Introduction

Radiographic imaging using X-rays has been the primary diagnostic imaging technology used within the dental profession since its discovery in 1895. However, it is an ionising radiation with some detrimental effects to our health1 and has limitations, for example, superimposition of anatomical structures and poor contrast of soft-tissue. Radiographic imaging is, to a large extent, technology limited and has advanced rapidly, particularly in recent years with the development of digital imaging. X-rays are also part of the electromagnetic spectrum. The aim of this literature review was to revisit diagnostic imaging considering not just technological advancements, but also the entire electromagnetic spectrum (Fig. 1). The objective is to identify alternative imaging techniques, X-ray or otherwise, which may deliver complementary diagnosis, lower dose or better diagnostic efficacy than current techniques. Some of the options available are discussed below.

Figure 1
figure 1

The electromagnetic spectrum

Full size image

Method

A comprehensive review of four databases, that is, Medline, PubMed, Science Direct and Web of Knowledge, was undertaken. General and restorative dental terminologies were combined with the imaging options from the electromagnetic spectrum and Boolean Operators were applied. Exclusion criteria included articles of language other than English.

Results

Magnetic resonance imaging (wavelength ∼100 to 102 m)

This uses a static 0.1-8 Tesla magnetic field and a radio pulse, both of which are non-ionising radiation, to produce 2D or 3D digital images of tissues, for example, tumour staging, TMJ and intra-cranial lesions. The ability to see through bone avoids superimposition of structures and calcified and non-calcified tissues are illustrated (Fig. 2).2 Functional images which provide physiological information are possible with real time sequences and good resolution (50 μm × 50 μm × 350 μm). However, specialist equipment of considerable size and expense is required and a contrast medium is often used for soft tissue, for example, gadolinium, which requires intra-venous administration.3 Gadolinium is contra-indicated in patients with renal impairment. Capture of the image is lengthy and the machines are very noisy and claustrophobic. The magnetic field is a problem for people with metallic implants of any description and dental restorations can distort the image.

Figure 2: MRI of the head and neck.
figure 2

Evaluating a lipoma of the neck

Full size image

Terahertz Imaging (Wavelength of ∼10−4 To 10−3 m)

Also known as sub-millimetre or T-waves. Production of these waves is possible by exposure of zinc-telluride crystals to pulses of visible or infra-red light and detection is now feasible with advancements in photoconductive detectors.4 Terahertz pulses are strongly absorbed by water, and tissue penetration appears to be limited to the micron level.5 Image production is from the reflected wave (Fig. 3)6 and in mineralised tissue it is proving difficult to focus the reflected beam.7 They are non-ionising and can give good spatial resolution of 1 μm but, at that level, digital reconstruction of the image is complicated,8 whereas 40-60 μm is achievable.9 To our knowledge, imaging devices are not readily available to the profession and long-term health effects are unknown.

Figure 3: Cutting edge: terahertz images of a tooth.
figure 3

Reprinted with permission from Macmillan Publishers Ltd: Nature News, doi:10.1038/news020506-5, copyright (2002)

Full size image

Infra-red (Wavelengths of ∼ 8 × 10−7 TO 10−3 m)

Infra-red waves are non-ionising, non-invasive and non-destructive, enabling repeat images to be taken. Portable devices are available.

This radiation may be directed at the subject as, in the near-infra-red range of 1,310 nm, enamel is highly transparent compared to visible light, permitting the detection of demineralisation as a dark spot from increased scattering (Fig. 4).10,11 Fluorosis, stains and pigmentation are recognisable from demineralisation.11

Figure 4: NIR images of interproximal lesions.
figure 4

Reprinted from: Fried D, Featherstone J D B, Darling C L, Jones R S, Ngaotheppitak P, Bühler C M. Early caries imaging and monitoring with near-infra-red light. Dent Clin North Am 2005; 49: 771–793 with permission from Elsevier

Full size image

Alternatively, the natural infra-red emissivity of a subject can also be recorded with a thermal camera operating in the 9,000 nm to 12,000 nm range. The first thermogram was produced in 1840, and the 1950s saw the first medical use.12 Temperature differentials of 0.025 °C are achievable, producing colour contrasting images with pixel resolutions of 640 × 480. Occlusal caries has been associated with reduced temperature compared to its surroundings. Evaporation of water from the porous demineralised area is detectable with a thermal camera (Fig. 5),13 and correlation of the lesion depth or mineral loss with surface temperature has been demonstrated.14 Sensitivity of 58% and specificity of 83% for occlusal lesions reaching the dentine has been shown. Soft-tissue lesions, such as basal cell carcinoma, can also be observed.15

