The dental industry is undergoing rapid changes in digitalization in line with the Fourth Industrial Revolution. As a result, the use of computer-aided design/computer-aided manufacturing (CAD/CAM) in the design and fabrication of prosthetics is increasingly used to produce implant guides, mouth guards, dental die models, denture tops, and resins for temporary dentures using 3D printers.

3D printers are broadly categorized into FDM, SLA, and Digital Light Processing (DLP) types [1]. Among them, DLP-type 3D printers have the advantage of relatively high precision and excellent surface finish [2]. DLP 3D printers are widely used because they use a laser to cure a photocurable liquid resin layer by layer, making it easier to reproduce complex shapes and reduce time by producing multiple prints at once [3,4].

Previous studies on 3D-printed dental prostheses have focused on physical properties, changes in material properties, wear resistance, and fitness depending on the final curing time after printing [1,5]. Among them, fitness is one of the factors that affects the clinical survival rate of dental prostheses, and many studies have been conducted on the fitness of 3D printed dental prostheses and several methods have been introduced for its evaluation[5,6]. However, there is still a lack of comparative site-specific conformity studies on the slice thickness of temporary dental restorative resins produced by additive manufacturing (DLP) 3D printers.

Therefore, in this study, we aimed to evaluate the marginal and medial fitness of temporary dental restorative resins produced with different slice thicknesses using a DLP-type 3D printer through 3D analysis by scanner using silicon replica measurement and to use it as a clinical performance predictor through verification of clinically acceptable fitness figures.

1. Making a working model

The study was conducted as shown in Figure 1. A maxillary left first molar abutment tooth from an oral model (D51DP-500A-MF, Nissin, Japan) was used as the reference model. An impression was taken of the reference model using silicone impression material (Elite double 22, Zhermack, Germany) to create the mold (Figure 2).

Figure 1. Study design.

Figure 2. Mold fabrication.

To fabricate the working model, dental hard plaster (FujiRock, GC, Japan) was stirred and injected into the mold according to the manufacturer’s instructions, and the mold was removed after one hour of curing to make a total of 50 working models (Figure 3).

Figure 3. Creating a working model.

2. Making a temporary denture resin specimen

A model scanner (E2, 3shape, Denmark) was used to obtain scan data of the working model (Figure 4). The scanned working model data was used to design a temporary dental resin specimen using the CAD design program (3shape, Copenhagen K, Denmark), with a cement space setting of 50 μm (Figures 2-5).

Figure 4. Scanning the working model.

Figure 5. Designing a resin specimen for a temporary denture.

Once the design was completed, the file was divided into five groups according to the slice thickness of the DLP 3D printer (ASIGA MAX, Asiga, Australia): 0.01 mm (A), 0.025 mm (B), 0.075 mm (C), 0.1 mm (D), and 0.15 mm (E), and a total of 50 specimens, 10 for each group, were printed using a dedicated resin solution (MAZIC D TEMP, VERICOM, Korea) (Figure 6).

Figure 6. Temporary denture resin specimen as a function of slice thickness.

They were later cleaned using an ultrasonic cleaner (Twin Tornado, MEDIFIVE, Korea) and post-treated with a UV lamp (U102H, Graphy, Korea) for final curing according to the manufacturer’s instructions.

3. Measuring fitness

After injecting a light body silicone having high fluidity (PERFECT-F VPS, HAN DAE CHEMICAL, Korea) into the fabricated temporary dental resin specimen, it was uniformly applied to the working model and pressed with a finger until cured. Once cured, the specimen was carefully removed from the working model to avoid separating the silicone (Figure 7).

Figure 7. Silicone impression.

A total of 100 scans were acquired using the scanner for 50 working models with no treatment and 50 working models with the inner surface of the specimen replicated by a thin film of silicone. At this time, titanium dioxide spray (Easy Scan, Alphadent, Korea) was applied to obtain accurate scan data.

To ensure accurate merging and identical boundaries between the scan data of the working model and the scan data of the working model with the replicated inner surface of the specimen, unnecessary areas below the dental line were removed, merged, and evaluated using the CAD design program.

After verifying if the two scan data were aligned in the correct position, a virtual plane was set up to pass through the buccal cusps of the working model to measure the thickness of the silicon. The virtual plane formed was set at the same coordinates in all 50 merged scans, and the gap between the working model and the silicon was measured on the side cut by this virtual plane. The gap between the working model and silicone was measured at six locations. Each measurement site was designated as buccal margin (BM), buccal axial wall (BA), buccal cusp (BC), lingual cusp (LC), lingual axial wall (LA), and lingual margin (LM). Therefore, six gaps were measured on each side for a total of 300 gaps (Figure 8).

Figure 8. Merge scan data and fitness measurement.

4. Statistical analysis

Statistical analysis was performed by ANOVA and repeated measures ANOVA to examine the differences in temporary dental resin by measurement site according to the slice thickness of the DLP 3D printer and Tukey’s HSD for post-hoc testing. Statistical judgments were made at the 0.05 level of significance, and data processing for this study was analyzed using SPSS version 28.0 (release 23.0; IBM, Chicago, USA).

1. Descriptive statistics of temporary dental resin fitness

The analysis of the marginal and internal fitness by slice thickness of five groups of DLP 3D printers is shown in (Table 1).

