Abstract

The article explores the burgeoning role of Artificial Intelligence (AI) in healthcare, particularly in anesthesia and surgical practices, with the ultimate aim of enhancing patient outcomes. It underscores the necessity for refining AI algorithms while addressing legal and ethical concerns. AI's potential applications in healthcare are manifold, including predicting perioperative risks, detecting intraoperative events, and identifying postoperative complications for early intervention. The integration of AI in the operating room (OR) aims to augment human capabilities rather than replace healthcare professionals, thereby improving surgical safety and outcomes.

Significantly, AI is posited to enhance efficiency and quality of care in healthcare settings. This includes automating referral management in anesthesia clinics, which face challenges like referral backlogs and labor-intensive processes. AI can streamline these processes, reduce paperwork, alleviate patient anxiety, and provide real-time feedback for more accurate and timely interventions. The article also highlights various AI applications in anesthesia, such as personalized anesthetic management, vital sign monitoring, and trend analysis in anesthesia practice.

Additionally, the article delves into AI's transformative potential in pharmaceutical research, particularly in Central Nervous System (CNS) therapeutics. It mentions a study named "ADVENTURE" by the University Hospital, Strasbourg, focusing on using AI for classifying and analyzing adverse events in anesthesia.

Furthermore, the use of AI in pediatric anesthesia for preoperative assessment, risk stratification, and managing various intraoperative challenges is discussed. AI's impact on reducing MRI scan times and enhancing ultrasound-guided regional anesthesia is also highlighted.

The article concludes with a discussion on the implementation of AI in healthcare. It emphasizes the need for accurate, diverse data sets and robust governance for successful AI integration. The goal is to streamline clinic operations, improve patient care quality, and increase patient satisfaction while ensuring AI's role as an aid, not a substitute, in clinical judgment.

Introduction

AI is emerging as a powerful tool in healthcare, with the ultimate aim of achieving better patient outcomes. To reach this goal, algorithm development must be refined and address any legal and ethical issues. AI could be used to accurately predict perioperative risks related to both anesthesia and surgery, as well as intraoperative events such as hypotension or delays in surgical steps. AI could also be used to identify the risk of postoperative complications such as sepsis or renal failure, and plan for early intervention. The use of AI in the OR is not intended to replace the surgeon or the anesthesiologist, but to expand human capacity and capability through enhanced vision, dexterity, and complementary machine intelligence for improved surgical safety and outcomes ^[1]. AI has the potential to improve efficiency and productivity for healthcare workers, while also enhancing the quality of care provided to patients ^[2,3]. Anesthesia clinics are often overwhelmed with referrals, many of which may be inappropriate or incomplete, resulting in a backlog of cases and a labor-intensive process for assessing referrals and ensuring necessary investigations and labs are done ^[4,5]. Artificial intelligence has the potential to impact the clinical practice of anesthesia in aspects ranging from perioperative support to critical care delivery to outpatient pain management. As research efforts advance and technology development intensifies, it is essential for practicing clinicians to provide practice-based insights to assist in the clinical translation of artificial intelligence ^[6].

Discussion

Automating an AI-assisted referral management system in Anesthesia ^[7] can streamline the referral process, automatically identifying appropriate referrals for further assessment and those that should be sent back to the referring surgeon for further information. Additionally, it will provide patients with information about anesthesia, surgery, and recovery before the clinic appointment, reducing time spent on paperwork, improving clinic efficiency, and reducing anxieties for patients. Furthermore, AI-assisted referral management systems can also be used to monitor patient progress and provide real-time feedback to clinicians, allowing for more accurate and timely interventions. In addition to the AI-assisted referral management system discussed above, there are several other potential applications of AI in anesthesia practice that could further improve efficiency and patient outcomes.

For example, AI-assisted decision support systems can be used to provide personalized recommendations for anesthetic management, helping to reduce the risk of adverse events and improve patient safety. AI-assisted systems can also be used to monitor vital signs and alert clinicians to any changes in patient status, allowing for more timely interventions. Finally, AI-assisted systems can be used to analyze patient data and provide insights into trends in anesthesia practice, helping to identify areas for improvement and optimize patient care ^[8,9].

Here are few examples:

1. Automated drug delivery: AI-powered algorithms could be used to automate the delivery of anesthesia medication to patients, ensuring that the correct dose is administered at the right time. This could reduce the risk of medication errors and improve patient safety.
2. Automated patient monitoring: AI-powered systems could be used to monitor patient vital signs in real-time, providing early warning alerts to caregivers if any changes in patient condition are detected.
3. Automated anesthesia documentation: AI-powered natural language processing (NLP) systems could be used to automatically generate structured anesthesia reports, saving time and reducing errors in documentation.
4. Automated anesthesia technique selection: AI-powered decision-making tools could be used to help anesthesiologists choose the most appropriate anesthesia technique for each patient based on factors such as patient characteristics, medical history, and the nature of the procedure.
5. Virtual Anesthesia Assistants: A virtual assistant that can help anesthesiologists with tasks such as monitoring patient vital signs, administering medication, providing reminders for essential tasks, and providing real-time feedback on patient conditions.

