A Sensorised Surgical Glove to Analyze Forces During Neurosurgery

From experimental to essential: The evolving role of augmented reality in neurosurgery (2012-2024)

Recent years have seen augmented reality (AR) transition from experimental to clinical practice. Advancements in hardware, software, and its integration with complementary technologies such as machine learning and robotics have improved its workflow and integration into the neurosurgical environment. This systematic review evaluates shifts in trends in AR adoption in neurosurgery from 2022 to 2024. A systematic review of PubMed was conducted following PRISMA guidelines. Studies published between January 2022 and December 2024 that had direct clinical or educational applications were included. Extracted data included the clinical context and geographical context from each study, and was analyzed with data from a previous systematic review from 2012 to 2021 to assess research evolution. A total of 275 new studies were identified, revealing a substantial increase in AR-related publications. Research trends have shifted towards more clinical embedded topics, particularly centered around neuronavigation (101), education (87), and spinal surgery (70), with the subspecialties exhibiting the most growth being spinal surgery, vascular surgery and neuro-oncology. Research output remained concentrated in high-income countries, led by the United states (53%), Switzerland (18.55%) and the UK (9.45%), reinforcing an expanding global disparity. Topic clustering analysis identified education as a central point of focus across subspecialties. As AR continues to become increasingly integrated within the neurosurgical workflow, future research should emphasize standardizing its clinical implementation and addressing global disparities in access and adoption.

Augmented reality technology shortens aneurysm surgery learning curve for residents

OBJECTIVES

We aimed to prospectively investigate the benefit of using augmented reality (AR) for surgery residents learning aneurysm surgery.

MATERIALS AND METHODS

Eight residents were included, and divided into an AR group and a control group (4 in each group). Both groups were asked to locate an aneurysm with a blue circle on the same screenshot after their viewing of surgery videos from both AR and non-AR tests. Only the AR group was allowed to inspect and manipulate an AR holographic representation of the aneurysm in AR tests. The actual location of the aneurysm was defined by a yellow circle by an attending physician after each test. Localization deviation was determined by the distance between the blue and yellow circle.

RESULTS

Localization deviation was lower in the AR group than in the control group in the last 2 tests (AR Test 2: 2.7 ± 1.0 mm vs. 5.8 ± 4.1 mm, p = 0.01, non-AR Test 2: 2.1 ± 0.8 mm vs. 5.9 ± 5.8 mm, p < 0.001). The mean deviation was lower in non-AR Test 2 as compared to non-AR Test 1 in both groups (AR: p < 0.001, control: p = 0.391). The localization deviation of the AR group decreased from 8.1 ± 3.8 mm in Test 2 to 2.7 ± 1.0 mm in AR Test 2 (p < 0.001).

CONCLUSION

AR technology provides an effective and interactive way for neurosurgery training, and shortens the learning curve for residents in aneurysm surgery.

Medical student perception of force application: An accuracy assessment and pilot training program

BACKGROUND

It is unclear how accurately students can reproduce specific forces that are often required for physical examination maneuvers. This study aimed to determine the baseline accuracy of force application for preclinical medical students, evaluate the effectiveness of a quantitative visual feedback intervention, and investigate whether certain demographics influence accuracy.

MATERIALS AND METHODS

First- and second-year medical students were enrolled and demographic data were collected. Students blindly applied their estimation of 15 lbs (6.8 kg), 3 lbs (1.4 kg), 10 lbs (4.5 kg), 1.5 lbs (0.7 kg), and 6 lbs (2.7 kg) of force on a scale. Visual feedback training was then performed wherein students applied a series of additional forces unblinded five times, and then blindly administered the same five initial forces 12 minutes and one week later. Accuracy was compared at each time point and a regression analysis was evaluated for predictors of accuracy.

RESULTS

Thirty-three students participated. The mean baseline accuracy was 38.3%, 41.1% immediately following intervention, and 35.6% one week later (P = 0.66). Accuracy was significantly higher at higher intended forces compared to lower forces (P < 0.05). The number of prior occupations was a positive independent predictor (P = 0.04), and the number of sports played was noted to be a negative predictor (P = 0.01), of baseline accuracy.

CONCLUSIONS

Medical students' ability to accurately reproduce clinically relevant forces is poor. There is a clear need to implement a robust training program in medical education, and students may need multiple training sessions to refine this skill.

Validation of a surgical simulator and establishment of quantitative performance thresholds-RealSpine simulation system for open lumbar decompressions

BACKGROUND CONTEXT

The majority of surgical training is conducted in real-world operations. High-fidelity surgical simulators may provide a safer environment for surgical training. However, the extent that it reflects real-world operations and surgical ability is often poorly characterized.

PURPOSE

(1) Assess the validity and fidelity of a surgical simulator; (2) Examine the quantitative relationship between simulation performance and markers of real-world ability; (3) Establish thresholds for surgical expertise, and estimate their external validity and accuracy.

STUDY DESIGN/SETTING

A cohort study of surgeons at a British neurosurgical center.

STUDY SAMPLE

Ten early-career "novice" surgeons and 8 board-certified "expert" neurosurgeons.

OUTCOMES MEASURES

(1) Face and content validity, and visual and haptic fidelity; (2) Construct validity; (3) Predictive and discriminative utility of quantitative performance thresholds.

