Gülden Karabiber, Hanife Nuray Yılmaz, Gamze Yıldırım

- Korean J Orthod 2025;55:37-47

I appreciate the effort and dedication you and your team have put into your study, “Are different photogrammetry applications on smartphones sufficiently reliable?”. Your research provides valuable insights into the accuracy and clinical applicability of smartphone-based photogrammetry, which is an area of great interest in modern orthodontics. I found your findings particularly thought-provoking, as they highlight both the potential and the current limitations of these emerging technologies.

A1. We think that advancements in calibration techniques and smartphone hardware have primary importance to overcome this problem. Technical studies that evaluating and comparing properties of hardware and calibration techniques are needed especially in the engineering field. Advancements in smartphone hardware, such as improved sensors, higher resolution cameras, and better lenses, could reduce lens distortion and enhance image quality, leading to more accurate three-dimensional (3D) models.¹ Calibration techniques also play a vital role in achieving reliable measurements.² The use of pre-calibration in the Motion-Multi Stereo View method increases the accuracy of 3D modeling when compared with self-calibration method.² Additionally, better algorithms and software could improve image alignment, feature matching, noise filtering and further refining results.¹

A2. The image processing time should not be confused with the time for capturing of images with a smartphone. The duration for serial capturing of the images are similar in both applications however the processing duration for reconstruction of 3D images takes longer time in Magiscan. This is a disadvantage if patient records need to be taken consecutively in busy clinics. The longer image processing time may be due to larger files that are initially generated from larger-sized images to obtain more detailed and accurate models. Additionally building a polygonal mesh and generating a texture map are the most time-consuming step so, the longer image processing time may also be due to the additional analyses.¹

A3. Future studies need bigger sample size to better understand this issue. It is a fact that smartphone-based photogrammetry systems such as Qlone and Magiscan in the present version provide less reliable results compared to 3dMD. This difference is especially important in facial soft tissue measurements and orthodontic applications that require metric accuracy.

Fortunately, smartphone cameras offer higher resolution and depth perception every year. With the widespread use of sensors such as light detection and ranging, the quality of capturing surface details is improving significantly.³ In addition, if photogrammetry software is supported by artificial intelligence and deep learning, a rapid development in reducing noise in images, correcting alignment errors and providing more accurate results might be seen in volumetric analyses.⁴ The use of pre-calibration methods can also improve the accuracy.²

However, the most important limitation is the longer image acquisition time with smartphone applications compared to 3dMD which may cause artefacts due to patient movement.⁵ Although conditions such as ideal light and background are provided for smartphone applications, it would be useful to develop multi simultaneous shooting camera systems that can be connected to the smartphone while the software and operating part is running on the phone.

Smartphone-based systems can currently be used in non-critical applications such as preliminary scans, patient follow-up, treatment motivation before-after visualization.² However, in fields that require high accuracy such as diagnosis, surgical planning, or official documentation of post-treatment results, it needs to go a little further to compete with systems such as 3dMD.