Figure 5: Thermal image of demineralised tooth.
figure 5

Reprinted from: Zakian C M, Taylor A M, Ellwood R P, Pretty I A. Occlusal caries detection by using thermal imaging. J Dent 2010; 38: 788–795 with permission from Elsevier

Full size image

However, the emissivity is a surface effect and there is low resolution of images acquired. Any fluid present (for example, saliva) will absorb the waves16 and, for the breathing subject, there is continuous fluctuation of air temperature and humidity, which hinder the accuracy of the readings. Attempts have been made to assess tooth vitality from the crown temperature but results have been inconclusive. Some studies demonstrate a higher temperature and quicker re-warm rate in vital than non-vital teeth,17,18,19 while others have demonstrated no detectable difference.20,21,22

Optical Coherence Tomography (Wavelengths of ∼10−7 m)

Waves in the near-infra-red are split then recombined, enabling a pattern from the interference and back-scattered waves to format a 2D image of the optical reflection23,24,25 and 3D real-time imaging is also achievable. It is a non-invasive, non-contact, non-ionising technique, with no biological effects to date. The chosen wavelength determines the depth of penetration and the resolution, which can reach 2.0 mm and 5-15 μm, respectively. Uses include assessment of oral soft-tissue lesions and caries-depth with polarisation26 (Fig. 6).24 Sensitivity for detection of squamous cell carcinoma from non-cancer tissue was reported at 0.931 and specificity was the same; for squamous cell carcinoma against other pathologies sensitivity was 0.931 with specificity of 0.973.25 However, the availability of the machine to the profession is a problem.27 The status of the tooth-surface needs consideration, as hydration of enamel affects signal intensity, which decreases when air is blown,28 as well as the structural orientation of dentine on the scattering of light.29

Figure 6: Cervical caries lesion (CL) viewed by OCT.
figure 6

E = enamel, D = dentine, DEJ = dentine enamel junction, G = gingivae. Reprinted from: Feldchtein F I, Gelikonov G V, Gelikonov V M et al. In vivo OCT imaging of hard and soft tissue of the oral cavity. Opt Express 1998; 3: 239–250 with permission from the Optical Society

Full size image

Laser Doppler (Wavelength ∼10−7 m)

Red30 or green31 light is used and tissue vascular supply is assessed by the Doppler effect.32 It is non-invasive, non-ionising with arbitrary units,33 which prevent comparison of successive readings. Future indications include assessment of grafts, osteomyelitis of bone and supporting-bone of dental implants.34 Unfortunately, the signal received can vary due to the location and angulation of the probe on the tooth35 and also due to the optical properties of the tooth.30 When investigating the pulpal vascular supply, there may be contamination of the signal from other vascular sources, for example, periodontal ligament,36 giving unreliable results. Its use is contra-indicated in heavily-restored dentitions.

Digital Fibre Optic Transillumination, DIAGNOdent and Quantitative Light-Induced Fluorescence (Wavelength ∼400 to 750 nm)

These diagnostic techniques draw on the visible spectrum for assessment of the mineralised coronal portion of the tooth. Fibre optic transillumination (FOTI) (Fig. 7) preceded the digital fibre optic transillumination (DiFOTI) which captures the transmitted photons from the light-source with a charge couple device (CCD), enabling the digital image to be displayed.37 Sensitivity of 14% and specificity of 95% for occlusal caries is reported, but proximal lesions result in 4% sensitivity and 100% specificity38 from FOTI. Quantitative assessment is possible from DiFOTI, with resolution of 43 pixels/mm.39 It is simple to use, non-ionising and non-invasive with the ability to provide real-time images, which are a great education tool. However, the depth of a lesion cannot be estimated and subgingival areas are not accessible.40

Figure 7
figure 7

FOTI image evaluating interproximal caries lesion, courtesy of A. F. Hall

Full size image

DIAGNOdent uses 655 nm red laser light to initiate fluorescence which is possibly of microbial origin,41 captured by a photocell delivering a numerical and acoustic signal.42 It is non-ionising and simple to use but active or arrested lesions are indistinguishable.43 Sensitivity and specificity for dentinal lesions have been cited as 0.75 and 0.96, respectively.44

Quantitative light fluorescence (QLF) uses light with a peak intensity of 370-410 nm to fluoresce the tooth and the emitted photons are captured by a CCD and digitised.45 Non-ionising, non-invasive, quick to use, but hydration of the tooth affects the results, with a stronger signal from dehydrated lesions due to increased scattering of the short-length photons.46 Accessibility can be a concern for interproximal surfaces and smooth surfaces can only be assessed to 500 μm4 (Fig. 8), with stains and white spot lesions appearing identical. Sensitivity of 0.68 and specificity of 0.70 for occlusal caries has been achieved.47

