Nova tecnologia para neurocirurgias minimamente invasivas guiadas por aplicativo de SMARTPHONE

Abstract

The performance of less invasive and therefore less bloody neurosurgical approaches requires precise methods of locating points on the surface of the skull derived from the evaluation of imaging exams. In countries with limited resources for health, it is necessary to develop low- cost alternatives. This work aims to develop smartphone applications to guide minimally invasive neurosurgery craniotomies and to evaluate the accuracy and precision of the developed tools. The project in question consisted of four phases: two experimental stages and two clinical ones. Neurokeypoint application was developed, with the React Native framework for Android and iOS platforms, and Neurokeypoint AR, in Swift programming language. From a randomly selected CT of the normal skull, a head model was printed on a Z-Corp Model 310 Rapid Prototyping System 3D printer. Twelve fiducial adhesives were fixed in this model, which was then submitted to a new CT, used for accuracy assessment. The clinical phases consisted of non-probabilistic convenience samples with prospective data collection. After development, the Neurokeyoint application remained operationally stable on both Android and iOS mobile platforms. Associated with a compass, the overall mean target registrations error (TRE) was 1.6 ± 1.0 mm. There was also no relevant difference between test and retest (1.7 ± 1.1 vs 1.4 ± 0.9 mm, p = 0.340). The Neurokeypoint AR application, using augmented reality, was developed exclusively for iOS and, under optimized environmental conditions, obtained an average TRE of 2.6 ± 1.6 mm with iPhone 6s and 2.4 ± 1.2 mm with iPhone 12 Pro. In the clinical stages, 122 patients had their surgical plans using the developed tools (88 with neurokeypoint and 34 with neurokeypoint AR). With Neurokeypoint, 41 tumors (24 gliomas, 13 metastases and 4 meningiomas), 17 traumatic brain injuries, 16 spontaneous intraparenchymal hematomas, seven ventriculoperitoneal shunts, four abscesses, two cavernomas and one arteriovenous malformation were operated on. Thirty-four patients were operated using Neurokeypoint AR as a planning tool. They had the following diseases: thirteen intracranial metastases, nine meningiomas, seven gliomas and five intraparenchymal hematomas. Craniotomies assisted by Neurokeypoint AR had a mean diameter of 4.3 ± 1.9 cm, ranging from 2.5 to 10.4 cm. The incisions were mostly straight (55.9%). All cases of meningiomas, 12 metastases and 4 cases of gliomas underwent gross total resection. The Neurokeypoint and Neurokeypoint AR applications have been successfully developed and enable both craniotomy and trephination approaches to head injuries in a precise, less invasive and low-cost manner.