Makine Öğrenimi, Evrişimli Sinir Ağları ve Anestezi
Özet
Bu çalışma, yapay zekâ (AI), makine öğrenimi (ML) ve evrişimli sinir ağlarının (CNN) sağlık sektörüne, özellikle de anesteziyoloji ve yoğun bakım alanlarına olan etkilerini incelemektedir. İnsan zihninin yorgunluk ve bilişsel hata payı gibi kısıtlamalarına karşılık bilgisayar bilimlerindeki donanım ve depolama ilerlemeleri, yapay zekânın tıbbın çeşitli alanlarında tanısal ve girişimsel uygulamalar bulmasını sağlamıştır. Tıbbi uygulamaların büyük kısmı denetimli öğrenme kapsamına girerken; denetimsiz ve pekiştirmeli öğrenme modelleri de ilaç sınıflandırması ve anestezi denetleyicileri gibi süreçlerde kullanılmaktadır. Anesteziyolojide AI; intraoperatif ortamda hipoksemi ve hipotansiyon risklerinin önceden tahmini, sedasyon uygulamalarının otomasyonu ve yapay görme teknolojileriyle entübasyon ile ultrason rehberliğindeki bölgesel anestezi (sinir bloğu) süreçlerinde karar destek aracı olarak öne çıkmaktadır. Ayrıca yoğun bakım süreçlerinde sepsisin erken tanısında, mortalite risk ayarlarında ve akut böbrek hasarı gibi komplikasyonların tahmininde ML tabanlı sistemlerin başarısı kanıtlanmıştır. Yapay zekanın bu vaatlerine rağmen önünde veri paylaşımında heterojenlik, algoritmaların şeffaf olmaması ("kara kutu" sorunu), kurumsal önyargılar ve sürekli öğrenen sistemlerin kilitlenmesini gerektiren FDA gibi düzenleyici kurumların mevzuat yetersizlikleri gibi engeller bulunmaktadır. Sonuç olarak AI teknolojileri, perioperatif süreçlerden kritik bakıma kadar anestezi uzmanlarının reaktif olmaktan ziyade proaktif kararlar almasını sağlayan güçlü birer yardımcı konumundadır.
This study examines the impact of artificial intelligence (AI), machine learning (ML), and convolutional neural networks (CNN) on the healthcare sector, specifically within the fields of anesthesiology and intensive care. In contrast to the limitations of the human mind, such as fatigue and cognitive errors, major advancements in computer science regarding hardware and storage have enabled AI to find diagnostic and interventional applications across various medical domains. While the majority of medical applications fall under supervised learning, unsupervised and reinforcement learning models are also utilized in processes like drug classification and anesthesia controllers. In anesthesiology, AI stands out as a decision support tool for predicting intraoperative hypoxemia and hypotension risks, automating sedation delivery, and optimizing difficult intubation or ultrasound-guided regional anesthesia (nerve blocks) through computer vision technologies. Furthermore, the success of ML-based systems has been demonstrated in intensive care settings for the early recognition of sepsis, ICU mortality risk adjustments, and the prediction of complications such as acute kidney injury. Despite these promises, AI faces significant challenges, including a lack of interoperability in data sharing, the opacity of algorithms (the "black box" problem), institutional biases, and regulatory deficiencies from bodies like the FDA, which currently require algorithms to be locked. Consequently, AI technologies serve as powerful tools that allow anesthesiologists to make proactive rather than reactive decisions, spanning from perioperative care to critical care delivery.
