The validation of Quantitative light-induced fluorescence to quantify acid erosion of human enamel

Archives of Oral Biology (2004) 49, 285—294

The validation of Quantitative light-induced fluorescence to quantify acid erosion of human enamel I.A. Prettya,*, W.M. Edgarb, S.M. Highamb a

Unit of Prosthodontics, Department of Restorative Dentistry, Turner Dental School, The University of Manchester, Higher Cambridge Street, Manchester M15 6FH, UK b Department of Clinical Dental Sciences, The University of Liverpool, Liverpool, UK Accepted 19 November 2003

KEYWORDS Dental erosion; Quantification; QLF; Longitudinal; In vitro

Summary Objective: The purpose of this study was to validate the Quantitative lightinduced fluorescence (QLF) device against transverse microradiography (TMR) with regard to the quantification of enamel erosion in vitro. Design: Longitudinal in vitro. Methods: Thirty previously extracted, caries free, human premolars were selected and prepared by gentle pumicing and coating in an acid-resistant nail-varnish save for an exposed window on the buccal surface. QLF baseline images were taken and the teeth then exposed to an erosive solution, 0.1% citric acid (pH 2.74). Teeth were removed at 30 min intervals, air-dried and QLF images taken. At this time one tooth was randomly selected, removed from solution and sectioned through the lesion at three sites. The polished sample (100 mm) was subjected to TMR and analysed for erosive mineral loss using proprietary software, with the DZ values noted. QLF images were analysed by a blinded examiner with DF and DQ values recorded. Data were entered into SPSS and the correlation between the DZ and DF, and DZ and DQ values calculated. Results: A wide range of erosive lesions was produced, with a steady increase in both DZ and DF over time; DZ (24.0 (S.D. 1.2)—6114.3 (S.D. 1177.57)); DF (1.8—11.2), DQ (2.5—202.6). The results were scatter plotted and a regression line calculated. A positive correlation between DZ and DF of 0.91 was found, and for DZ and DQ; 0.87. Conclusions: The ability for QLF to detect and longitudinally monitor in vitro erosion has been shown. The strong positive correlation of DF with DZ suggests that percentage fluorescence loss as measured by QLF could be of great value in the development of a nondestructive, longitudinal tool for use in vitro, in situ and possibly in vivo. ß 2004 Elsevier Ltd. All rights reserved.

Introduction Enamel erosion

*

Corresponding author. Present address: University Dental Hospital of Manchester, Higher Cambridge Street, Manchester M15 6FH. Tel.: þ44-161-275-6849; fax: þ44-161-275-6640. E-mail address: [email protected] (I.A. Pretty).

As caries incidence decreases an increased interest in other forms of tooth surface loss including attrition, abrasion and erosion has developed.1 A significant cause of enamel loss, especially in younger individuals, erosion has a wide and complex aetiology. Erosion is found when tooth surfaces are

0003–9969/$ — see front matter ß 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.archoralbio.2003.11.008

286

I.A. Pretty et al.

exposed to frequent contact with acids that can be separated into intrinsic or extrinsic acids.1 Examples of intrinsic include gastric reflux disease,2,3 chronic alcoholism,4 certain eating disorders,5—7 or rumination while extrinsic sources include fruit drinks,8 alcoholic drinks, citric fruits or acidic work environments.9,10 Numerous studies have been published that have tested, in vitro and in situ the erosive potential of products such as fruit and carbonated drinks,11,12 alchopops,13 dietary acids, medicines,14 and mouthrinses.15