Table 1 . Fitness measurement results of temporary dental resin (nunit: μm)

Slice Thickness	BM	BA	BC	LC	LA	LM	F (p)
	Mean (SD)
A	105.80 (21.99)	127.90 (18.17)	121.30 (25.51)	131.00 (17.58)	122.10 (18.77)	106.60 (15.62)	6.959* (.026)
B	115.30 (24.31)	145.30 (38.94)	135.20 (33.92)	149.00 (45.30)	129.90 (24.40)	97.90 (32.59)	5.613* (.041)
C	117.00 (16.21)	145.30 (35.85)	174.20 (44.22)	233.20 (85.17)	130.70 (25.22)	101.80 (23.79)	8.534* (.017)
D	118.80 (6.96)	157.70 (29.63)	215.30 (58.54)	244.00 (22.74)	142.10 (27.03)	107.40 (28.28)	32.198*** (.001)
E	119.10 (21.22)	177.70 (45.10)	225.70 (63.93)	245.10 (102.07)	179.40 (53.54)	116.60 (22.72)	5.249* (.046)
F (p)	0.815 (.522)	2.805* (.037)	9.605*** (.000)	7.562*** (.000)	4.976** (.002)	0.776 (.547)

By repeat measure ANOVA (post-test Tukey’s HSD) for three or more groups.

When looking at the mean values of the descriptive statistics of gaps by slice thickness for DLP 3D printers, there was no significant difference for BM (p=0.522) and LM (p=0.547). There was a significant difference in the case of BA (p<0.05), BC (p<0.001), LC (p<0.001), and LA (p<0.01).

2. Post-hoc testing of temporary dental resin fitness

When examined by slice thickness, 0.01 ([<2] p<0.05), 0.025 ([<4] p<0.05), 0.075 ([<6] p<0.05), 0.1 (p<0.001), and 0.15 (p<0.05) all showed significant differences. Post-hoc Tukey’s HSD test of differences by slice thickness showed that LC was the highest in the 0.01 and 0.15 groups, while LM and BM were relatively low. In the 0.025 group, LC and BA were the highest, while LM showed relatively low values. The 0.075 and 0.1 groups had the highest LC, while LM showed relatively low values.

When broken down by measurement site, BM, LM (p>0.05), except for BA (p<0.05), BC (p<0.001), LC (p<0.001), and LA (p<0.01) all showed significant differences. The post-hoc Tukey’s HSD test for differences by measurement site showed that BA and LA had the highest values in the 0.15 group, while the 0.01 group showed relatively low values. BC and LC were highest in the 0.15 and 0.1 groups, while the 0.01 group showed relatively low values (Table 2).

Table 2 . Tukey’s HSD post-hoc test

		Tukey’s HSD	F (p)
Slice Thickness (mm)	0.01	LC>BA>LA, BC>LM, BM	6.959* (.026)
	0.025	LC, BA>BC>LA>BM>LM	5.613* (.041)
	0.075	LC>BC>BA>LA>BM>LM	8.534* (.017)
	0.1	LC>BC>LA, BA>BM>LM	32.198*** (.001)
	0.15	LC>BC>LA, BA>BM, LM	5.249* (.046)
Measurement site (μm)	BM	-	.815 (.522)
	BA	0.15>0.1>0.025=0.075>0.01	2.805* (.037)
	BC	0.15, 0.1>0.075>0.025>0.01	9.605*** (.000)
	LC	0.15, 0.1>0.075>0.025>0.01	7.562*** (.000)
	LA	0.15>0.1>0.075, 0.025>0.01	4.976** (.002)
	LM	-	.776 (.547)

*p<0.05, **p<0.01, ***p<0.001.

3. Repeated measures statistical analysis of temporary dental resin fitness

The final results of the measurements were evaluated for the significance of each of the measurement sites and slice thickness, as well as the significance of the combined effect of measurement sites and slice thickness, which showed a significant difference by slice thickness (F=17.181, p<0.001). Significant differences were observed by measurement site (F=41.815, p<0.001). For the combined effect of Slice Thickness and Timing of Measurement Site (F=3.580, p=0.001), there was a significant combination effect (Table 3) (Figure 9).

Table 3 . Statistical analysis of interval repeat measurement by slice thickness

Source	Type III Sum of Squares	df	Mean Square	F
Slice Thickness	139756.380	4	34939.095	17.181*
Measurement site	316981.387	5	63396.277	41.815*
Slice Thickness *Measurement site	108542.980	20	5427.149	3.580*

*p<0.001

Figure 9. Graph of area measurement results by slice thickness.

4. Measurement results graph

In this study, ten temporary dental resin specimens of five groups with different slice thicknesses were produced using a DLP-type 3D printer, and the measurement results of marginal and medial fitness were evaluated by 3D analysis through a model scanner using silicone replica measurements, and the following results were obtained by ANOVA.

1. According to the slice thickness of the five groups, there were differences in fitness for each measurement site. For the axial walls and cusps, except for margins, the thinner the slice thickness, the more improvement in fitness was observed.

2. There was no significant difference in the temporary denture resin margins of the slice thickness, and all margins of the five groups showed excellent results below 120 μm.

These findings suggest that the temporary dental restorative resin produced by the DLP method used in this study shows positive performance predictions for clinical applications. Therefore, further research is needed to improve the clinical utility of temporary dental restorations produced by DLP 3D printers.

No potential conflict of interest relevant to this article was reported.