Using Digital Technology to Enhance Restorative Dentistry

There is a growing array of digital technology and computerized systems for restorative dental treatment. In general, all of the current systems follow the basic workflow of computer-assisted design computer-assisted machining (CAD/CAM) to create a restoration. There are three main sequences to this workflow. The first sequence is to capture or record the intraoral condition to the computer. This involves the use of a scanner or intraoral camera. Once the data has been recorded to the computer, a software program (CAD) is used to complete the custom design of the desired restoration. This may involve a full-contour design of the restoration or just the internal coping or substructure of the final restoration. The final sequence requires a milling device to fabricate the actual restoration from the design data in the CAD program. Although there are a number of possible techniques for the milling systems, currently the most common technique is a wet grinding subtractive milling process during which a preformed block of material is shaped by cutting instruments.

A number of manufacturers have developed equipment to accomplish the CAD/CAM workflow, and the capabilities and differences between the equipment is the most common discussion dentists experience when considering whether to invest in CAD/CAM technology for their office. This is essentially a discussion about the process of making a digital restoration. However, it is the final restoration that determines the clinical outcome of treatment. And the outcome is the most important aspect of any discussion about digital systems. Only after determining the usefulness of the outcome is a discussion of the process a significant consideration. It is not how quickly or easily a restoration is fabricated that is the most critical consideration, but rather how well the restoration functions and performs for the patient.

Digital Systems

Commercially available digital systems for the dental office are commonly divided into two categories: digital impression systems and chairside CAD/CAM systems. Both types of systems must be able to accurately record the intraoral condition to a computer data file with a scanner or camera. It is what the systems accomplish after recording the data file that distinguishes them.

Digital impression systems are designed to electronically transmit the recorded data file to the dental laboratory for restoration fabrication. Once dental laboratories have downloaded the file, they can have models processed from the file. Any conventional laboratory process can be used to fabricate the restoration once the laboratory receives the processed models. Alternatively, the dental laboratory can use the transmitted data file in a computer design (CAD) program to create a full-contour restoration or coping that can be refined on the processed models.¹ The two most common examples of digital impression systems are the Lava™ Chairside Oral Scanner C.O.S. (3M ESPE, www.3MESPE.com) and the iTero™ system (Cadent, www.cadentinc.com).

The CEREC^® Acquisition Center (AC) (Sirona Dental Systems, www.sirona.com) and E4D Dentist™ system (D4D Technologies, www.e4dsky.com) are the two available chairside CAD/CAM systems.^2,3 Chairside CAD/CAM systems complete all three steps of the imaging, design, and milling process in the dental office to fabricate full-contour restorations within the time span of a single appointment.

Both chairside CAD/CAM systems have manufacturer-specific software programs that permit the production of single-tooth ceramic or composite inlays, onlays, veneers, and crowns. Both chairside CAD/CAM systems also offer the option to be used as purely digital impression systems. The CEREC Connect system for the CEREC AC unit and the E4D Sky Network for the E4D Dentist system are upgrades that allow for electronic transmission of the digital scan file to the dental laboratory for restoration fabrication.

Digital Cameras and Scanners

The one component common to all digital systems is the intraoral camera or scanner used to record data files of the intraoral condition. Recording the digital file negates the need for a conventional impression to fabricate a restoration and is a significant paradigm shift in the clinical workflow for dentists.

The CEREC AC system includes the Bluecam—a light-emitting diode (LED) camera that records a series of overlapping single images that the software calculates into a 3-dimensional (3-D) virtual model of the dentition. The camera records the image data reflected from the surface of the teeth and soft tissues, and a titanium dioxide powder is required to create a uniform reflective surface. The shorter wavelength of the LED blue light has been measured to have a higher resolution compared to that of a red laser light.⁴ The Bluecam can be used in either a manual or automatic mode. When the camera is used in the automatic mode, it is unable to record data while the camera is moving or shaking. This prevents the capture of blurred images that would be inaccurate.

The E4D Dentist IntraOral Digitizer is a single-image camera that uses a red laser light to record intraoral images.⁵ It also works by recording reflected data from the hard and soft tissues. However, it only requires the use of a reflective medium (E4D Accent liquid) when scanning through a thin, transparent area of the cavity preparation. The camera can be used in either a manual image capture mode or automatic image capture mode with Rapid Scan. A series of separate images are recorded from the occlusal, lingual, and facial views for a true 3-D capture.⁶ The software provides immediate feedback on each scanned image to ensure all images are accurately scanned.

