Artificial Intelligence in Neurosurgery

• Advances and Integrations of Computer-Assisted Planning, Artificial Intelligence, and Predictive Modeling Tools for Laser Interstitial Thermal Therapy in Neurosurgical Oncology
• A Data-Centric Approach to Deep Learning for Brain Metastasis Analysis at MRI
• Subthalamic nucleus or globus pallidus internus deep brain stimulation for the treatment of parkinson's disease: An artificial intelligence approach
• Clinical feasibility of AI Doctors: Evaluating the replacement potential of large language models in outpatient settings for central nervous system tumors
• World's First Artificial Intelligence-Based Evaluation of Rist Catheter Stability in Transradial Procedures: A Feasibility Study
• Spinal Instability Neoplastic Score as a Predictor of Vertebral Fracture in Patients Undergoing Radiation Therapy for Spinal Metastases: A Single-Institution Study
• Learning to combine top-down context and feed-forward representations under ambiguity with apical and basal dendrites
• Clinical outcome and deep learning imaging characteristics of patients treated by radio-chemotherapy for a "molecular" glioblastoma

Artificial intelligence (AI) is revolutionizing the field of neurosurgery by improving diagnosis, neurosurgical planning, intraoperative guidance, and postoperative care. Integrating AI-driven tools in neurosurgery enhances precision, reduces complications, and optimizes patient outcomes.

1. Radiomics & Deep Learning: AI-powered algorithms analyze MRI, CT, and PET scans to detect brain tumors, vascular malformations, and degenerative conditions with high accuracy.

1. Automated Segmentation: Machine learning models assist in the segmentation of brain structures and pathology (e.g., gliomas, aneurysms) for improved diagnostic accuracy.
3. Predictive Analytics: AI predicts disease progression or recurrence likelihood based on imaging patterns and clinical data.

1. Neuro-navigation Systems: AI enhances precision in preoperative planning by integrating multimodal imaging data (MRI, CT, tractography).

1. Virtual & Augmented Reality: AI-based simulations enable neurosurgeons to practice complex procedures in a risk-free environment before surgery.

1. Predictive Models: Machine learning algorithms forecast potential complications and outcomes based on patient-specific parameters.

1. Robot-Assisted Surgery: AI-driven robotic systems (e.g., ROSA, Mazor X) improve accuracy in procedures like deep brain stimulation (DBS) and spinal fusion.

1. Real-Time Image Processing: AI enhances intraoperative imaging (e.g., fluorescence-guided surgery) for better tumor margin identification.

1. AI-Enhanced Microsurgery: AI-powered tools assist in delicate procedures by optimizing microscope settings and providing real-time guidance.

1. Automated Monitoring Systems: AI-based ICU monitoring helps detect early signs of deterioration in neurosurgical patients.

1. AI for Rehabilitation: Machine learning models predict recovery trajectories and optimize rehabilitation protocols for stroke and spinal cord injury patients.

1. Outcome Prediction Models: AI predicts long-term functional outcomes based on perioperative data.

1. Natural Language Processing (NLP): AI-powered literature review tools analyze vast amounts of neurosurgical research to identify trends and breakthroughs.

1. AI-Based Training Platforms: AI-driven simulators provide interactive learning experiences for neurosurgical residents.

1. Data Privacy & Security: AI relies on vast datasets that must be protected to ensure patient confidentiality.
2. Bias & Generalizability: AI models must be trained on diverse populations to prevent biases in clinical decision-making.
3. Regulatory & Legal Issues: The use of AI in neurosurgery must comply with medical regulations to ensure patient safety and accountability.

### Conclusion AI is transforming neurosurgery by enhancing diagnosis, surgical precision, and patient care. While challenges remain, ongoing advancements in AI technology will continue to revolutionize the field, ultimately improving outcomes for neurosurgical patients.

Would you like a more detailed breakdown on a specific AI application in neurosurgery?