Referanslar
Bellman R. An introduction to artificial intelligence : can computers think? [Internet]. Wikipedia. San Francisco: Boyd & Fraser Pub. Co.; 1978 [cited 2021 Dec 9]. Available from: https://es.wikipedia.org/wiki/Inteligencia_artificial
Kahneman Daniel. Thinking, fast and slow /Daniel Kahneman. – National Library [Internet]. [cited 2021 Dec 9]. Available from: https://www.nlb.gov.sg/biblio/14241661
Thrall JH, Li X, Li Q, Cruz C, Do S, Dreyer K, et al. Artificial Intelligence and Machine Learning in Radiology: Opportunities, Challenges, Pitfalls, and Criteria for Success. J Am Coll Radiol [Internet]. 2018 Mar 1 [cited 2021 Dec 9];15(3 Pt B):504–8. Available from: https://pubmed.ncbi.nlm.nih.gov/29402533/
Salto-Tellez M, Maxwell P, Hamilton P. Artificial intelligence—the third revolution in pathology. Histopathology [Internet]. 2019 Feb 1 [cited 2021 Dec 9];74(3):372–6. Available from: https://onlinelibrary.wiley.com/doi/full/10.1111/his.13760
Deo RC. Machine learning in medicine. Circulation [Internet]. 2015 Nov 17 [cited 2021 Dec 9];132(20):1920–30. Available from: https://www.ahajournals.org/doi/abs/10.1161/CIRCULATIONAHA.115.001593
Hashimoto DA, Rosman G, Rus D, Meireles OR. Artificial Intelligence in Surgery: Promises and Perils. Ann Surg [Internet]. 2018 Jul 1 [cited 2021 Dec 9];268(1):70–6. Available from: https://pubmed.ncbi.nlm.nih.gov/29389679/
FDA permits marketing of artificial intelligence-based device to detect certain diabetes-related eye problems | FDA [Internet]. [cited 2021 Dec 9]. Available from: https://www.fda.gov/news-events/press-announcements/fda-permits-marketing-artificial-intelligence-based-device-detect-certain-diabetes-related-eye
Artificial Intelligence A Modern Approach Third Edition.
Kendale S, Kulkarni P, Rosenberg A, Anesthesiology JW-, 2018 undefined. Supervised machine-learning predictive analytics for prediction of postinduction hypotension. pubs.asahq.org [Internet]. 2018 [cited 2021 Dec 9]; Available from: https://pubs.asahq.org/anesthesiology/article-abstract/129/4/675/19988
Wanderer J, Anesthesiology JR-, 2018 undefined. Machine learning for anesthesiologists: A primer. pubs.asahq.org [Internet]. [cited 2021 Dec 15]; Available from: https://pubs.asahq.org/anesthesiology/article-pdf/129/4/A29/385010/20181000_0-00005.pdf
Bisgin H, Liu Z, Fang H, Xu X, Tong W. Mining FDA drug labels using an unsupervised learning technique - topic modeling. BMC Bioinformatics. 2011 Oct 18;12(SUPPL. 10).
Hakonarson H, Bjornsdottir US, Halapi E, Bradfield J, Zink F, Mouy M, et al. Profiling of genes expressed in peripheral blood mononuclear cells predicts glucocorticoid sensitivity in asthma patients. Proc Natl Acad Sci U S A [Internet]. 2005 Oct 11 [cited 2021 Dec 15];102(41):14789–94. Available from: https://pubmed.ncbi.nlm.nih.gov/16203992/
Sutton R, Barto A. Reinforcement learning: An introduction. 2018 [cited 2021 Dec 9]; Available from: https://books.google.com/books?hl=tr&lr=&id=uWV0DwAAQBAJ&oi=fnd&pg=PR7&ots=miqNn3W-h1&sig=NQ23uZKpHN-G2ylYFtFVwEfuM4o
Padmanabhan R, Meskin N, Processing WH-BS, 2015 undefined. Closed-loop control of anesthesia and mean arterial pressure using reinforcement learning. Elsevier [Internet]. 2015 [cited 2021 Dec 9];22:54–64. Available from: https://www.sciencedirect.com/science/article/pii/S1746809415000920
notes AN-CL, 2000 undefined. CS229 Lecture notes. backspaces.net [Internet]. [cited 2021 Dec 15]; Available from: http://backspaces.net/temp/ML/CS229.pdf
systems LZ-F sets and, 1978 undefined. Fuzzy sets as a basis for a theory of possibility. Elsevier [Internet]. [cited 2021 Dec 10]; Available from: https://www.sciencedirect.com/science/article/pii/0165011478900295
Baig MM, GholamHosseini H, Kouzani A, Harrison MJ. Anaesthesia monitoring using fuzzy logic. J Clin Monit Comput. 2011 Oct;25(5):339–47.