Figure 8
figure 8

QLF image evaluating caries lesions, courtesy of A. F. Hall

Full size image

Ultraviolet (UV) (Wavelength of 10−8 to 4 × 10−7 m)

UV light can produce fluorescence in enamel48 which, if demineralised, will lack fluorescence but it is difficult to separate caries from developmental defects.49 UV digital viewers and camcorders use 396 nm UV-rays and can give resolution of 640 × 480 pixels. The UV-rays may be reflected or absorbed in surface layers and aid forensic medicine with detection of bitemarks and bruises50 and the fluorescence properties identify dental materials.51 Basal cell carcinomas can be recognised as dark patches and oral squamous cell carcinoma may be discerned.52

It is non-ionising but the subject and operator need protection due to risk of cataracts and possible damage to DNA formation.15 Some subjects will be particularly sensitive to UV-rays and care is needed, for example, systemic lupus erythematous and xeroderma pigmentosa.53

X-rays (Wavelength ∼10−8 to 10−10 m)

X-rays whose absorption is dependent on the tissues' atomic number (Z) produce image contrast primarily due to the photoelectric effect. An energy range of 65 to 70 kV for intra-oral dental views, and 90 to 120 kV for cone beam computed tomography, is employed. This yields a variety of analogue or digital images from full-field or scanning sequences (Fig. 9).

Figure 9
figure 9

A variety of radiographic images: a) Extra-oral panoramic, b) Intra-oral bitewings, c) Intra-oral periapical, d) Axial CT

Full size image

Sensitivity of 0.95 and specificity of 0.83 are reported for enhanced digital images for proximal caries detection.47 Subtraction of digital images is also possible and, when monitoring lesions, may be very useful but the geometric reproducibility needs to be exemplary. There are numerous digital detector systems available54 which convert the radiation into an electric signal which may be wired or wireless, for example, solid state (CCD and CMOS) or photostimulable phosphor plates. Considerations such as patient-comfort, image quality (for example, spatial resolution and contrast), radiation dose, speed of image production and cross-infection risk influence the operator's choice. The detector may integrate the signal over time or record every signal event. Digital detectors do compare with analogue film for spatial resolution, achieving 20 lp/mm. However, X-rays are ionising with associated detrimental health-effects, that is, somatic deterministic and stochastic, genetic stochastic.1 The geometric accuracy may lead to errors in assessment of lesion-depths, such as caries.42 Guidelines for dental radiography are shown in Table 1.

Table 1 Guidelines for selection criteria in dental radiography
Full size table

Tomography

Conventional tomography can deliver the single-slice dental panoramic image but, with technological advances, multiple-slices of computed tomography (CT) can provide a 3D image from a CT scanner. However, the equipment is very expensive and occupies a large space. Dental cone beam CT (CBCT) – also known as digital volumetric tomography (DVT) – can use a vertically-positioned patient, reducing the space requirement compared to the horizontal CT scan. This is available and affordable to the general dental practitioner. Spatial resolution of between 0.07-0.4 mm55,56 with isotropic voxels is possible with CBCT. CT voxels may be anisotropic or isotropic. Multi-slice CT may deliver equivalent resolution from equivalent dose compared to CBCT56 but it has also been stated CBCT can deliver reduced effective dose (Table 2).57,58 Both CT and CBCT can have an adjustable field of view (FoV).

Table 2 Effective radiation doses for dental radiography
Full size table

CT has been used for assessing intracranial disease and damage following trauma to the head and neck, facial fractures, tumour-staging in the head and neck, implant-planning and investigating the TMJ1 but, with a possible reduction in effective radiation dose, CBCT may be more appropriate in some situations, for example, implant-planning. Other uses of CBCT may include: assessment of dento-alveolar pathology, maxillofacial surgery, orthodontics, nerve-position, endodontics, periodontics and general and forensic dentistry.59

Measurement accuracy of CBCT can be good (within 1 pixel longitudinally and up to 2.35 pixels horizontally60) with good geometric accuracy (mean deviations 0.13 ± 0.09 mm from three coordinate axes61). However, with increasing voxel size there can be a tendency to underestimate volumetric measurements which can become significant above 300 μm.62

Many manufacturers market CBCT machines, each with their own limitations, and there is a need to establish image-quality criteria, irrespective of machine.58

CBCT also has increased noise and scatter, leading to less soft-tissue contrast than CT.55 Additional limitations of CBCT include: arbitrary greyscale values,63 streak or star artefacts from metallic objects (for example, amalgam restorations, reducing diagnostic yield64) and motion artefacts. The motion may be unavoidable due to respiration and cardiac rhythm or to head-tremor65 and needs further research to enable correction during image reconstruction.