METHODS

Participants performed unilateral lumbar decompressions on high-fidelity spinal simulators that replicate the bony and soft tissue anatomy along with physiological processes such as bleeding and CSF leaks. Operating times, measured from first surgical action to either self-perceived satisfactory decompression or the end of allocated time, were recorded. The performance was also assessed independently by 2 blinded spinal subspecialist neurosurgeons using OSATS, a validated surgical assessment tool that utilizes 5-point scales on a variety of technical domains to grade the overall technical proficiency. Validity and fidelity were assessed by expert neurosurgeons using quantitative questionnaires. Construct validity was assessed by ordinal regression of simulation performance against real-world surgical grade and portfolio. Thresholds of expert status by simulation performance was established, and their predictive and discriminative utility assessed by crossvalidation accuracy and AUC-ROC.

RESULTS

Operating time and expert assessments of simulation performance (OSATS) were strong and significant prdictors of surrogate markers of real-world surgical ability. The thresholds for expert status were operating time of 15 minutes and modified OSATS score of 15/20. These thresholds predicted expert status with 84.2% and 71.4% accuracy respectively. Strong discriminative ability was demonstrated by AUC-ROC of 0.95 and 0.83 respectively. All expert surgeons agreed that RealSpine simulators demonstrate high face validity, and high visual and haptic fidelity, with overall scores showing statistically significant agreement on these items (all scores at least 4/5, p<.0001). There was less consensus on content validity, but with still significant overall agreement (average score: 3.75/5, p=.023).

CONCLUSIONS

Real-world surgical ability and experience can be accurately predicted by defining objective quantitative thresholds on high-fidelity simulations. The thresholds established here, along with other data presented in this paper, may inform objectives and standards to be established in a spinal surgical training curriculum.

Computational Modeling, Augmented Reality, and Artificial Intelligence in Spine Surgery

Over the past decade, advancements in computational modeling, augmented reality, and artificial intelligence (AI) have been driving innovations in spine surgery. Much of the research conducted in these fields is from the past 5 years. In 2021, the market value for augmented reality and virtual reality reached around $22.6 billion, highlighting the rise in demand for these technologies in the medical industry and beyond. Currently, these modalities have a wide variety of potential uses, from preoperative planning of pedicle screw placement and assessment of surgical instrumentation to predictions for postoperative outcomes and development of educational tools. In this chapter, we provide an overview of the applications of these technologies in spine surgery. Furthermore, we discuss several avenues for further development, including integrations between these modalities and areas of improvement for more immersive, informative surgical experiences.

A Deep Learning Approach to Classify Surgical Skill in Microsurgery Using Force Data from a Novel Sensorised Surgical Glove

Microsurgery serves as the foundation for numerous operative procedures. Given its highly technical nature, the assessment of surgical skill becomes an essential component of clinical practice and microsurgery education. The interaction forces between surgical tools and tissues play a pivotal role in surgical success, making them a valuable indicator of surgical skill. In this study, we employ six distinct deep learning architectures (LSTM, GRU, Bi-LSTM, CLDNN, TCN, Transformer) specifically designed for the classification of surgical skill levels. We use force data obtained from a novel sensorized surgical glove utilized during a microsurgical task. To enhance the performance of our models, we propose six data augmentation techniques. The proposed frameworks are accompanied by a comprehensive analysis, both quantitative and qualitative, including experiments conducted with two cross-validation schemes and interpretable visualizations of the network's decision-making process. Our experimental results show that CLDNN and TCN are the top-performing models, achieving impressive accuracy rates of 96.16% and 97.45%, respectively. This not only underscores the effectiveness of our proposed architectures, but also serves as compelling evidence that the force data obtained through the sensorized surgical glove contains valuable information regarding surgical skill.

A synthetic model simulator for intracranial aneurysm clipping: validation of the UpSurgeOn AneurysmBox

Background and objectives

In recent decades, the rise of endovascular management of aneurysms has led to a significant decline in operative training for surgical aneurysm clipping. Simulation has the potential to bridge this gap and benchtop synthetic simulators aim to combine the best of both anatomical realism and haptic feedback. The aim of this study was to validate a synthetic benchtop simulator for aneurysm clipping (AneurysmBox, UpSurgeOn).

Methods

Expert and novice surgeons from multiple neurosurgical centres were asked to clip a terminal internal carotid artery aneurysm using the AneurysmBox. Face and content validity were evaluated using Likert scales by asking experts to complete a post-task questionnaire. Construct validity was evaluated by comparing expert and novice performance using the modified Objective Structured Assessment of Technical Skills (mOSATS), developing a curriculum-derived assessment of Specific Technical Skills (STS), and measuring the forces exerted using a force-sensitive glove.

Results

Ten experts and eighteen novices completed the task. Most experts agreed that the brain looked realistic (8/10), but far fewer agreed that the brain felt realistic (2/10). Half the expert participants (5/10) agreed that the aneurysm clip application task was realistic. When compared to novices, experts had a significantly higher median mOSATS (27 vs. 14.5; p < 0.01) and STS score (18 vs. 9; p < 0.01); the STS score was strongly correlated with the previously validated mOSATS score (p < 0.01). Overall, there was a trend towards experts exerting a lower median force than novices, however, these differences were not statistically significant (3.8 N vs. 4.0 N; p = 0.77). Suggested improvements for the model included reduced stiffness and the addition of cerebrospinal fluid (CSF) and arachnoid mater.

Conclusion

At present, the AneurysmBox has equivocal face and content validity, and future versions may benefit from materials that allow for improved haptic feedback. Nonetheless, it has good construct validity, suggesting it is a promising adjunct to training.