Artificial intelligence (AI) and machine learning (ML) have the potential to revolutionize pharmaceutical research by extracting novel and important insights from the vast amount of complex data generated from the drug discovery process. In recent years, AI/ML-based methods have been widely applied to many therapeutic areas and achieved state-of-the-art performance in addressing diverse problems in drug discovery. AI/ML algorithms have also shown promise in the development of CNS therapeutics, the most challenging area in drug discovery. AI/ML will play an increasingly critical role in CNS drug discovery towards personalized medicine, particularly in the areas of patient subtyping, identification of key disease drivers, prediction of cell type-specific drug response, autonomous design of novel drugs, and disease-specific BBB permeability testing.

Recently, University Hospital, Strasbourg, France, have completed a trial which they called “Analysis of Adverse Events in Anesthesia Using Artificial Intelligence (ADVENTURE)”. The aim of this research was to develop a natural language processing approach to establish a classification of adverse events observed during the perioperative period and to facilitate their analysis. The main objective of their study was to identify what a “naive” unsupervised model would discover based on Adverse Event (AE) descriptions. Their second goal was to identify apparently unrelated events whose combination could favor the occurrence of an AE. The results of the trial which was completed end of Jan 2023 are not published yet ^[11].

It is clear that AI/ML technologies have the potential to revolutionize the field of neuropharmacology and provide more effective treatments for CNS diseases. However, it is important to note that the successful implementation of these AI technologies would require the use of large, accurate, and diverse data sets, as well as careful validation and testing to ensure that the systems are safe and effective. Additionally, robust data governance, security, and privacy protocols must be in place to protect patient information ^[12]. The integration of AI in anesthesia practice can bring many benefits but it is important to keep in mind that it should be used as an aid and not a replacement for clinical judgment.

Ideas

Using supervised and customized machine learning to analyze historical data from prior clinics would enable the clinic to identify high-risk cases, prioritize them, and provide a better care plan. By analyzing patterns in referral letters, accompanied charts, and anesthesia clinic outcomes, an AI-powered system would be able to categorize patients based on their characteristics, such as sex, age, and ASA status. Medicine is a complex and demanding domain for the implementation of computer-based advisors. There is a broad range of different types of information that can be brought to bear on medical problems, including fundamental biomedical principles, clinical observations linking disease states to clinical findings, anecdotal case-based knowledge, and critical interpretation of the clinical literature. Granting the complexity and diversity of medical knowledge, it should not be surprising that the development of robust, sophisticated computer-based advisors poses challenging problems ^[13]. Lonsdale et al and colleagues have proposed the use of AI for the following purposes in Pediatric Anesthesia Preoperative Assessment and Risk Stratification: Machine Learning and Anesthesia Procedures (vascular access, video-based systems or sonography use, nerve blocks), Airway Management and Monitoring, Prediction of Bleeding Risk, Planning of Intraoperative Transfusion Requirements, Predicting Complications following Blood Transfusion, Prediction of Mortality, Need for PICU Admission, Prediction of Cardiac Arrest during Critical Care Unit Admission ^[14]. Additionally, AI-powered systems can be used to identify and predict potential adverse events, such as anaphylaxis, and to provide personalized care plans for high-risk patients. By leveraging the power of AI, the clinic can ensure that the right care is provided to the right patient at the right time.

The pediatric population often requires sedation or general anesthesia to undergo an MRI exam, which can be a lengthy and costly process. However, thanks to a new AI-powered algorithm developed by Dr Qiang Zhang of the Radcliffe Department of Medicine, the time spent in an MRI scanner can be drastically reduced from 30-45 minutes to just 15 minutes. This not only cuts the cost of the scan, but also produces clearer and more interpretable images. Even more importantly, this technology can reduce the need for sedatives, anesthesia, and contrast, and the associated side effects and reactions. Mark R Sullivan was ahead of his time in recognizing the potential of mobile phones, and now AI systems in healthcare have the potential to have a similar impact, particularly in the field of ultrasound-guided regional anaesthesia. AI technologies can help optimize performance, increase uptake, and standardize training, and anaesthetists should take advantage of this opportunity to ensure that these technologies are used to enhance the specialty in a meaningful way.

Results

The implementation of the proposed AI platform would reduce the number of unnecessary referrals to the clinic, increase the time available for specialists and nurses to focus on providing the highest quality care, and improve the quality of care provided. Additionally, this approach would allow clinics to target the most complex cases and provide better care to the patients, by knowing their specific characteristics, increase patient satisfaction and reduce anxieties. By automating and streamlining the referral process, anesthesia clinics can reduce their workload, improve clinic efficiency, and provide a higher level of care to patients.

Conclusion

It's important to note that the implementation should be done carefully to ensure that the data used is accurate and up-to-date, and that the system is trained on a diverse patient population to generalize well on new cases. Additionally, the use of AI should be monitored to ensure that it is providing the desired results and that the quality of care is not being compromised.