The scan settings (for example, FoV, resolution and X-ray parameters) determine effective dose66 which may vary by a factor of 4.6 to 5.2 from the lowest to highest dose67 and can influence whether sensitive organs are inside or outside the direct beam. This influences image quality and diagnostic potential of the CBCT image. A more standardised approach is needed to achieve the 'as low as reasonably achievable' dose.68 Copper filtration can reduce effective dose by 43% with adjusted kVp69 without loss of image quality. The use of collimation and patient protection has been suggested to minimise dose to structures such as the eyes and thyroid.70 However, the thyroid should not be in the primary beam.

For dental applications imaging high and medium-contrast tissues, CBCT may be the correct choice compared to CT but the profession needs to review CBCT guidelines regularly.

Future developments for X-rays

Phase contrast

On entry through tissue, X-rays refract like light waves through glass and the index of refraction deviates from 1, that is, unity:

n = 1 − δ − ίβ

n = index of refraction

δ = phase shift incorporating the refractive effects

ίβ = absorption of incident rays

This will result in a loss of coherence of the waves generating a shift in wave phase. The refractive proportion is greater than the absorption component normally used to assemble an image and if captured can exhibit detail not currently achievable (Figs 10 and 11).71,72,73 This can operate at the higher energy levels of X-rays, reducing exposure of the subject but achieving greater image contrast. The greatest contrast is seen with soft-tissues, not the mineralised tissue.73

Figure 10: Phase-contrast radiography – evaluating root morphology.
figure 10

Reprinted from: Sato A, Nakamoto N, Abe T, Fukushima Y, Tomaru Y, Sakata Y, et al. Preliminary Results of a Study Comparing Conventional Radiography with Phase-contrast Radiography for Assessing Root Morphology of Mandibular Third Molars. Dentomaxillofacial Radiology 2011; 40: 91–95.

Full size image
Figure 11: Post-mortem images of a chicken wing.
figure 11

a) Conventional transmission image, b) differential phase-contrast image, c) dark-field image. Reprinted from: Bech M, Jensen T H, Bunk O et al. Advanced contrast modalities for X-ray radiology: phase-contrast and dark-field imaging using a grating interferometer. Z Med Phys 2010; 20: 7–16 with permission from Elsevier

Full size image

Sophisticated equipment is required to produce the phase contrast of the waves, for example, a synchrotron, which was unacceptable clinically due to size and cost, and historically crystals split the beams to enable phase differences, but these were unstable and gave a small field of view. Micro-focus X-ray tubes are being successfully operated to generate the beam but also have limited field of view.74 X-ray interferometery is now achievable by exposing the beam to a series of gratings. One grating is located close to the wave-source (source-grating) before the subject, and two gratings are positioned after the subject. These are described as a phase-grating and absorption-grating, respectively. The phase-contrast is generated by the last two gratings which are strategically placed to produce a linear periodic fringe pattern which is aligned with the absorption grating placed in front of the detector.73 Changes in oscillations within the detector when the gratings are scanned are assessed according to subject. These signals are digitised to generate the phase contrast image for analysis.

Darkfield

X-ray grating interferometry, as described above, is actually a multimodal imaging technique and can be drawn on to fabricate not only the absorption and phase-contrast image but also the darkfield image.73 The darkfield image, as with visible light, is manufactured from the scattered X-rays. The image contrast is created by the small angle scattering of the waves generated from a conventional X-ray tube source. The detector, which is a technological achievement, has the capacity to detect the scattered waves and decipher the unscattered waves which can be removed. The greater the scatter, the greater the contrast of the image and this is seen with the mineralised tissue (Fig. 11c).73 Directional darkfield imaging has been demonstrated on a dry slice of tooth.75 This allows structural information smaller than the image resolution and maximises the different angles created by the scattered rays.

The detector will primarily register the perpendicular scattered waves in darkfield imaging but, in directional darkfield, the actual angle of the scattered wave is analysed as well. Dentine showed the strongest signal and the strength of the signal decreases with increasing distance from the pulp-chamber. This may correlate with the dentinal tubules, as enamel which is a reasonably homogenous mineralised tissue, generated little scatter signal. Whether this will actually happen with a vital tooth with dentinal fluid remains to be seen, as this equipment is not currently available to the profession. Is it feasible to hypothesise that selective demineralisation of enamel prisms in the early carious lesion would induce greater scattering enabling detection?

