RESUMO
PURPOSE: To evaluate whether a commercial AI tool for intracranial hemorrhage (ICH) detection on head CT exhibited sociodemographic biases. METHODS: Our retrospective study reviewed 9736 consecutive, adult non-contrast head CT scans performed between November 2021 and February 2022 in a single healthcare system. Each CT scan was evaluated by a commercial ICH AI tool and a board-certified neuroradiologist; ground truth was defined as final radiologist determination of ICH presence/absence. After evaluating the AI tool's aggregate diagnostic performance, sub-analyses based on sociodemographic groups (age, sex, race, ethnicity, insurance status, and Area of Deprivation Index [ADI] scores) assessed for biases. χ2 test or Fisher's exact tests evaluated for statistical significance with p ≤ 0.05. RESULTS: Our patient population was 50% female (mean age 60 ± 19 years). The AI tool had an aggregate accuracy of 93% [9060/9736], sensitivity of 85% [1140/1338], specificity of 94% [7920/ 8398], positive predictive value (PPV) of 71% [1140/1618] and negative predictive value (NPV) of 98% [7920/8118]. Sociodemographic biases were identified, including lower PPV for patients who were females (67.3% [62,441/656] vs. 72.7% [699/962], p = 0.02), Black (66.7% [454/681] vs. 73.2% [686/937], p = 0.005), non-Hispanic/non-Latino (69.7% [1038/1490] vs. 95.4% [417/437]), p = 0.009), and who had Medicaid/Medicare (69.9% [754/1078]) or Private (66.5% [228/343]) primary insurance (p = 0.003). Lower sensitivity was seen for patients in the third quartile of national (78.8% [241/306], p = 0.001) and state ADI scores (79.0% [22/287], p = 0.001). CONCLUSIONS: In our healthcare system, a commercial AI tool had lower performance for ICH detection than previously reported and demonstrated several sociodemographic biases.
RESUMO
STUDY DESIGN: Feasibility study. PURPOSE: A phantom model was used to evaluate the accuracy of a novel augmented reality (AR) system for cervical screw placement. OVERVIEW OF LITERATURE: The use of navigation systems is becoming increasingly common in spine procedures. However, numerous factors limit the feasibility of regular and widespread use of navigation tools during spine surgery. AR is a new technology that has already demonstrated utility as a navigation tool during spine surgery. However, advancements in AR technology are needed to increase its adoption by the medical community. METHODS: AR technology that uses a fiducial-less registration system was tested in a preclinical cervical spine phantom model study for accuracy during spinal screw placement. A three-dimensional reconstruction of the spine along with trajectory lines was superimposed onto the phantom model using an AR headset. Participants used the AR system to guide screw placement, and post-instrumentation scans were compared for accuracy assessment. RESULTS: Twelve cervical screws were placed under AR guidance. All screws were placed in an acceptable anatomic position. The average distance error for the insertion point was 2.73±0.55 mm, whereas that for the endpoint was 2.71±0.69 mm. The average trajectory angle error for all insertions was 2.69°±0.59°. CONCLUSIONS: This feasibility study describes a novel registration approach that superimposes spinal anatomy and trajectories onto the surgeon's real-world view of the spine. These results demonstrate reasonable accuracy in the preclinical model. The results of this study demonstrate that this technology can assist with accurate screw placement. Further investigation using cadaveric and clinical models is warranted.
RESUMO
Management of chronic subdural hematoma (cSDH) poses unique challenges and can be fraught with complications. Understanding the spatial relationships of cSDH to adjacent brain tissue and skull topography is critical for successful surgical treatment. The aim of this report is to highlight the feasibility and efficacy of a novel augmented reality (AR) overlay tool for surgical planning with technical description of two surgical cases using AR for surgical management of cSDH. This report describes a fiducial-less AR system for surgical planning of surgical evacuation of cSDH. The AR system was used to superimpose 3D anatomy onto the patients head to provide image guidance during two cases of evacuation. Imaging demonstrated convexity cSDH. A 3D model of the patient's anatomy was created and registered onto the patients' heads using a novel AR system. Surgical evacuation of the cSDH was completed in each case with surgical planning assisted by AR overlay.
RESUMO
Background and importance: Vestibular schwannomas are benign tumors and are the most common tumor found in the cerebellopontine angle. Surgical management of these lesions involves consideration of various operative approaches, which can have profound effects on procedural course and patient outcomes. Therefore, a comprehensive understanding of the location of the tumor and surrounding anatomical structures is vital for a positive outcome. We present a case of a 47-year-old female patient with vestibular schwannoma. A novel mixed reality (MR) system was used to register patient-specific 3D models onto the patient's head for operative planning and anatomical visualization. Case description: A 47-year-old female presented with a history of left-sided hearing loss, tinnitus, and episodic left facial tingling. Magnetic Resonance Imaging (MRI) demonstrated a 3.3â cm enhancing lesion in the left cerebellopontine angle at the with mass effect on the brachium pontis and medulla. Surgical resection was performed via retrosigmoid craniotomy. Conclusions: In this study, we report the use of Augmented Reality (AR) visualization for planning of vestibular schwannoma resection. This technology allows for efficient and accurate registration of a patient's 3D anatomical model onto their head while positioned in the operating room. This system is a powerful tool for operative planning as it allows the surgeon to visualize critical anatomical structures where they lie on the patient's head. The present case demonstrates the value and use of AR for operative planning of complex cranial lesions.