Neurosurgery has evolved alongside technological innovations; however, these advances have also introduced greater complexity into clinical practice. Neurosurgery remains a demanding and high-risk field that requires a broad range of skills. Artificial intelligence (AI) has immense potential in neurosurgery given its ability to rapidly analyze large volumes of clinical data generated in modern clinical environments. An expanding body of literature has demonstrated that AI enhances various aspects of neurosurgery, including diagnostics, prognostication, decision-making, data management, education, and clinical studies. AI applications are expected to reduce medical errors and costs, broaden healthcare accessibility, and ultimately boost patient safety and surgical education. Nevertheless, AI application in neurosurgery remains practically limited because of several challenges, such as the diversity and volume of clinical training data collection, concerns regarding data quality, algorithmic bias, transparency (explainability and interpretability), ethical issues, and regulatory implications ¹⁾

Artificial intelligence is used in various aspects of neurosurgery, including diagnosis, neurosurgical planning and navigation, treatment outcome prediction, intraoperative monitoring, robot-assisted neurosurgery, and rehabilitation. However, challenges, such as data bias, ethical issues, costs, and regulations, remain. In Japan, issues such as the uneven distribution and decline of neurosurgeons, the collapse of regional healthcare, and the increase in the number of patients with spinal disorders due to aging have been highlighted. The “AI and Robot-Assisted Surgery Moonshot Plan” serves as a guide to overcome the challenges of neurosurgery in Japan and establish a sustainable medical system ²⁾.

The emergence of artificial intelligence (AI) in the medical field holds promise in improving medical management, particularly in personalized strategies for the diagnosis and treatment of brain tumors. However, integrating AI into clinical practice has proven to be a challenge. Deep learning (DL) is very convenient for extracting relevant information from large amounts of data that has increased in medical history and imaging records, which shortens diagnosis time, that would otherwise overwhelm manual methods. In addition, DL aids in automated tumor segmentation, classification, and diagnosis. DL models such as the Brain Tumor Classification Model and the Inception-Resnet V2, or hybrid techniques that enhance these functions and combine DL networks with support vector machine and k-nearest neighbors, identify tumor phenotypes and brain metastases, allowing real-time decision-making and enhancing preoperative planning. AI algorithms and DL development facilitate radiological diagnostics such as computed tomography, positron emission tomography scans, and magnetic resonance imaging (MRI) by integrating two-dimensional and three-dimensional MRI using DenseNet and 3D convolutional neural network architectures, which enable precise tumor delineation. DL offers benefits in neuro-interventional procedures, and the shift toward computer-assisted interventions acknowledges the need for more accurate and efficient image analysis methods ³⁾

What is the level of confidence, knowledge and the attitude of the global neurosurgical community towards AI basic concepts and applications?

A 24-item survey was designed and distributed. The survey results reported on level of knowledge, confidence and interest in AI, perspectives and attitude towards the application of AI technologies in neurosurgery. The potential influence of demographics and work-related environment features on AI knowledge was investigated.

They received a total of 250 responses from 61 countries. The correct definition of 'Machine Learning', 'Deep Learning' and main Big Data features were identified by respectively 42%, 23% and 23% of the respondents. The survey unveiled a strong interest and a positive attitude towards the introduction of AI in the neurosurgical practice. The main concerns included trustworthiness and liability, the main barriers to implementation were considered lack of funding, infrastructure, knowledge and multidisciplinary collaboration.

There is a low familiarity with basic AI concepts in the neurosurgical community. Nevertheless, there is a strong interest and a positive attitude towards AI implementation. The systematization of training and the production of educational resources will be key in guaranteeing a successful implementation of AI in the evolving Neurosurgery History ⁴⁾

Artificial Intelligence for Preoperative Planning in Neurosurgery.

Intraoperative Assistance: Explore AI-based navigation systems, robotic surgery assistants, and augmented reality (AR) applications that provide real-time guidance during surgery.

Postoperative Care: Mention AI’s role in monitoring patients post-surgery, predicting complications, and managing long-term outcomes.

3. Case Studies and Examples

1. Local Applications in Valencia and Murcia: If there are specific examples of AI being used in neurosurgery within the Valencia and Murcia regions, highlight these. This could include local hospitals, research institutions, or startups.
2. Global Innovations: Present notable cases from around the world, such as the use of AI in robotic neurosurgery (e.g., ROSA robot), predictive analytics for brain tumor growth, or AI-driven personalized treatment plans.

4. Benefits of AI in Neurosurgery

1. Increased Precision and Accuracy: How AI improves surgical outcomes by providing enhanced visualization and precision.
2. Reduced Surgical Time and Costs: Efficiency gains from AI that reduce operation times and associated costs.
3. Improved Patient Outcomes: How AI helps in reducing surgical risks and improving recovery times.