Future developments of three dimensional darkfield images require technological furtherance but, with the ability to capture the data to produce the absorption, phase-contrast and darkfield image, the potential improvement in diagnostic yield is vast.

Spectral

Initially proposed in 197676 spectral imaging exploits the energy spectrum of X-rays. Manipulation of X-rays has enabled the production of specific incident energy spectrums (spectral shaping) and detection of photons in specific energy bands (spectral imaging).77

An energy-dependent detector charts the charge release from incoming photons and, the higher the energy, the higher the weighting documented. This improves the resolution of the image. Counting of each individual photon in a set energy range is possible (Fig. 12)78 and eliminates any weighting for the higher energies, again increasing the image resolution.

Figure 12: Energy dependent spectral dental images.
figure 12

Reprinted from: Norlin B, Fröjdh C. Energy Dependence in Dental Imaging with Medipix2. In Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment. pp 19–23. Proceedings of the 6th International Workshop on Radiation Imaging Detectors, 2005.

Full size image

The scanning beam method is beneficial for the subject, as the majority of scatter is blocked. In addition, the actual number of photons produced interacting with the patient's tissue is reduced. Transmitted information needs to be accurately and efficiently detected by X-ray optics, capillary optics or an array of refractive X-ray lenses.

Comparison between photon-counting detectors (for example, Medipix1) and CCD used within dentistry for diagnostic imaging, demonstrated dose reductions without loss of resolution contrast.79 Further research revealed an increase in contrast of 18% is possible with the Medipix2.78

Characterisation of tissue-type (Fig. 13) and element enhancement for contrast studies80 from the energy of the detected photons (spectral imaging) is feasible. This is due to each tissue-type having its own signature because of its atomic number (Z).81,82 This enables optimising the incident beam energy for a chosen tissue (spectral shaping)83 which can reduce the dose to the patient from non-informative low- and high-energy photons.

Figure 13: Early results of spectral imaging from Medical Physics Department, Leeds Teaching Hospitals.
figure 13

1) Conventional X-ray, 2) spectral image (each colour highlights a tissue-type)

Full size image

The possible benefits of spectral imaging include:

  • Energy weighting

  • Dual energy subtraction

  • Identify tissue and quantify it.

This system is not currently commercially available to the dental profession.

Gamma rays (Wavelength ∼10−13 to 10−10 m)

A radioisotope, for example, technetium (99mTc) is obligatory, usually amalgamated with a pharmaceutical and administered intravenously or orally. The radioisotopes nuclei are unpredictably volatile and spontaneously disintegrate, producing gamma rays, making the patient the provenance of ionising radiation. A specialised gamma camera captures the emitted radiation to produce the image, for example, single photon emission computed tomography (2D – scintigraphy, slices – SPECT) or positron emission tomography (PET) which employs two photons from the annihilation radiation. A functional image can be generated, enabling all tissues to be assessed from one visit and recognising physiological change before anatomical change.84 However, resolution is poor.85 Anatomical location can improve accuracy by superimposition of a PET scan with a CT or MRI Scan.86,87

It is expensive, time-consuming and the images, unfortunately, are not disease-specific. The radiation dose is substantial following intravenous administration of radioisotope (5-7 mSv and, when combined with CT scanning, approximately 25 mSv).88 Dental restorations can produce artefacts89 and it is used for diagnosis of osteomyelitis, osteoblastic metastatic tumours, Paget's disease (Fig. 14) and salivary gland function.

Figure 14
figure 14

Single plane nuclear medicine image demonstrating Paget's disease affecting the mandible

Full size image

Ultrasound does not employ waves from the electromagnetic spectrum and is not discussed within the remit of this article.

Summary

The use of computer technology has enhanced imaging techniques, as shown with the development of the digital image and detector, for example, Medipix, along with the accessible stable beam sources for the production of phase contrast, darkfield and spectral imaging. Not all techniques are available to the clinician yet but they indicate what the future may hold. It must be remembered that technology may also be the limiting factor for any imaging technique. Signals may be received and converted to digital images but the monitor they are viewed on may not resolve adequately, with loss of diagnostic information.54 The format the images are stored in may lead to loss of data, as well. The correct environment and setting for viewing the images must be rigidly observed,90 which leads to operator capabilities. There needs to be quality control, not just for the acquisition of images, but also for those undertaking the reporting of the image.91

Conclusion

Non-ionising imaging techniques, for example, infra-red, are becoming more commensurate with traditional radiographic techniques as technology progresses, and need further exploration. Nevertheless, X-rays continue to be the leading diagnostic image for dentists and are being nurtured for improved diagnostic potential with elevated contrast and resolution with reduced doses, hopefully diminishing the potential health-risk to our patients.