Measurement of erosion Several methods of detection and measurement have been employed in these studies, including profilometry,15,16 ultrasonification,17 calcium 18 loss, photographic analysis,19 scanning electron microscopy,12 photography, surface microhardness,12 clinical indices,20 analysis of sequential dental study casts and transverse microradiography (TMR).21 Many of these techniques are either destructive, and thus limiting their use to in vitro work, or provide information on surface effect only, without considering any subsurface mineral loss. The ability to measure erosion longitudinally is difficult with destructive techniques such as TMR, and so data can only be obtained at one time interval during a study. This may result in the loss of important time based data, for example, erosion by proprietary mouthwashes was demonstrated using TMR, however a longitudinal technique showed that this only happened after 14 h of continuous use.22 Eroded lesions (at least in vitro lesions) are comprised of two distinct characterising areas; the first is the crater formed by actual loss of tooth surface. Directly beneath this crater is the second layer of

softened, demineralised enamel that, given sufficient time and erosive assault (possibly augmented by attrition or abrasion) will be lost, contributing to further crater formation and a new softened surface. The first event, therefore, in the development of any eroded lesion, is surface softening. Techniques such as microhardness monitoring will measure only the surface softening area of the lesion and are unable to detect or quantify the cratered area as they have no reference to the original, sound surface. Photographic and profilometric techniques, and those which involve the use of clinical indices or monitoring of dental study casts, describe only the crater formation, i.e. that amount of tooth substance that has been lost, and cannot measure the softened, demineralised surface layer lying at the base of the crater. It is possible to combine measurement systems, for example by using profilometry to establish crater morphology and subsequently microhardness the lesion’s properties can be elucidated, however these are time consuming methods and can only be used within in vitro or in situ models. While not universally accepted, TMR has been shown to be an acceptable gold standard for the quantification of eroded lesions from in situ/in vitro experiments23 (see Fig. 1). The transverse sections through the lesion enable a visualisation of the forming crater and also the softened enamel laying at the base of the lesion. By analysing each of these elements a truer picture of the lesion is obtained. TMR is, however, a time consuming and destructive technique that requires thin, planoparallel polished enamel slabs in order to ensure accurate measurement. With TMR there is also a risk of ‘selection bias’. As sections through the lesion are taken, and, inevitably, some of the lesion is lost within the cutting process, it is impossible to map the entire

Figure 1 TMR Image of an in vitro eroded lesion in enamel.

The validation of Quantitative light-induced fluorescence to quantify acid erosion of human enamel 287

lesion. For example, if one is using the maximal crater depth as the reportable value in a study, in reality this would be only the maximal crater depth of recovered slices. It is impossible to know what the aspects of the lesion were lost due the preparation process. This selection effect may however, only be of minimal importance when assessing artificially created eroded lesions that tend to be highly homogeneous in their morphology as compared to demineralised lesions which have been shown to be more heterogeneous.24 There is therefore a need for a measurement system that can detect both components of the eroded lesion, that can longitudinally monitor the erosive process and that is non-destructive, objective and expedient. The ability to use such a system in both laboratory and clinical experimentation would be of great use.

QLF Quantitative light-induced fluorescence (QLF) has been employed in the detection of early demineralisation from enamel and, with the addition of a fluorescent dye, dentine.25—28 Numerous studies have investigated the technique’s use in the detection of early caries in permanent teeth, adjacent to restorations and orthodontic brackets, in primary teeth, and in clinical trials of dentifrice products.25,29—35 Recently, there have been a number of researchers who have suggested that QLF technology can be used in a number of other applications including the detection of failing fissure sealants, tooth whitening and planimetric plaque measurements.25,31,36 Ku ¨hnisch et al. were the first to suggest that QLF could be employed as a device to monitor erosion progression in vitro.37

Objective The objective of this study was to determine if QLF could measure in vitro enamel erosion longitudinally. By comparing the data from the QLF device with TMR, a method previously demonstrated to measure both surface softening and crater depth (a gold standard), an assessment of validity could be conducted.

Methods and materials Sample preparation Thirty previously extracted human premolars were selected based upon their lack of enamel defect, caries or extraction damage on their buccal surface

as determined visually. The teeth had been stored in distilled water with thymol. Each was gently cleaned with pumice and wet-and-dry paper and coated with transparent acid-resistant varnish save for an exposed window of approximately 5 mm 5 mm. Baseline QLF images of each tooth were taken using a laboratory jack to standardise the position of the QLF camera to the specimen. The prepared teeth were subsequently mounted on glass rods using dental wax and suspended in a buffered citric acid solution (pH 2.74) with gentle agitation.