Hashimoto DA, Witkowski E, Gao L, Meireles O, Rosman G. Artificial Intelligence in Anesthesiology: Current Techniques, Clinical Applications, and Limitations. Anesthesiology [Internet]. 2020 [cited 2021 Dec 10];132(2):379. Available from: /pmc/articles/PMC7643051/
Hu YJ, Ku TH, Jan RH, Wang K, Tseng YC, Yang SF. Decision tree-based learning to predict patient controlled analgesia consumption and readjustment. BMC Med Inform Decis Mak. 2012;12:131.
McCulloch W, biophysics WP-T bulletin of mathematical, 1943 undefined. A logical calculus of the ideas immanent in nervous activity. Springer [Internet]. [cited 2021 Dec 10]; Available from: https://link.springer.com/article/10.1007%252FBF02478259
Lecun Y, Bengio Y, Hinton G. Deep learning. Nature [Internet]. 2015 May 27 [cited 2021 Dec 10];521(7553):436–44. Available from: https://pubmed.ncbi.nlm.nih.gov/26017442/
Jha S, Topol EJ. Adapting to Artificial Intelligence: Radiologists and Pathologists as Information Specialists. JAMA [Internet]. 2016 Dec 13 [cited 2021 Dec 10];316(22):2353–4. Available from: https://pubmed.ncbi.nlm.nih.gov/27898975/
Bickford RG. Automatic electroencephalographic control of general anesthesia. Electroencephalogr Clin Neurophysiol. 1950;2(1–4):93–6.
Brogi E, Cyr S, Kazan R, Giunta F, Hemmerling TM. Clinical performance and safety of closed-loop systems: A systematic review and meta-analysis of randomized controlled trials. Anesth Analg [Internet]. 2017 Feb 1 [cited 2021 Dec 12];124(2):446–55. Available from: https://journals.lww.com/anesthesia-analgesia/Fulltext/2017/02000/Clinical_Performance_and_Safety_of_Closed_Loop.15.aspx
Pasin L, Nardelli P, Pintaudi M, Greco M, Zambon M, Cabrini L, et al. Closed-Loop Delivery Systems Versus Manually Controlled Administration of Total IV Anesthesia: A Meta-analysis of Randomized Clinical Trials. Anesth Analg [Internet]. 2017 Feb 1 [cited 2021 Dec 12];124(2):456–64. Available from: https://pubmed.ncbi.nlm.nih.gov/28099320/
sedasys - U.S. Food and Drug Administration Search Results [Internet]. [cited 2021 Dec 12]. Available from: https://search.usa.gov/search?query=sedasys&affiliate=fda1
Goudra B, Singh PM. Failure of sedasys: Destiny or poor design? Anesth Analg [Internet]. 2017 Feb 1 [cited 2021 Dec 12];124(2):686–8. Available from: https://journals.lww.com/anesthesia-analgesia/Fulltext/2017/02000/Failure_of_Sedasys__Destiny_or_Poor_Design_.43.aspx
Hemmerling TM, Arbeid E, Wehbe M, Cyr S, Taddei R, Zaouter C, et al. Evaluation of a novel closed-loop total intravenous anaesthesia drug delivery system: a randomized controlled trial. BJA Br J Anaesth [Internet]. 2013 Jun 1 [cited 2021 Dec 12];110(6):1031–9. Available from: https://academic.oup.com/bja/article/110/6/1031/245822
Zaouter C, Hemmerling TM, Lanchon R, Valoti E, Remy A, Leuillet S, et al. The Feasibility of a Completely Automated Total IV Anesthesia Drug Delivery System for Cardiac Surgery. undefined. 2016 Oct 1;123(4):885–93.