Abbreviations

AI: Artificial Intelligence

OR: Operating Room

CNS: Central Nervous System

MRI: Magnetic Resonance Imaging

NLP: Natural Language Processing

ML: Machine Learning

ADVENTURE: Analysis of Adverse Events in Anesthesia Using Artificial Intelligence

PICU: Pediatric Intensive Care Unit

Declarations

Ethical Approval and Consent to participate

Not applicable

Consent for publication

Not applicable

Availability of supporting data

Peer review Articles

Competing interests

None to declare

Funding

None to declare

Authors' contributions

None to declare

References

1. Artificial intelligence in perioperative management of major gastrointestinal surgeries, Sohan Lal Solanki 1, Saneya Pandrowala 2, Abhirup Nayak 3, Manish Bhandare 2, Reshma P Ambulkar 3, Shailesh V Shrikhande 2, PMID: 34135552 PMCID: PMC8173379. DOI: 10.3748/wjg.v27.i21.2758
2. Artificial Intelligence in Healthcare, Ivana Ognjanovic, PMID: 32990674 DOI:10.3233/SHTI200677
3. HealthTech Feb 2021, AI Can Increase Efficiency in Healthcare, even in a Pandemic, David Lubarsky and Nick Culbertson. Preoperative Consultations, Stephan R Thilen, Duminda N Wijeysundera, Miriam M Treggiari, PMID: 26927736 DOI:10.1016/j.anclin.2015.10.003
4. Referral Systems Development and Survey of Perioperative and Critical Care Referral to Anesthetists, P L Narendra, Harihar V Hegde, Maroof Ahmad Khan, Dayanand G Talikoti, Samson Nallamilli, PMID: 28928575 PMCID: PMC5594794 DOI: 10.4103/0259-1162.207072
5. Design and development of referrals automation, a SMART on FHIR solution to improve patient access to specialty care; Anobel Y Odisho, Hansen Lui, Ramakrishna Yerramsetty, Felicisimo Bautista, Nathaniel Gleason, Edwin Martin, Jerry J Young, Michael Blum, Aaron B Neinstein;PMID: 33215075 PMCID: PMC7660949 DOI: 10.1093/jamiaopen/ooaa036
6. Anesthesiology. Author manuscript; available in PMC 2020 Nov 5., Published in final edited form as: Anesthesiology. 2020 Feb; 132(2): 379–394; doi: 10.1097/ALN.0000000000002960
7. Artificial Intelligence in Anesthesiology: Current Techniques, Clinical Applications, and Limitations; Daniel A Hashimoto, Elan Witkowski, Lei Gao, Ozanan Meireles, Guy Rosman, PMID: 31939856 PMCID: PMC7643051 DOI: 10.1097/ALN.0000000000002960
8. Application of Machine Learning Algorithm in Anesthesia, Jin Xu, Xiaoming Deng, Fuxia Yan, PMID: 33131527 DOI: 10.3881/j.issn.1000-503X.12356
9. Machine-Learning Implementation in Clinical Anesthesia: Opportunities and Challenges; Char, Danton S. MD; Burgart, Alyssa MD; Anesthesia & Analgesia 130(6):p 1709-1712, June 2020. | DOI: 10.1213/ANE.0000000000004656
10. Artificial intelligence and machine learning‐aided drug discovery in central nervous system diseases: State‐of‐the‐arts and future directions, Medd Res Rev. 2021 May; 41(3): 1427–1473; Published online 2020 Dec 9. doi: 10.1002/med.21764; PMCID: PMC8043990; NIHMSID: NIHMS1650992; PMID: 33295676
11. Paul-Michel MERTES, MD, PhD, Contact: Said CHAYER, PhD, HDR University Hospital, Strasbourg, France, ClinicalTrials.gov Identifier: NCT05185479
12. Artificial intelligence and anesthesia: A narrative review; Madhavi Singh, Gita Nath. PMID: 35261595 PMCID: PMC8846233 DOI: 10.4103/sja.sja_669_21
13. Artificial Intelligence in Healthcare. 2020: 295–336; Published online 2020 Jun 26. doi: 10.1016/B978-0-12-818438-7.00012-5; PMCID: PMC7332220; Ethical and legal challenges of artificial intelligence-driven healthcare; Sara Gerke, Timo Minssen, and Glenn Cohen
14. Machine Learning and Artificial Intelligence in Pediatric Research: Current State, Future Prospects, and Examples in Perioperative and Critical Care; Hannah Lonsdale MBChB, Ali Jalali PhD, Luis Ahumada PhD and Clyde Matava MD; Journal of Pediatrics, The, 2020-06- 01, Volume 221, Pages S3-S10
15. Dr Qiang Zhang Scientist, Radcliffe Department of Medicine; World Economic Forum, annual meeting, 10 Oct 2022
16. Artificial intelligence for image interpretation in ultrasound-guided regional anaesthesia, J. Bowness, K. El-Boghdadly, D. Burckett-St Laurent, First published: 29 July 2020 https://doi.org/10.1111/anae.15212