The iTero intraoral scanner uses a parallel confocal white light and red laser light camera to record a series of single images to create a 3-D model.^5,7 The scanner emits a beam of light that is reflected off the tooth surface. Only data reflected back through the filtering device at the correct focal distance is recorded.⁸ The scanner captures 100,000 points of laser light and focuses accurately to 300 focal depths spaced 50 µm apart.^6,8 Although light is reflected from the surface of the tooth to record the data, no reflective powder or coating is required and the camera can be placed in contact with the teeth. The operator is prompted to record a series of scans from the occlusal, facial, lingual, mesio-proximal, and disto-proximal angles of the prepared tooth and additional scans for adjacent teeth. The opposing dentition is scanned separately.⁸ The scanned series of images is not continuous, so individual images may be retaken until adequate data is obtained.⁶ A total of 15 to 30 scanned images may be required to record the preparation, opposing teeth, and occlusal relationships.⁸ Ultimately, the software merges common data from all scans and proposes a virtually articulated cast.⁵

Although most cameras or scanners record a series of overlapping images that are combined by the respective software program into a virtual 3-D model, the Lava Chairside Oral Scanner C.O.S. camera is a video camera that can process and fabricate a 3-D virtual model in real-time to visualize it on the computer monitor.⁹ It uses “active wavefront sampling” to record up to 20 3-D data sets per second. Three sensors record the clinical situation from varying perspectives and use proprietary image-processing algorithms to process the model.¹⁰ The virtual model can be visualized as either a 2-dimensional (2-D) or 3-D virtual model; it can also be visualized with 3-D glasses to verify aspects of the preparations and models. The entire model does not need to be recorded in a single recording pass. The dentition can be recorded in strips or sections, and the computer can assemble the strip scans in a real-time single 3-D virtual model.¹¹ The operator has a field of view of approximately 10 mm x 13.5 mm, and the camera must stay a working distance of between 5 mm and 15 mm from the surface being recorded; otherwise, a fail-safe component stops capturing data, preventing poor data from being included in the scan.⁵ A Scan Rewind function allows the user to rewind and delete 10-second portions of the scanned video rather than delete the entire scan. Smart Scan function automatically removes extraneous data that is recorded in one scan and not duplicated in successive scans. For example, if a portion of the tongue or cheek is recorded in the first strip scan and is not recorded in the second overlapping strip scan, the computer deletes the extraneous information from the virtual model.

Camera Size

An obvious concern of many dentists is the size of the camera. Cameras are much larger than a dental handpiece or a curing light, so this leads to the issue of how much space is needed for the camera to fit in the mouth and record, for example, a second molar. While the compact size of a camera head can be an advantage intraorally, the overall larger size of the camera is due to the need for convenient, ergonomic handling. The body of the camera does not go into the mouth any more than an entire curing light is designed to fit into the mouth. The camera body is designed to be balanced in the user’s hand and comfortable to handle to facilitate the camera’s rotational or translational movements while recording the dentition. So although the small size of the camera head is a commonly touted attribute of most systems, dentists must be comfortable handling the camera body to determine the ergonomic “feel” and evaluate their ability to manipulate the camera for capturing images.

Single-Image Cameras

The CEREC Bluecam, E4D Intraoral digitizer, and iTero scanner are considered single-image cameras. These cameras capture a series of individual digital images that overlap one another. The overlapping images are “stitched together” by the computer software program to process a single 3-D virtual model. All cameras work in a single line of sight. This means that the camera can only record data that is in the direct line of sight of the camera. Generally, the first image recorded is from an occlusal direction. The data cervical to the height of contour of an unprepared tooth would not be visible to the direct line of sight of the camera. A series of images must be captured as the camera is rotated towards the facial and lingual directions to capture data cervical to the height of contour.

A common comparison between systems is the number of images a system requires for a case. The number of overlapping digital images recorded will determine the size of the virtual model on the computer. The more teeth desired in the working model, the more images are necessary to record it. Although some of the systems indicate that only a limited number of images are required, in actual practice, more images are generally used, as the size of the model desired often mandates the need for additional images. The number of images recorded is not as important a factor as the time or efficiency in recording the images for the case. The longer the time required, the longer the time the operating field must be well-isolated and free of moisture for the images to be accurate.