Longitudinal monitoring At 30 min intervals all the teeth were removed from the solution, rinsed in distilled water and allowed to bench dry for 2 h to enable complete dehydration of the lesion. QLF images were then taken and the resultant images saved to the PC’s hard drive for later analysis. One tooth at each time interval was randomly selected and removed from the sample and retained for subsequent TMR. This was repeated 30 times over the 15 h exposure period. Samples were stored in distilled water between exposures.

TMR analysis At the conclusion of the erosive process the retained teeth, each representing an increasing exposure to the erosive solution, were subjected to TMR. Each sample was sectioned (approximately 250 mm) using a water-cooled diamond saw (Well, Walter Ebner, Switzerland) through the eroded lesion. Three sections from each tooth were obtained, each bearing an eroded area and a sound enamel reference for reconstruction. These were mounted on to custom brass anvils with varnish and polished using a diamond disc to give planoparallel specimens of 100 mm thickness. The fine sections were then mounted on to a microradiographic plate-holder with an aluminium stepwedge (25 mm steps). Kodak high-resolution plates (type 1A) were employed with a 15 min exposure using a Cu Ka Xray source (Philips B.V., The Netherlands) operating at 25 kV and 10 mA at a focus-specimen distance of 30 cm. Plates were developed using Kodak brand materials following manufacturers’ instructions. Following developing the microradiographs were analysed using a Leica DMRB microscope (Leica, Germany) with image capture via a CCD video camera (Sony, Japan) connected to a PC. TMRW v.1.22 (Inspektor Research Systems BV, The Netherlands) was used to determine the integrated mineral loss (vol.% mm, DZ), lesion depth (mm) and width (mm) using a two-stage analysis procedure. Values of DZ were recorded and entered in SPSS.

288

I.A. Pretty et al.

QLF analysis QLF analyses were conducted by a blinded single examiner following a set of predetermined rules38 for patch placement using QLF Version 1.99 (Inspektor Research Systems BV, The Netherlands). Values for DF (percentage fluorescence loss) at the 5% threshold level and DQ (DF integrated by lesion area in mm2) were calculated and entered into SPSS. Correlations between DZ and DQ and DZ and DF were performed.

Results The DZ values for each of the eroded samples are shown in Fig. 2. TMR images showed that successive teeth presented craters of increased depth with an area of mineral loss beneath. DZ values increased with increased exposure to the erosive solution. DZ values (mean, S.D.) ranged from 25.0 (1.2) to 6114.3 (1177.57). The DQ values at the 5% threshold for each of the samples analysed are shown in Fig. 3. There was no differences detected when pre-varnished and baseline images were compared. QLF images demonstrated a successive loss of fluorescence in the exposed enamel window representing increasing mineral loss. DQ values ranged from 3.0 (0.0) to 1177.23. No standard deviation was available for

this final value as only one tooth remained in the experiment at the 15 h stage. The DF values for each of the eroded samples are shown in Fig. 4. QLF images DF values increased with increased exposure to the erosive solution. DF values ranged from 1.8 (0.5) to 11.2. Again, no standard deviation was available for this final value as only one tooth remained in the experiment at the 15 h stage. Values of both DFx  1 and DZ, and DQx  1 and DZ were scatter plotted against each other and a regression line calculated, Fig. 5, showing a positive correlation. Pearson’s correlation coefficient was calculated and r ¼ 0:91 when DQx  1 was compared to DZ and r ¼ 0:87 when DFx  1 and DZ were compared.

Discussion The question of validity is ‘‘Is the device measuring what it claims to measure?’’.39 In this study, the measurement is that of mineral loss from enamel during the erosive process. By comparing a novel device, QLF, with an established gold standard, TMR, it is possible to compare the data through a correlation coefficient and determine their agreement. The decision to use TMR as the gold standard was based upon its ability to quantify both the absolute loss of tooth surface (the crater) and also

8000

7000

6000

Delta Z

5000

4000

3000

2000

1000

0 1

2

3

4

5

6

7

8 9

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Sample Number (Increasing Time) (30 minute intervals)

Figure 2 DZ values with increasing exposure to erosive challenge. Error bars are standard deviations.