5. Challenges and Ethical Considerations

1. Data Privacy and Security: Discuss the importance of protecting patient data and the challenges of ensuring data privacy in AI systems.
2. Bias and Fairness: The risk of biases in AI algorithms, which could lead to disparities in care.
3. Regulatory Issues: Overview of the regulatory landscape in Spain and Europe for AI in healthcare, and how these regulations impact the adoption of AI technologies.

6. Future Directions

1. Emerging Technologies: Mention future possibilities such as AI in brain-computer interfaces (BCIs), predictive models for neurological disorders, and personalized medicine.
2. Research and Development: Highlight ongoing research in the Valencia and Murcia regions and opportunities for collaboration.

7. Conclusion

1. Summary of Key Points: Recap the benefits and challenges of AI in neurosurgery.
2. Call to Action: Encourage neurosurgeons and residents to engage with AI technologies and participate in research and training.

8. Q&A Session

1. Open the floor for questions and discussions to address specific interests or concerns from the audience.

### Tips for Your Presentation

- Use Visual Aids: Incorporate videos, diagrams, and interactive elements like AR or VR if possible to showcase how AI tools function in neurosurgery. - Stay Updated: Make sure your information is current, especially regarding local examples and the latest advancements in AI technologies. - Engage the Audience: Tailor your presentation to the specific interests of neurosurgeons and residents. Use clinical language and case studies that they can relate to. - Highlight Local Relevance: Discuss AI research and applications that are specific to the Valencia and Murcia regions to make the content more relatable and engaging.

### Resources to Include

- Research Papers and Articles: Provide a list of recommended readings for further exploration. - Online Tools and Demos: Share links to online AI tools or demo platforms where they can experiment with AI applications. - Networking Opportunities: Mention any upcoming conferences, workshops, or seminars in the Valencia and Murcia regions focused on AI in healthcare.

This structure should provide a comprehensive overview of AI in neurosurgery, tailored to your audience's needs and interests.

Preoperative Planning:

AI can assist in the planning of neurosurgical procedures by providing detailed anatomical information from medical images (e.g., MRI and CT scans).

AI can help identify and visualize critical structures and regions within the brain, aiding in surgical strategy development.

Image Analysis:

AI algorithms can analyze medical images to detect and classify neurological conditions, such as tumors, vascular malformations, and aneurysms.

Computer vision models can assist in the segmentation and 3D reconstruction of brain structures, improving visualization for surgeons.

Surgical Navigation:

AI-powered navigation systems can help guide surgeons during surgery, allowing for more precise targeting of lesions or abnormalities. These systems can take real-time imaging data and overlay it on the patient's anatomy, providing visual guidance. Robot-Assisted Surgery:

AI-controlled robotic systems are used to perform delicate neurosurgical procedures with a high degree of precision. These robots can stabilize the surgeon's hand movements and enable minimally invasive techniques. Risk Assessment and Predictive Models:

AI can assess patient data to predict surgical outcomes and identify potential complications. It can also help estimate patient-specific risk factors, which aids in decision-making and patient counseling. Intraoperative Monitoring:

AI can continuously monitor patients' vital signs, anesthesia levels, and other critical parameters during surgery. It can provide real-time alerts to surgical teams if any anomalies are detected. Postoperative Care and Monitoring:

AI systems can assist in postoperative monitoring and recovery management. They can track patient progress, provide decision support, and help with personalized rehabilitation plans. Data Analytics and Research:

AI helps analyze vast amounts of medical data, including patient records and clinical trials, to identify patterns and trends in neurological conditions. It aids in the development of new treatment approaches and therapies. Telemedicine and Remote Consultations:

AI-powered telemedicine platforms enable neurosurgeons to offer remote consultations and second opinions to patients and colleagues. This is especially valuable for patients in remote areas or those with limited access to specialized care. Education and Training:

AI can be used in virtual reality (VR) and augmented reality (AR) simulations for training neurosurgeons. These platforms offer a safe environment to practice surgical techniques and decision-making. Automated Documentation:

AI can assist in automating clinical documentation, reducing the administrative burden on healthcare professionals and improving record-keeping accuracy. AI in neurosurgery has the potential to enhance patient outcomes, increase the precision of surgical procedures, and improve the overall quality of care. While the integration of AI technologies in neurosurgery is ongoing, it is important to ensure that these tools are rigorously tested, validated, and integrated safely into clinical practice to maximize their benefits and minimize risks.

Previous studies have used artificial intelligence to attempt to expedite the diagnosis of Intracranial hemorrhage pathology on neuroimaging. However, these studies have used local, institution-specific data for training of networks that limit deployment of across broader hospital networks or regions because of data biases.