Open and Closed Architecture

All of the computerized systems record the digital impression to a data file in the manufacturer’s software program. The initial versions of these proprietary data files were designed with a “closed architecture” concept in that the digital files could only be read and used by equipment from the same manufacturer using the manufacturer’s software program. This is still true for the digital files used by the CEREC AC and E4D systems for full-contour chairside restorations. The digital files cannot be moved between the two manufacturers’ systems for processing the chairside restoration. However, with the advent of digital impression systems, dental laboratories were faced with the potential problem of having to acquire systems from each manufacturer in order to manage all the data files they received from their client doctors. More recently, manufacturers of computerized systems have moved to “open architecture” with their digital files. Multiple corporate partnerships have developed, allowing the use of a specific manufacturer’s digital files across a number of different software programs and CAD/CAM equipment. A number of laboratories use laboratory-based CAD/CAM systems such as Dental Wings (Dental Wings, Inc., www.dentalwings.com) or 3Shape System (3Shape A/S, 3Shape Dental Systems, www.3shape.com) to process digital files from every computerized manufacturer available for the dental office.

Digital Impressions

A well-accepted principle of restorative dentistry is that the final restoration can be only as accurate and well adapted as the final impression. The clinical challenge is to provide an accurate final impression of the intraoral condition to the laboratory. This concept is equally true for digital impressions. The final restoration can only be as accurate as the recorded data file.

All digital impression systems and chairside CAD/CAM systems rely on the ability to accurately record the data file, and there are several principles that are common to all the cameras that significantly influence the outcome. The first is that digital impressions are as sensitive to moisture contamination as traditional impression materials. Blood and saliva obscures the surface of the tooth or margin from the camera and prevents an accurate recording. At best, the camera records the moisture as an inaccurate surface contour; at worst, no data is recorded where moisture has collected. In either situation, an accurate restoration cannot be fabricated.

A second principle is that inadequate management and retraction of soft tissues may prevent visualization of the marginal areas, resulting in an inaccurate recording with the camera. As desirable as it may be to scan through soft tissues, this is not possible with current systems. Digital cameras can only record data that is directly visible to the camera lens.

Soft-tissue retraction is somewhat different for digital impressions compared to traditional impressions. Traditional impressions generally require that soft tissues be retracted laterally as well as vertically past the tooth preparation margins. The lateral soft-tissue retraction allows for a bulk of impression material at the margin to avoid tearing it upon removal. The vertical soft-tissue retraction allows for impressing tooth structure cervical to the margin to ensure the margin is accurately recorded. Digital impressions only require the soft tissues to be retracted sufficiently laterally to visualize the margins. This may be as little as 150 µm to register the margin of the tooth preparation separately from the soft tissues. This is a primary reason diode lasers are particularly popular adjunctive instruments for digital impressions, as they efficiently create lateral retraction while preventing bleeding and ensuring a dry field of view.

A common concern relative to the use of digital impression systems is the degree to which they can record subgingival margins. As described earlier, all cameras will record data that is visible to the camera lens. Therefore, the degree to which a subgingival margin can be recorded is dependent on the ability to visualize it. As long as the camera can visualize the margin, it will be recorded. A more critical concern may be how far subgingivally a margin should be recorded. This is especially an issue if an adhesive restoration is being planned for the case, as the outcome of the final restoration is also dependent on the ability to predictably bond the restoration to the tooth, and careful isolation is a requirement for adhesive success.

A general concern expressed by those unfamiliar with digital impression systems is the amount of time it takes to make a digital impression compared to a traditional impression. One obvious factor is that the operator’s comfort level while using an intraoral camera significantly impacts the length of time needed to record the images. For those who have never used a wand-type intraoral camera, the initial learning objective is to be comfortable using the camera intraorally while visualizing the image on the computer monitor. A recommended learning technique is to practice scanning volunteer “patients” without tooth preparations to be comfortable with the camera prior to focusing on scanning tooth preparations. Most anecdotal reports by experienced users of digital systems indicate that digital impressions are more efficient than traditional impressions, which usually require 5 to 7 minutes of set time. For example, the maximum scan time for a single scan with the Lava C.O.S. system is 7 minutes, with most quadrants requiring only 2 to 3 minutes. Most quadrant scans with the CEREC AC Bluecam can be completed in less than 45 seconds. However, this may not be true for all digital impression systems. A recent randomized clinical trial questioned the efficacy of digital impressions. The study compared crowns fabricated with either iTero digital impressions or a conventional impression technique. The authors reported that both impression and crown adjustment times were significantly longer when the digital technique was used, even though no significant differences were found in marginal fit.¹²