The validation of Quantitative light-induced fluorescence to quantify acid erosion of human enamel 289

1600

1400

1200

Delta Q

1000

800

600

400

200

0 1

2

3

4

5

6

7

8

9

10 11 12

13 14 15 16 17 18

19 20

21 22 23 24 25 26

27 28 29 30

Sample Number (Increasing Time) (30 minute intervals)

Figure 3 DQ values with increasing exposure to erosive challenge. Error bars are standard deviations.

14

12

10

Delta F

8

6

4

2

0 1

2

3

4

5

6

7

8

9

10

11 12

13 14

15 16

17 18 19 20

21 22 23

24 25

26 27 28 29

Sample Number (Increasing Time) (30 minute intervals)

Figure 4

DF values with increasing exposure to erosive challenge. Error bars are standard deviations.

30

290

I.A. Pretty et al.

6000

5000

∆Q vs ∆Z Regression line (0.91)

4000

TMR ∆ Z

3000

2000

1000

0

-1000

-2000 0

2

4

6

8

10

12

14

QLF ∆F (x-1)

Figure 5

Scatterplot of DQ vs. DZ values with regression line.

the pre-loss mineral dissolution below this (surface softening).23 In order to ensure that a validation comparison is scientifically sound, a wide spread of possible values is required. Figs. 2 and 3 illustrate that the samples within this experiment achieved this requirement, and also serve to demonstrate the predictable nature of erosion in in vitro samples, with an increase in the two measurements over time.40 The correlation coefficient of 0.91 represents a strong association between the two measurements, and is similar to coefficients of QLF compared to TMR for the assessment of early carious lesions in permanent teeth, r ¼ 0:98,41 and deciduous teeth, r ¼ 0:96.32 The ability to relate any given DF value with a DZ value using the scatter plot enables the severity of the lesion to be established with ease. The longitudinal ability of the QLF device enables the elucidation of the erosive process over time, as demonstrated in Fig. 6, which shows the erosive lesion on sample 14 after each of the 30 min immersions in citric acid. There are several theories that have been suggested to explain the decrease in auto-fluorescence seen in incipient lesions of enamel under QLF conditions.42 These have been well described in the literature42 but are briefly summarised here:

(a) a greater degree of light scattering occurs within demineralised enamel than that in sound enamel. The light path in the lesion is therefore shorter and the light absorption per volume is less resulting in lower fluorescence43; (b) the light scattering occurring within the demineralised enamel prevents further light penetration to and from the underlying fluorescing dentine; (c) the demineralisation process causes a change in the molecular environment of the chromophore causing a decrease in fluorescence44; (d) the demineralisation process removes the proteininc chromophores. Recent work employing a Monte Carlo simulation of the optical properties of enamel during the QLF process suggests that it is light-scattering that is the most plausible explanation for the mechanism of QLF.45 The fluorescent nature of teeth is thought by most investigators to be due the underlying dentine and most particularly the dentinal-enamel junction (DEJ). Recently, however, a study has offered evidence to suggest that (in the case of bovine enamel, at least) the dentine, and indeed the DEJ, are not required for accurate QLF assessment

The validation of Quantitative light-induced fluorescence to quantify acid erosion of human enamel 291

Figure 6

Example of the QLF detection of acid erosion over time.

of demineralisation.46 However, this is in contrast with a number of other studies25 and further research is required before the mechanistics of auto-fluorescence are fully understood. The decrease in the fluorescence seen in the longitudinal monitoring of erosive samples represents a possible paradox for the functioning of QLF. It has been noted that the eroded lesion (at least in vitro) is comprised of two separate entities, the crater and the softened surface.47 The quantification of the softened surface is explainable by any number of the possible theories for loss of fluorescence, however, it appears that, due to the considerable correlation between QLF and TMR, that the crater is influencing the appearance of the lesion. How can the QLF device measure mineral loss from tissue that is no longer present?

