is to clear out blood in order to enable visual inspection of blood vessels, for example. Other lumens are less problematic (e.g. larynx, pharynx, esophagus, ear canal, nostrils) but some are collapsed in their regular condition and must be manipulated in order to exploit the full view with the miniature camera aid.
In natural cavities, one of the main challenges is to provide sufficient illumination (if the cavity is spacious and a global view is required). For large cavities, the finite dynamic range of the cameras also complicates simultaneous viewing of nearby and more distant objects. However, even for a large dynamic range, a real time application usually calls for minimal latency and therefore limits the maximum allowed exposure time to compensate for poor illumination. This is a major difference between still images of all sorts, or semi-still (namely low video rate, see above discussion of the "pill camera") and a high video rate camera. The current standard rate is 30 fps (frames per second).
One of the most complicated tasks is the a-priori simulation or design of the illumination conditions prevalent within a cavity or a lumen. The unknown structure, reflections, tissue condition and so on, prevent the designers from applying a single solution to multiple applications. As a result, the development process of a device for specific applications is necessarily an iterative process that should encompass all aspects of the expected environment.
Single incision applications may also serve as an appropriate platform for miniature cameras, with or without robotic technology (Oleynikov D 2005, Ahmed I 2011, Otten ND 2011, Greaves N 2011). Such procedures can replicate laparoscopic procedures with smaller incisions and trocars, to bring the latter closer to the object under inspection and to circumvent the inconvenience and limitations of rigid laparoscopes. This is particularly relevant in view of the smaller diameter of the miniature camera that makes room for more steering mechanisms.
Rigid endoscopes enjoy the same benefits and are usually needed in one of two occasions: when lateral force or moment is applied in order to adjust the direction of view so that it meets the region of interest (e.g. in arthroscopy) or when the camera itself acts as a needle to penetrate through a wall.
In general, the flexibility of the miniature camera enables it to wind around obstacles and within lumens, and to penetrate strictures and walls if used in the context of rigid endoscopes. This is also evident when used in combination with the mother-endoscope vehicle. The mother-endoscope provides the gross steering, the capability to apply force on the lumen or tissues and acts as a protective shield for the more delicate mini-camera. The mini-camera goes the "last mile" and peeks at more distant objects where the mother-scope dimensions do not allow a close look. It can also use the mother outlet structure as a leaning wall or even a direction changer, like in ERCP procedures (Chamberlain RS 2009).
The specific application is also what dictates the level of disinfection or sterilization the miniature camera needs to withstand (a single time, during production). This may range from simple disinfection for truly non-invasive applications (e.g. ear canal) all the way to high sterilization degree when introduced into a blood vessel.
Clearly, the advantages of the small dimensions of such cameras are fully exploited in combination with other tools such as etchers, biopsy collectors, object removers, etc. A separate, dedicated working channel for the vision device or a common channel that accommodates other tools according to the immediate needs are both possible.
Recently, a few innovative augmenting elements have been introduced to make yet better use of miniature cameras. Some are mature and ready for integration with miniature camera technology, while others are at a more preliminary stage, but nonetheless hold significant promise for new types of treatment. These elements include a miniature integrated zoom mechanism, an electric steering mechanism, magnetic steering capabilities, stereoscopic view as derived by multiple camera utilization, and more.
5 Summary and Conclusions
Miniature cameras can be merely referred to as gadgets, however a closer look at the enabling technology they introduce, reveal their real benefits. These cameras are already changing the way physicians are gathering the necessary information in order to reach at correct patient management decisions. The ability to monitor procedures, simple or complicated, a larynx inspection or ERCP stone removal, modifies the way these procedures are conducted.
The demand for sophisticated medical interventions and technological progress are two elements in a closed loop that feed each other. Even though miniature cameras are not yet as good, in terms of image quality, as big, amateur cameras, they have become indispensable. The need to see inner organs through small windows or through orifices together with the desire to monitor therapies are by now part of the standard care for some conditions. We foresee that in the near future, it would consider mal practice when a physician will not be using the option to judge by direct vision and to treat with eyes wide open. Every organ that is accessible will be accessed – big and small. The miniature cameras will serve this function to provide better, more precise patient care management.6 Acknowledgements
Ts.K. would like to thank Aaron Jaffe for his valuable comments on the manuscript.
Abu Gazala, Shussman, Abu Gazala, Schlager, Elazary, Ponomernco, Khalaila, Rivkind, and Mintz. "Miniature Camera for Enhanced Visualization for Single-port Surgery and Notes." Journal of Laparoendoscopic & Advanced Surgical Techniques , 2012: 22 (10) 984-8.
Ahmed I, Paraskeva P. "A clinical review of single-incision laparoscopic surgery." The Surgeon : Journal of the Royal Colleges of Surgeons of Edinburgh and Ireland, 2011: 9(6):341-351.
Bankman, I. Handbook of Medical Imaging. Processing and Analysis. Burlington: Academic Press, 2008.
Chamberlain RS, Sakpal SV. "A comprehensive review of single-incision laparoscopic surgery (SILS) and natural orifice transluminal endoscopic surgery (NOTES) techniques for cholecystectomy." Journal of Gastrointestinal Surgery : Official Journal of the Society for Surgery of the Alimentary Tract , 2009: 13(9):1733-1740.
Greaves N, Nicholson J. "Single incision laparoscopic surgery in general surgery: a review." Annals of the Royal College of Surgeons of England, 2011: 93(6):437-440.
Henning Müller, Nicolas Michoux, David Bandon, Antoine Geissbuhler. "A review of content-based image retrieval." International Journal of Medical Informatics, 2004: (73) 1-23.
Oleynikov D, Rentschler ME, Dumpert J, Platt SR, Farritor SM. "In vivo robotic laparoscopy." Surgical Innovation, 2005: 12 (2) 177-181.
Otten ND, Farritor SM, Lehman AC, Wortman TD, McCormick RL, Markvicka E, Oleynikov D. "Miniature in vivo cameras for use in single-incision robotic surgery - biomed 2011." Biomedical Sciences Instrumentation, 2011: 47:165-170.
P. Coan, A. Peterzol, S. Fiedler, C. Ponchut, J. C. Labiche and A. Bravin. "Evaluation of imaging performance of a taper optics CCD `FReLoN' camera designed for medical imaging." J. Synchrotron Rad., 2006: (13) 260-270.
Rabbetts, Ronald B. Clinical Visual Optics. London: Elsevier, 2007.
Rosen M, Ponsky J. "Minimally invasive surgery." Endoscopy, 2011: 33(4):358-366.
Simi M, Pickens R, Menciassi A, Herrell SD, Valdastri P. "Fine Tilt Tuning of a Laparoscopic Camera by Local Magnetic Actuation: Two-Port Nephrectomy Experience on Human Cadavers." Surg Innov., 2012.