Once the virtual models for both the prepared quadrant or arch and the opposing quadrant or arch have been calculated, they must be “mounted” or aligned in the software program prior to fabricating the model or the restoration. Several techniques have been utilized as the systems have evolved. One technique is to make a static bite registration with a polyvinylsiloxane material. The surface of the bite registration is scanned instead of an opposing dentition model and the software program matches the bite registration model to the preparation model (Figure 1). The surface of the bite registration is used to refine the occlusal contours and contacts of the restoration proposal. The most common technique is to use a buccal bite recording. The patient is brought into centric occlusion and the maxillary and mandibular teeth are recorded from the buccal direction (Figure 2). This records the static position of the dentition, and the software program uses it to align the opposing digital models to record the patient’s occlusal relationships. No computerized system has the ability to digitally record the functional movements of the mandible to transfer them to the software program. For larger cases, where the lateral guidance is critical, a facebow and conventional check bites can be used to mount the models on a semi-adjustable or fully adjustable articulator.

The virtual alignment of the opposing recorded models on the computer is an obvious critical issue in determining the accuracy of occlusion for the resulting restoration. An in vitro study evaluated the change in vertical dimension in scanned models of quadrants and full arches based on two common clinical conditions using the Lava C.O.S. system.¹³ Tooth No. 14 was prepared on a typodont to simulate a single crown preparation with a vertical stop distal to the preparation. Teeth Nos. 2 and 3 were prepared on a typodont to simulate adjacent crown preparations with no distal stop. Both clinical conditions were scanned with the Lava C.O.S. system, using both quadrant model scanning and full-arch scanning. The files were transmitted to the laboratory, where the stereolithography resin models were fabricated and mounted. The vertical dimension of each of the models was measured and compared with the vertical dimension of the master models. There was no significant difference in the vertical dimension of the models for the single crown preparation case whether a quadrant or full-arch scan was used. However, for the adjacent crown preparation case, the full-arch scan resulted in a more accurate recording of the vertical dimension compared to the quadrant scan. The resin models were not inaccurate; the difference in the vertical dimension was traced to the flexibility of the plastic articulators. The conclusion is that use of a full-arch or quadrant scan is equally accurate in recording vertical dimension; however, for larger cases, more care is required in the choice of the articulator to ensure an accurate vertical dimension.

Software

Digital systems offer a unique alternative technique for clinical dentistry based on application of computer technology. This represents a significant change in the clinical workflow for dentistry. Many dentists are not interested in becoming software engineers in order to practice. Digital system companies have each developed unique software programs to manage the digital impression process as well as the design and milling process.

Digital impression systems’ software programs basically record patient identification information, serve as a guide for the model-recording process, and then provide an online prescription form to electronically transmit the recorded digital file to the dental laboratory for restoration fabrication. The learning curve for these software programs is short, because there is little variation in the process of recording a digital impression, and data input is straightforward.

The software programs for the chairside CAD/CAM systems are more involved than those for the digital impression systems, because a CAD program must provide the necessary tools to virtually design the restoration on the computer monitor. A variety of tools—some intuitive, some not so intuitive—are featured in each manufacturer’s software program as aids in customizing the restoration design to the specific needs of the clinical case at hand. Generally, a 2-day basic training course is provided for the chairside CAD/CAM systems to learn the clinical workflow, including the design process. Although this may seem like a much longer and more drawn out learning curve, in most productive offices using chairside CAD/CAM systems, it is not the dentist but the dental assistants or dental laboratory technicians who are the primary “designers.” As with most technology advances, those who use it the most, use it the best.

A report on the largest group of private-practice dentists who have integrated chairside CAD/CAM technology deals with a corporate dental group that involves nine western states.¹⁴ From 2007 to 2012, 272 chairside CAD/CAM systems were integrated into practices, resulting in the fabrication of more than 425,000 chairside restorations. A unique aspect of this successful integration is that the clinical practice model is a team-based approach with the design-fabrication process managed almost exclusively by dental assistants.