RESUMO
With its recent release, the Apple Vision Pro (Apple Inc., Cupertino, CA) represents a promising technological advancement of mixed reality in the field of neurosurgery and medicine more broadly. With all new technologies, it is critical to facilitate early use and assessment of the technology to facilitate adoption by the larger medical community. A 44-year-old female with a history of ruptured intracranial aneurysm status post anterior communicating artery aneurysm clipping presented with worsened confusion and intermittent headache. CT imaging revealed evidence of hydrocephalus due to the malfunction of a previous right parietal ventriculoperitoneal (VP) shunt. Prior to the case, the Apple Vision Pro was used in the operating room to visualize and interact with a 3D model of the patient's anatomy for the patient undergoing a VP shunt placement. A visualization of the 3D model through the headset was used to plan the approach and entry point. At the conclusion of the procedure, all clinicians and operating staff who used the technology for planning completed a survey about their initial impressions of the headset. Overall, users felt the 3D models felt realistic (4.5/5), that the display of the user's real-world view felt natural (4.3/5), and that the headset did not cause eye strain or fatigue (4.5/5). The majority of users responded that they would continue to use the headset for cases (4/5). This represents one of the first known clinical uses of the Apple Vision Pro. It is a cutting-edge technology that will likely provide immense value for healthcare providers as it becomes more integrated into clinical care.
RESUMO
Image guidance technologies can significantly improve the accuracy and safety of intracranial catheter insertions. Augmented reality (AR) allows surgeons to visualize 3D information overlaid onto a patient's head. As such, AR has emerged as a novel image guidance technology that offers unique advantages when navigating intracranial targets. A 71-year-old woman with a history of brain metastasis from breast cancer and prior resection surgery and chemotherapy presented with altered mental status and generalized weakness worse on her left side. Magnetic resonance imaging (MRI) demonstrated right frontotemporoparietal edema with a contrast-enhancing mass. MR perfusion confirmed an active tumor with an enlarging right temporal pole cyst. A cyst aspiration was performed via Ommaya reservoir placement. Neuro-navigation (BrainLab, Munich, Germany) and AR navigation were used to plan the trajectory from the temporal gyrus to the cyst. Post-operative computed tomography (CT) demonstrated good placement of the reservoir, reconstitution of the temporal horn of the lateral ventricle with decreased external mass effect, and no areas of hemorrhage. AR has tremendous potential in the field of neurosurgery for improving the accuracy and safety of procedures. This case demonstrates an encouraging application of AR and can serve as an example to drive expanded clinical use of this technology.
RESUMO
OBJECTIVE: External ventricular drain (EVD) placement is a common neurosurgical procedure that can be performed at bedside. A frequent complication following EVD placement is catheter-associated hemorrhage (CAH). The hemorrhage itself is rarely clinically significant but may be complicated in patients taking anticoagulant or antiplatelet (AC/AP) medications. METHODS: A total of 757 patients were who underwent EVD placement at bedside were included as part of a retrospective study at a large academic medical center. Demographic factors, use of AC/AP therapies, and several other clinical variables were recorded and assessed in univariate and multivariate regression analysis for association with CAH and mortality. RESULTS: One hundred (13.2%) patients experienced CAH within 24 hours of the procedure. After univariate analysis, in 2 tandem-run multivariate regression analyses after stepwise variable selection, use of 2 or more AC/AP agents (odds ratio [OR] = 2.362, P = 0.020) and dual antiplatelet therapy with aspirin and clopidogrel (OR = 3.72, P = 0.009) were significantly associated with CAH. Use of noncoated catheters was a protective factor against CAH compared to use of antibiotic-coated catheters (OR = 0.55, P = 0.019). Multivariate analysis showed age, multiagent therapy, and thrombocytopenia were significantly associated with increased mortality. CONCLUSIONS: There was increased risk of CAH after EVD placement in patients taking more than one AC/AP agent regardless of presenting pathology. In particular, use of aspirin and clopidogrel combined was associated with significantly higher odds of CAH, although it was not associated with higher mortality. In addition, there appears to be an association between use of antibiotic-coated catheters and CAH across univariate and multivariate analysis.
Assuntos
Anticoagulantes , Inibidores da Agregação Plaquetária , Humanos , Inibidores da Agregação Plaquetária/efeitos adversos , Anticoagulantes/efeitos adversos , Estudos Retrospectivos , Clopidogrel , Neurocirurgiões , Drenagem/efeitos adversos , Drenagem/métodos , Hemorragia/etiologia , Aspirina , Catéteres/efeitos adversos , Ventriculostomia/efeitos adversos , Antibacterianos/uso terapêuticoRESUMO
Background: Ventriculoperitoneal (VP) shunt placement is one of the most commonly performed neurosurgical procedures, yet failure rates remain very high. Surface landmarks are typically used to guide VP shunt placement, but they are not reliable in identifying the target anatomy. Augmented reality (AR) is a promising new technology that has the potential to improve the accuracy and effectiveness of neurosurgical procedures. We describe the use of AR for the surgical planning of a VP shunt. Case Description: A 62-year-old male with a history of subarachnoid hemorrhage presented with delayed hydrocephalus. A computed tomography scan was obtained that confirmed dilated ventricles, requiring a right VP shunt. The patient was brought to the operating room, where the AR system was used for visualization and planning. Conclusion: In this study, we describe the use of AR for VP shunt placement. The AR system consists of a Microsoft HoloLens 2 head-mounted display and a novel markerless registration system, which was used to register patient-specific 3D models onto the patient's head for visualizing target anatomy and planning an operative approach. The AR system was used to plan the VP shunt placement in the operating room. This system is easy to use and provides a visualization of the patient's anatomy, which can be used to plan an optimal trajectory. We believe that this has the potential to improve the accuracy and outcomes of VP shunt placements, and further studies are needed to characterize the system's accuracy and benefits.