Due to the growing importance of anterior segment measurements, numerous devices based on different principles have been developed. The most commonly used principles include standard corneal topography (with a Placido disk), Scheimpflug imaging, AS-OCT, color light-emitting diode (LED), OLCR as well as swept-source OCT (SS-OCT).

The SW-9000, a new OLCR-based optical biometer, has not yet been investigated. Any new device needs to be validated before it can be widely adopted in the clinical setting. It is considered eligible only if it is proven to be reliable and can be interchangeably used with other instruments. The MS-39 is the only device that combines SD-OCT and Placido. Previous studies have revealed high repeatability and reproducibility of MS-39 for anterior segment analysis [14, 19]. Good agreement was found when comparing the MS-39 with the Pentacam HR (OCULUS, Wetzlar, Germany) and Sirius (CSO, Florence, Italy), both of which are based on Scheimpflug imaging [14, 20, 21]. Similar results were obtained when investigating agreement between the MS-39 and SS-OCT-based devices such as Argos (Movu, Komaki, Japan) and ANTERION (Heidelberg, Heidelberg, Germany) [19, 22]. Besides, AS-OCT system like RTVue (Optovue, Fremont, CA) and PCI system like IOLMaster 500 (Carl Zeiss Meditec, Jena, Germany) had also been compared with MS-39. Nevertheless, no study has compared the MS-39 with the devices which adhered to OLCR principle. In order to comprehensively evaluate the accuracy of SW-9000, we compared the agreement and difference between the SW-9000 and MS-39.

When measuring CCT, our results revealed a significant difference between the MS-39 and SW-9000 (P < 0.001). The mean difference was 2.21 ± 2.67 μm, with the 95% LoA between − 3.01 and 7.44 μm. Comparison between Argos and MS-39 [19] showed that the mean difference was 5.78 ± 4.84 μm, and the 95% LoA was − 3.70 to 15.25 μm, nevertheless, the values in that study were larger than ours. Oh et al. [23] found excellent agreement between ANTERION based on SS-OCT and CASIA 2 (Tomey, Nagoya, Japan), with a mean difference value of 2.30 ± 6.30 μm and 95% LoA of − 10.06 to 14.65 μm. Hashemi et al. [24] compared an OLCR-based system (Lenstar LS900, Haag-Streit AG, Koeniz, Switzerland) and a Scheimpflug-Placido topographer (Pentacam HR) for measuring CCT, and obtained high agreement, where the mean difference was − 5.14 ± 7.52 μm, and 95% LoA was − 19.88 to 9.60 μm. The interval we obtained was narrower than those in most previous studies. Given the small mean difference value and narrow LoA in our study, we suggest that the two devices can be used interchangeably for CCT measurement.

In terms of ACD and AQD measurements, the SW-9000 presented lower values than the MS-39, as the mean difference were − 0.10 ± 0.03 mm and − 0.10 ± 0.04 mm, while the 95% LoAs were − 0.16 to − 0.05 mm and − 0.18 to − 0.03 mm, respectively. These intervals were slightly narrower than those reported by Ruan et al., who separately assessed agreement between the IOLMaster 700 (Carl Zeiss Meditec, Jena, Germany) and CASIA 2, yielding 95% LoAs of − 0.03 to 0.24 mm for ACD and 0.04 to 0.25 mm for AQD [25]. Similar results were reported when comparing the MS-39 and the Argos (− 0.01 ± 0.03 mm) for ACD and AQD measurements, which were insufficient to produce noticeable differences in clinic [19]. According to previous studies, the IOL power would change by 0.1 D with 0.1 to 0.2 mm change in ACD [26, 27]. Mean difference of 0.1 mm corresponds to a change of IOL power of 0.1 D, which has no influence in clinical practice. Consequently, although the differences obtained in our study were statistically significant, they were too small to have any noticeable impact on the refractive outcome.

As for the measurement of mean keratometry, the mean difference of − 0.01 ± 0.22 D indicated no significant difference between the two devices (P = 0.773), with 95% LoA was − 0.45 to 0.43 D. However, these results were slightly larger than those reported for the MS-39 and Argos [19]. Mehdizadeh et al. found that the calculated IOL power varies by 0.9 to 1.3 D for a 1.0 D change in K [28]. In addition, Jasvinder et al. reported that a difference of 1.0 D and 0.5 D in average K translates to approximately 1.0 D and 0.5 D difference in IOL power [29]. Hence, we strictly set the threshold of clinical difference at 0.5 D to ensure the visual acuity after IOL implantation. Under these circumstances, the absolute maximum value of the limit of 95% LoA (0.45 D) in this study was still less than the cutoff value. Apparently, when measuring Km, the difference between the two devices was clinically irrelevant.

CD, is an important parameter for determining the optical area in corneal refractive surgery and predict the vault after phakic IOL implantation. The mean difference in CD measurement between the MS-39 and SW-9000 was 0.20 ± 0.16 mm, with 95% LoA ranging between − 0.12 and 0.51 mm. When the MS-39 was compared to the Pentacam and Sirius, the 95% LoAs ranged between − 0.46 to + 0.19 mm and − 0.54 to + 0.47 mm, respectively, showing agreement close to the results obtained in a previous study [14]. Variations in detection methods as well as dissimilar methods of defining the limbus for various devices usually lead to non-optimal agreement [30,31,32]. Given that CD has been widely used for phakic IOL implantation and phakic IOLs are sized to the nearest 0.50 mm, ≥ 0.50 mm was set as the threshold for clinical difference [33, 34]. Therefore, 0.51 mm, which is a close approximation of the above threshold, indicates that the two instruments are interchangeable.

This study has some limitations. First, only young healthy people were included. Given the presence of multiple pathological eyes, such as keratoconus and post-corneal refractive surgery eyes, the good agreement between the two instruments in this study was not completely representative for different populations. Second, with increasing age, the corneal senile ring becomes more common, and whether this will worsen the agreement of CD measurement needs to be studied.