Digital Restorations

Considerable research has been invested in the margin fit and internal adaptation of CAD/CAM restorations, as computerized systems introduced significantly different fabrication processes compared to those used for conventional dental restorations. The CEREC system was introduced in 1985 and has more than 25 years of published research relative to the accuracy of the chairside CAD/CAM process. A study that compared the fit of CEREC and laboratory-fabricated onlays reported no significant difference in the marginal gap, with a margin of 85 µm for the CEREC onlays.¹⁵ Another study measured the degree of crown preparation taper relative to the margin fit and reported a margin gap on the order of 53 µm to 67 µm for CEREC crowns.¹⁶ A third study compared the fit of CEREC crowns to several different laboratory-fabricated crowns and reported a margin gap of 47.5 + 19.5 µm for the CEREC crowns.¹⁷ Additionally, initial laboratory studies have documented the marginal fit and adaptation of restorations fabricated with more recently introduced systems.¹⁸

There are also a significant number of published clinical studies on chairside CAD/CAM restorations.¹⁹ A systematic review of clinical trials between 1986 and 1997 reviewed 29 reports on CEREC restorations.²⁰ They reported a failure rate of 2.6% and a mean survival rate of 97.4% over a 4-year period based on 2,862 restorations over 15 clinical trials. Another study evaluated 200 CEREC restorations in a private practice and reported a Kaplan-Meier survival probability of 90.4% after 10 years.²¹ A study of 2,328 CEREC inlays and onlays placed in 794 patients reported a Kaplan-Meier survival probability of 97.4% at 5 years, and 95.5% at 9 years.²² The longest clinical study to date on CEREC restorations regularly evaluated 1,011 over 18 years and reported a Kaplan-Meier survivability of 95% at 5 years, 91.6% at 7 years, 90% at 10 years, and 84.9% at 16.7 years.²³

A growing list of published studies document the accuracy of the more recently introduced digital impressions as at least equal to that of conventional impressions.^24-26 One study compared the margin fit of Lava crowns and porcelain-fused-to-metal (PFM) crowns made with a conventional polyvinylsiloxane impression and a digital impression with the Lava C.O.S. system.²⁷ The studies found no significant differences in the fit of the crown margins based on the impression technique. Additional studies provide some evidence that digital impressions are more accurate than traditional impressions. A recent randomized clinical trial measured the amount of time required to adjust matched zirconia crowns made from a traditional polyvinylsiloxane impression and a digital impression with the Lava C.O.S. system.²⁸ The study reported no significant difference in the time required to adjust the crowns prior to cementation based on the impression technique. However, the marginal fit and internal adaptation of the crowns was significantly better for those crowns made with digital impressions compared with traditional polyvinylsiloxane impressions. Similar results have also been reported for Lava C.O.S. digital impressions in laboratory studies.^10,22

Conclusion

Computerized dental systems offer innovative materials and techniques for restorative dental treatment. They are being incorporated into the workflow of many dental offices across the country with overwhelming patient acceptance. However, all the benefits of the new workflow may not be immediately realized until the learning curve has been completed. Predictable success with digital impressions is still dependent on traditional skills such as achieving optimal soft-tissue retraction and maintaining moisture control and isolation that will enable consistently accurate digital image acquisition. And, of course, accurate impressions—digital as well as traditional—are critical for precise, well-fitting dental restorations.

About the Author

Dennis J. Fasbinder, DDS, ABGD

Clinical Professor, Director of the Advanced Education in General Dentistry Program, Director of Computerized Dentistry (CompuDent) Unit, University of Michigan School of Dentistry, Ann Arbor, Michigan; Private Practice, Ann Arbor, Michigan

References

1. Fasbinder DJ. Digital workflow for the Lava COS system. Inside Dentistry. 2009;5(9)114-117.

2. Levine N. To the sky and beyond. Dental Products Report. 2009;Oct:116.

3. Mörmann WH. The evolution of the CEREC system. J Am Dent Assoc. 2006;137 suppl:7S-13S.

4. Mehl A, Ender A, Mörmann W, Attin T. Accuracy testing of a new intraoral 3D camera. Int J Comput Dent. 2009;12(1):11-28.

5. Kachalia PR, Geissberger MJ. Dentistry a la carte: in-office CAD/CAM technology. J Calif Dent Assoc. 2010;38(5):323-330.

