All the procedures of forming an implant cavity were unified as follows according to protocols. Firstly, the implant cavity was constructed by using a 2.0-mm-depth drill after marking the implant site using a guide drill. Secondly, a 3.1-mm pilot drill and a 3.1-mm twist drill were used to form a cavity for an implant with 3.8-mm diameter, and a 4.3-mm pilot drill and a 4.3-mm twist drill were used to form a cavity for an implant with 5.0-mm diameter. Of note, this experiment was performed under a non-irrigation environment because the rise of temperature would not be problematic.

Handy Type Torque Meter (HTG2-200NC, IMADA-SS Corp, Aichi, Japan) was used for the placement of an implant, and the maximum torque values were measured. The specifications of the Handy Type Torque Meter were as follows: the measurement unit, Ncm; the accuracy, within ±0.5% FS; the measurement maximum torque, 200 Ncm; and the minimum resolution function, 0.1 Ncm. Thus, the device had enough coverage of the torque measurement in this study (96.0 Ncm at maximum).

Osstell Mentor® (Integration Diagnostics AB, Gôteborg, Sweden) was used as a resonance frequency analysis device to measure the ISQ values. The accompanying smart peg (type 1) was attached to the implant for the measurement at every 90°, and the average value was calculated.

An implant simulation software (Landmarker ver. 5.0 with special specifications for study purposes, iCAT, Osaka, Japan) was used for analysis. The evaluation site was selected on the smallest area that was as adjacent to the implant as possible so as not to include the area that was immune to the primary stability. Specifically, the width of the measurement site was defined as 0.50 mm, i.e., from 0.25 mm inside (the to-be-compressed area at the time of placement) to 0.25 mm outside of the virtual implant. The area adjacent to the bottom of the implant was excluded from the measurement site because the maximum torque value and the ISQ value were both subject to the lateral force a great deal (Figures 4 and 5). Then, the voxel values of the measurement site were extracted into the comma-separated values (CSV) files. Subsequently, the average of the voxel values of the measurement site was calculated and defined as the voxel value of the whole bone around the implant. Additionally, the average of the voxel values was calculated at every 0.1-mm depth from the surface of the bone to determine the thickness of the cortical bone from the voxel values. The threshold of the voxel value for the border of the cortical bone and the spongy bone was decided to be 350, which had a strong correlation (Pearson's correlation coefficient, 0.897) with the actual measurement in the preliminary examination.