Further complications arise from the evidence on the impact of enamel thickness provided by Ando et al.48 This study found that the thinner the enamel the greater the fluorescence using an in vitro design employing enamel and dentine cores with artificial caries.48 As the crater is formed during the prolonged erosive challenge within the current study, the enamel in the eroded lesion is thinner than in the sound areas. Given that the softened surface will cause a loss in fluorescence this may be mediated by a small increase due to the proximity of the newly formed surface to the DEJ. The dynamic erosive process, as seen in vitro, results in an increasing amount of mineral loss from a softened surface which, at a critical point, is then lost and crater formation begins, followed by another surface softening and so on. This process

292

I.A. Pretty et al.

has been confirmed by studies employing ultrasonication following lesion production.17 This process creates a wall of sound enamel surrounding the eroded crater and its demineralised base, which alters in its mineralisation status during the erosive process as tooth substance is lost.23 Could these walls may be responsible for a loss of fluorescence in addition to that already contributed by the demineralised surface? Several hypotheses could explain how this may happen. If we consider that the QLF method is explained best by the scattering effect of carious enamel preventing penetration of the excitation energy to the underlying dentine, based upon the Monte Carlo simulations45,49 and Ando et al.’s48

work on enamel thickness, then the following may account for the results of the current work: (a) that the walls create a shadowing effect, preventing excitation energy reaching the enamel surface; (b) that the walls hinder the release of the fluorescing light due to further scattering effects; (c) a combination of these. Fig. 7 illustrates the suggested model and explains the apparent contradiction between the data from this study and the conclusions reached by Ando et al.48 from their enamel thickness data. The methodology employed within that experiment did

Enamel (sound) Dentine / enamel junction Demineralised enamel Fluorescing chromophores Excitation light (blue) Fluorescence (yellow)

(b) Crater only – arrested eroded lesion

(d) Ando et al – initial images; i) thick enamel, ii) thin enamel

(a) Classic subsurface lesion

(c) Eroded lesion, crater and surface softening

(e) Ando et al – demineralised images; i) thick enamel, ii) thin enamel

(f) Current study. While the crater decreases enamel thickness, the wall effect will i) disrupt excitation light and ii) hinder escape of fluorescence

Figure 7 Proposed model for the measurement of erosion by QLF.

The validation of Quantitative light-induced fluorescence to quantify acid erosion of human enamel 293

not produce walls of sound enamel as the entire surface of the enamel core was demineralised and the fluorescent difference calculated by comparison to an image acquired prior to the demineralisation process. This in vitro study has employed an aggressive erosion challenge that is unlikely to occur in vivo. This is justified as the purpose of the design was to obtain validation data rather than assess the mechanics of the erosive process itself. TMR images demonstrate that the erosive lesions obtained were of the normal morphology. Further work on the impact of saliva exposure and remineralisation therapies on the appearance of eroded lesions under QLF conditions is required.

Conclusion The QLF device has been shown to effectively quantify the erosive process in vitro when compared to TMR. Further research is required to determine the potential for QLF to detect enamel erosion in vivo, enamel erosion from deciduous teeth and also dentinal erosion of root surfaces.

References 1. Maupome G, Ray JM. Structured review of enamel erosion literature (1980—1998): a critical appraisal of experimental. Oral Dis 2000;6:197—207. 2. Aine L, Baer M, Maki M. Dental erosions caused by gastroesophageal reflux disease in children. ASDC J Dent Child 1993;60:210—4. 3. Bartlett DW, Evans DF, Smith BG. The relationship between gastro-oesophageal reflux disease and dental erosion. J Oral Rehabil 1996;23:289—97. 4. Smith BG, Robb ND. Dental erosion in patients with chronic alcoholism. J Dent 1989;17:219—21. 5. Roberts MW, Li SH. Oral findings in anorexia nervosa and bulimia nervosa: a study of 47 cases. J Am Dent Assoc 1987; 115:407—10. 6. Anderson L, Shaw JM, McCargar L. Physiological effects of bulimia nervosa on the gastrointestinal tract. Can J Gastroenterol 1997;11:451—9. 7. Ohrn R, Enzell K, Angmar-Mansson B. Oral status of 81 subjects with eating disorders. Eur J Oral Sci 1999;107: 157—63. 8. Jarvinen VK, Rytomaa II, Heinonen OP. Risk factors in dental erosion. J Dent Res 1991;70:942—7. 9. Skogedal O, Silness J, Tangerud T, Laegreid O, Gilhuus-Moe O. Pilot study on dental erosion in a Norwegian electrolytic zinc factory. Community Dent Oral Epidemiol 1977;5:248—51. 10. Pretty IA, Addy LD. Associated post-mortem dental findings as an aid to personal identification. Sci Justice 2002;42: 65—74. 11. Lussi A, Kohler N, Zero D, Schaffner M, Megert B. A comparison of the erosive potential of different beverages in primary and permanent teeth using an in vitro model. Eur J Oral Sci 2000;108:110—4.