To demonstrate the creation of a neural network based on an openly available imaging data tested on data from our institution demonstrating a high-efficacy, institution-agnostic network.

A data set was created from publicly available noncontrast computed tomography images of known ICH. These data were used to train a neural network using distinct windowing and augmentation. This network was then validated in 2 phases using cohort-based (phase 1) and longitudinal (phase 2) approaches.

The convolutional neural network was trained on 752 807 openly available slices, which included 112 762 slices containing intracranial hemorrhage. In phase 1, the final network performance for intracranial hemorrhage showed a receiver operating characteristic curve (AUC) of 0.99. At the inflection point, our model showed a sensitivity of 98% at a threshold specificity of 99%. In phase 2, we obtained an AUC of 0.98 after analysis of 726 scans with a negative predictive value of 99.70% (n = 726).

Hopkins et al. demonstrated an effective neural network trained on completely open data for screening ICH at an unrelated institution. This study demonstrates a proof of concept for screening networks for multiple sites while maintaining high efficacy ⁵⁾.

Reliable intraoperative delineation of tumor from healthy brain tissue is essentially based on the neurosurgeon's visual aspect and tactile impression of the considered tissue, which is-due to inherent low brain consistency contrast-a challenging task. Development of an intelligent artificial intraoperative tactile perception will be a relevant task to improve the safety during surgery, especially when-as for neuroendoscopy-tactile perception will be damped or-as for surgical robotic applications-will not be a priori existent. Here, we present the enhancements and the evaluation of a tactile sensor based on the use of a piezoelectric tactile sensor.

METHODS: A robotic-driven piezoelectric bimorph sensor was excited using multisine to obtain the frequency response function of the contact between the sensor and fresh ex vivo porcine tissue probes. Based on load-depth, relaxation and creep response tests, viscoelastic parameters E1 and E2 for the elastic moduli and η for the viscosity coefficient have been obtained allowing tissue classification. Data analysis was performed by a multivariate cluster algorithm.

RESULTS: Cluster algorithm assigned five clusters for the assignment of white matter, basal ganglia and thalamus probes. Basal ganglia and white matter have been assigned to a common cluster, revealing a less discriminatory power for these tissue types, whereas thalamus was exclusively delineated; gray matter could even be separated in subclusters.

CONCLUSIONS: Bimorph-based, multisine-excited tactile sensors reveal a high sensitivity in ex vivo tissue-type differentiation. Although, the sensor principle has to be further evaluated, these data are promising ⁶⁾.

The creation of medical notes in software applications poses an intrinsic problem in workflow as the technology inherently intervenes in the processes of collecting and assembling information, as well as the production of a data-driven note that meets both individual and healthcare system requirements. In addition, the note writing applications in currently available electronic health records (EHRs) do not function to support decision making to any substantial degree.

Deliberato et al. suggest that artificial intelligence (AI) could be utilized to facilitate the workflows of the data collection and assembly processes, as well as to support the development of personalized, yet data-driven assessments and plans ⁷⁾.

Mathematically modeling are used in the natural sciences (such as physics, biology, earth science, meteorology) and engineering disciplines (such as computer science, artificial intelligence), as well as in the social sciences (such as economics, psychology, sociology, political science).

Machine learning (ML) is a domain of artificial intelligence that allows computer algorithms to learn from experience without being explicitly programmed.

What can the hypotheses of Artificial Intelligence (AI) that brains work like computers contribute to the question whether neurotransplantations are permissible or not. My attitude is primarily critical. I point out that the believe that we could exchange parts of the brain like parts of a computer is erroneous. Many mathematical functions are not absolute but only relative computable. The computation of the latter is only possible by using an oracle (a stock of knowledge), which has to be implemented in the computer respectively in the brain. Hence, one should know in advance, before a transplantation is performed, in which part of the brain the oracle is located. Otherwise we would not know whether the oracle will be damaged by a neurotransplantation, and, hence, whether the persons ability to think will be changed. Of course, this is only a presupposition, not the solution of the ethical question of the legitimacy of neurotransplantations ⁸⁾.

Panesar S, Cagle Y, Chander D, Morey J, Fernandez-Miranda J, Kliot M. Artificial Intelligence and the Future of Surgical Robotics. Ann Surg. 2019 Mar 19. doi: 10.1097/SLA.0000000000003262. [Epub ahead of print] PubMed PMID: 30907754.