Evaluation of Accuracy and Reliability of a Smartphone Stereotactic Coordinates Checking Application

Júlio César de Almeida,Marcos Dellaretti,Pedro Henrique Piaulino Benvindo Ferreira,Tancredo Alcântara Ferreira Jr,Leyzeane Marques do Nascimento,Marcello Penholate Faria
DOI: https://doi.org/10.1016/j.wneu.2020.05.121
Abstract:Background: Surgical assistance applications for smartphones have the potential to be used in daily practice; however, regular reviews of these tools are required. StereoCheck (Mevis, São Paulo, São Paulo, Brazil) is a mobile application (app) designed to compute stereotactic coordinates as a checking tool. The present study evaluated the accuracy and reliability of the StereoCheck app. Methods: The present observational and prospective study included 26 patients who had undergone frame-based stereotactic brain biopsy. A standard stereotactic planning software (Framelink, version 5.0 [Medtronic, Minneapolis, Minnesota, USA]) was used to define the target coordinates. The surgical planning images were transferred to StereoCheck using 2 image input methods: photographs of the screen monitor and digitally exported images of the screen monitor. Five examiners performed the target coordinate definitions using StereoCheck on 2 occasions. The accuracy and reliability of the app were evaluated in tests that compared the planning methods (app vs. standard software), multiple examiners, and sequential tests. Results: The StereoCheck accuracy using photographs of the screen was 2.71 ± 0.86 mm (95% confidence interval, 2.37-3.06). Using the digitally exported images, it was 0.82 ± 0.61 mm (95% confidence interval, 0.58-1.07). The accuracy between the 2 methods was greater using the exported images (P < 0.01) and was not affected by the clinical and radiological features. The accuracy of StereoCheck among multiple examiners and in sequential tests showed a mean distance between the targets and coordinates of <1.00 mm. Thus, an excellent level of reliability for the StereoCheck coordinates (intraclass correlation coefficient ≥0.8) was verified in all contexts. Conclusion: StereoCheck showed satisfactory accuracy and reliability. The use of photographs to compute the coordinates could lead to a significant decrease in the accuracy of the app.
What problem does this paper attempt to address?