12. Lussi A, Hellwig E. Erosive potential of oral care products. Caries Res 2001;35:52—6. 13. O’Sullivan EA, Curzon ME. Dental erosion associated with the use of ‘alcopop’–—a case report. Br Dent J 1998;184: 594—6. 14. Giunta JL. Dental erosion resulting from chewable vitamin C tablets. J Am Dent Assoc 1983;107:253—6. 15. Pontefract H, Hughes J, Kemp K, Yates R, Newcombe RG, Addy M. The erosive effects of some mouthrinses on enamel. A study in situ. J Clin Periodontol 2001;28:319—24. 16. West NX, Hughes JA, Addy M. The effect of pH on the erosion of dentine and enamel by dietary acids in vitro. J Oral Rehabil 2001;28:860—4. 17. Eisenburger M, Hughes J, West NX, Jandt KD, Addy M. Ultrasonication as a method to study enamel demineralisation during acid erosion. Caries Res 2000;34:289—94. 18. Larsen MJ, Richards A. Fluoride is unable to reduce dental erosion from soft drinks. Caries Res 2002;36:75—80. 19. Al-Malik MI, Holt RD, Bedi R. Clinical and photographic assessment of erosion in 2—5-year-old children in Saudi Arabia. Community Dent Health 2001;18:232—5. 20. Johansson AK, Johansson A, Birkhed D, Omar R, Baghdadi S, Carlsson GE. Dental erosion, soft-drink intake, and oral health in young Saudi men, and the development of a system for assessing erosive anterior tooth wear. Acta Odontol Scand 1996;54:369—78. 21. Hall AF, Sadler JP, Strang R, de Josselin de Jong E, Foye RH, Creanor SL. Application of transverse microradiography for measurement of mineral loss by acid erosion. Adv Dent Res 1997;11:420—5. 22. Pretty IA, Edgar WM, Higham SM. The erosive potential of commercially available mouthrinses as measured by QLF. J Dent 2003;31:313—9. 23. Amaechi BT, Higham SM, Edgar WM. Use of transverse microradiography to quantify mineral loss by erosion in bovine enamel. Caries Res 1998;32:351—6. 24. ten Cate JM, Exterkate RA. Use of the single-section technique in caries research. Caries Res 1986;20:525—8. 25. Amaechi BT, Higham SM. Quantitative light-induced fluorescence: a potential tool for general dental assessment. J Biomed Opt 2002;7:7—13. 26. Ando M, Hall AF, Eckert GJ, Schemehorn BR, Analoui M, Stookey GK. Relative ability of laser fluorescence techniques to quantitate early mineral loss in vitro. Caries Res 1997;31:125—31. 27. de Josselin de Jong E, Sundstrom F, Westerling H, Tranaeus S, ten Bosch JJ, Angmar-Mansson B. A new method for in vivo quantification of changes in initial enamel caries with laser fluorescence. Caries Res 1995;29:2—7. 28. Pretty IA, Ingram G, Edgar WM, Higham SM. The use of fluorescein to monitor de- and re-mineralisation of in vitro root caries. J Oral Rehabil 2003;30(12):1151—6. 29. Lagerweij M, Beiswanger AJ, van der Veen M, Ferreira Zandona AG, Stookey G. Interexaminer variability of analysis of QLF images obtained from a patient. J Dent Res 1998;77:713. 30. Pretty IA, Edgar WM, Higham SM. Detection of in vitro demineralisation of primary teeth using quantitative lightinduced fluorescence (QLF). Int J Paediatr Dent 2002;12: 158—67. 31. Amaechi BT, Higham SM. Development of a quantitative method to monitor the effect of a tooth whitening agent. J Clin Dent 2002;13:100—3. 32. Ando M, van der Veen M, Schemehorn BR, Stookey G. A comparison of quantitative light-induced florescence (QLF) on white-spots in permanent and deciduous enamel. Caries Res 1997;31:324.