6. Birnbaum NS, Aronson HB. Dental impressions using 3D digital scanners: virtual bes reality. Compend Contin Educ Dent. 2008;29(8):494-505.

7. Garg AK. Cadent iTero’s digital system for dental impressions: the end of trays and putty? Dent Implant Update. 2008;19(1):1-4.

8. Henkel GL. A comparison of fixed prostheses generated from conventional vs digitally scanned dental impressions. Compend Contin Educ Dent. 2007;28(8):422-431.

9. McMaster D, Cohen B, Spitz SD. Digital workflow. Dental Economics. 2008;98(8):30-36.

10. Syrek A, Reich G, Ranftl D, et al. Clinical evaluation of all-ceramic crowns fabricated from intraoral digital impressions based on the principle of active wavefront sampling. J Dent. 2010;38(7):553-559.

11. Fasbinder DJ. Digital dentistry: innovation for restorative treatment. Compend Contin Educ Dent. 2010;31 spec no 4:2-12.

12. Givan DA, Burgess JO, O’Neal SJ, Aponte AA. Prospective evaluation of ceramic crowns by digital and conventional impressions [abstract]. J Dent Res. 2011;90(spec iss A): Abstract 380.

13. Poticny DJ, Fasbinder DJ. Accuracy of digital model articulation for chairside CAD/CAM restorations [abstract]. J Dent Res. 2012;91(spec iss A): Abstract 705.

14. Rodgers C. A worldwide leader in digital dentistry. Inside Dentistry. 2012;8(7):68.

15. Denissen H, Dozić A, van der Zel J, van Waas M. Marginal fit and short-term clinical performance of porcelain-veneered CICERO, CEREC, and Procera inlays. J Prosthet Dent. 2000;84(5):506-513.

16. Nakamura T, Dei N, Kojima T, Wakabayashi K. Marginal and internal fit of Cerec 3 CAD/CAM all-ceramic crowns. Int J Prosthodont. 2003;16(3):244-248.

17. Ellingsen LA, Fasbinder DJ. In vitro evaluation of CAD/CAM ceramic crowns [abstract]. J Dent Res. 2001;81(spec iss A): Abstract 2640.

18. Plourde J, Harsono M, Fox L, et al. Marginal and Internal Fit of E4D CAD/CAM All-Ceramic Crowns. J Dent Res. 2011;90(spec iss A): Abstract 638.

19. Fasbinder DJ. Clinical performance of chairside CAD/CAM restorations. J Am Dent Assoc. 2006;137 suppl:22S-31S.

20. Martin N, Jedynakiewicz NM. Clinical performance of CEREC ceramic inlays: a systematic review. Dent Mater. 1999;15(1):54-61.

21. Otto T, De Nisco S. Computer-aided direct ceramic restorations: a 10-year prospective clinical study of Cerec CAD/CAM inlays and onlays. Int J Prosthodont. 2002;15(2):122-128.

22. Posselt A, Kerschbaum T. Longevity of 2328 chairside Cerec inlays and onlays. Int J Computr Dent. 2002;6(3):231-248.

23. Reiss B. Clinical results of CEREC inlays in a dental practice over a period of 18 years. Int J Comput Dent. 2006;9(1):11-22.

24. Scotti R, Cardelli P, Baldissara P, Monaco C. Clinical fitting of CAD/CAM zirconia single crowns generated from digital intraoral impressions based on active wavefront sampling. J Dent. 2011; Oct:17. [Epub ahead of print]

25. Ender A, Mehl A. Full arch scans: conventional versus digital impressions—an in-vitro study. Int J Comput Dent. 2011;l4(1):11-21.

26. Ogledzki M, Wenzel K, Doherty E, Kugel G. Accuracy of 3M-Brontes stereolithography models compared to plaster models [abstract]. J Dent Res. 2011;90(spec iss A): Abstract 1060.

27. Sorensen JA, Sorensen PN, Mizuro K. Marginal fidelity of crowns made with optical versus conventional impressions [abstract]. J Dent Res. 2009;88(spec iss A): Abstract 1599.

28. Fasbinder DJ, Neiva GF, Dennison JB, et al. Evaluation of zirconia crowns made from conventional and digital impressions [abstract]. J Dent Res. 2012;91(spec iss A): Abstract 644.