294

33. Ando M, van Der Veen MH, Schemehorn BR, Stookey GK. Comparative study to quantify demineralised enamel in deciduous and permanent teeth using laser- and light-induced fluorescence techniques. Caries Res 2001;35:464—70. 34. Tranaeus S, Al-Khateeb S, Bjorkman S, Twetman S, AngmarMansson B. A six-month clinical study with QLF for comparison of F varnish and professional tooth cleaning in caries-active children. Caries Res 1999;32:291. 35. van der Veen MH, de Josselin de Jong E. Application of quantitative light-induced fluorescence for assessing early caries lesions. Monogr Oral Sci 2000;17:144—62. 36. Pretty IA, Edgar WM, Higham SM. The use of QLF to quantify in vitro whitening in a product testing model. Br Dent J 2001;191:566—9. 37. Kuhnisch J, Heinrich-Weltzien R, Rohrig A, van der Veen M, Stosser L. Monitoring of erosive lesions with quantitative light-induced fluorescence. Caries Res 2001;35:295. 38. Pretty IA, Hall AF, Smith PW, Edgar WM, Higham SM. The intraand inter-examiner reliability of quantitative light-induced fluorescence (QLF) analyses. Br Dent J 2002;193:105—9. 39. Brunette DM. Critical thinking: understanding and evaluating dental research. London: Quintessence Publication Co.; 1996. 40. Azzopardi A, Bartlett DW, Watson TF, Sherriff M. The measurement and prevention of erosion and abrasion. J Dent 2001;29:395—400. 41. Ferreira Zandona AG, Isaacs RL, van der Veen M, Eckert G, Stookey G. Comparison between light-induced fluorescence

I.A. Pretty et al.

42.

43.

44.

45.

46.

47. 48.

49.

and clinical evaluations for caries detection. Caries Res 1998;32:296. Angmar-Mansson B, ten Bosch JJ. Quantitative light-induced fluorescence (QLF): a method for assessment of incipient caries lesions. Dentomaxillofac Radiol 2001;30:298—307. ten Bosch JJ. Light scattering and related methods. In: Stookey GK, editor. Early detection of dental caries. Proceedings of the 1st Annual Indiana Conference, 1996. Indianapolis: Indiana University; 1996. p. 81—90. Spitzer D, ten Bosch JJ. Luminescence quantum yields of sound and carious dental enamel. Calcif Tissue Res 1977;24: 249—51. van der Veen MH, Ando M, Stookey GK, de Josselin de Jong E. A Monte Carlo simulation of the influence of sound enamel scattering coefficient on lesion visibility in lightinduced fluorescence. Caries Res 2002;36:10—8. Rousseau C, Vaidya S, Creanor SL, et al. The effect of dentine on fluorescence measurements of enamel lesions in vitro. Caries Res 2002;36:381—5. Amaechi BT, Higham SM. In vitro remineralisation of eroded enamel lesions by saliva. J Dent 2001;29:371—6. Ando M, Schemehorn BR, Eckert GJ, Zero DT, Stookey GK. Influence of enamel thickness on quantification of mineral loss in enamel using laser-induced fluorescence. Caries Res 2003;37:24—8. Ko CC, Tantbirojn D, Wang T, Douglas WH. Optical scattering power for characterisation of mineral loss. J Dent Res 2000;79:1584—9.