Restoux A, Grassin-Delyle S, Liu N, Paugam-Burtz C, Mantz J, Le Guen M. Pilot study of closed-loop anaesthesia for liver transplantation. Br J Anaesth [Internet]. 2016 Sep 1 [cited 2021 Dec 12];117(3):332–40. Available from: https://pubmed.ncbi.nlm.nih.gov/27543528/
Joosten A, Rinehart J, Bardaji A, Van Der Linden P, Jame V, Van Obbergh L, et al. Anesthetic Management Using Multiple Closed-loop Systems and Delayed Neurocognitive Recovery: A Randomized Controlled Trial. Anesthesiology [Internet]. 2020 [cited 2021 Dec 12];132(2):253–66. Available from: https://pubmed.ncbi.nlm.nih.gov/31939839/
Culley DJ, Connor CW. Artificial Intelligence and Machine Learning in Anesthesiology. Anesthesiology [Internet]. 2019 Dec 1 [cited 2021 Dec 12];131(6):1346–59. Available from: http://pubs.asahq.org/anesthesiology/article-pdf/131/6/1346/489029/20191200_0-00029.pdf
Alexander JC, Romito BT, Çobanoǧlu MC. The present and future role of artificial intelligence and machine learning in anesthesiology. Int Anesthesiol Clin. 2020;58(4):7–16.
In Two Moves, AlphaGo and Lee Sedol Redefined the Future | WIRED [Internet]. [cited 2021 Dec 14]. Available from: https://www.wired.com/2016/03/two-moves-alphago-lee-sedol-redefined-future/
Lundberg SM, Nair B, Vavilala MS, Horibe M, Eisses MJ, Adams T, et al. Explainable machine-learning predictions for the prevention of hypoxaemia during surgery. Nat Biomed Eng [Internet]. 2018 Oct 1 [cited 2021 Dec 10];2(10):749–60. Available from: https://pubmed.ncbi.nlm.nih.gov/31001455/
Hatib F, Jian Z, Buddi S, Lee C, Settels J, Sibert K, et al. Machine-learning Algorithm to Predict Hypotension Based on High-fidelity Arterial Pressure Waveform Analysis. Anesthesiology [Internet]. 2018 [cited 2021 Dec 10];129(4):663–74. Available from: https://pubmed.ncbi.nlm.nih.gov/29894315/
Evans L, Rhodes A, Alhazzani W, Antonelli M, Coopersmith CM, French C, et al. Surviving Sepsis Campaign: International Guidelines for Management of Sepsis and Septic Shock 2021. Crit Care Med. 2021 Nov;49(11):e1063–143.
Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, Sharma S, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Vol. 34, Critical Care Medicine. 2006. p. 1589–96.
Calvert JS, Price DA, Chettipally UK, Barton CW, Feldman MD, Hoffman JL, et al. A computational approach to early sepsis detection. Comput Biol Med [Internet]. 2016 Jul 1 [cited 2021 Dec 11];74:69–73. Available from: https://pubmed.ncbi.nlm.nih.gov/27208704/
Desautels T, Calvert J, Hoffman J, Jay M, Kerem Y, Shieh L, et al. Prediction of Sepsis in the Intensive Care Unit With Minimal Electronic Health Record Data: A Machine Learning Approach. JMIR Med informatics [Internet]. 2016 Jul 1 [cited 2021 Dec 11];4(3). Available from: https://pubmed.ncbi.nlm.nih.gov/27694098/
Shimabukuro DW, Barton CW, Feldman MD, Mataraso SJ, Das R. Effect of a machine learning-based severe sepsis prediction algorithm on patient survival and hospital length of stay: a randomised clinical trial. BMJ open Respir Res [Internet]. 2017 Dec 1 [cited 2021 Dec 11];4(1). Available from: https://pubmed.ncbi.nlm.nih.gov/29435343/
Nemati S, Holder A, Razmi F, Stanley MD, Clifford GD, Buchman TG. An Interpretable Machine Learning Model for Accurate Prediction of Sepsis in the ICU. Crit Care Med [Internet]. 2018 Apr 1 [cited 2021 Dec 11];46(4):547–53. Available from: https://pubmed.ncbi.nlm.nih.gov/29286945/
Peng X, Ding Y, Wihl D, Gottesman O, Komorowski M, Lehman LWH, et al. Improving Sepsis Treatment Strategies by Combining Deep and Kernel-Based Reinforcement Learning. AMIA Annu Symp Proc [Internet]. 2018 [cited 2021 Dec 11];2018:887. Available from: /pmc/articles/PMC6371300/
Pirracchio R, Petersen ML, Carone M, Rigon MR, Chevret S, van der Laan MJ. Mortality prediction in intensive care units with the Super ICU Learner Algorithm (SICULA): a population-based study. Lancet Respir Med [Internet]. 2015 Jan 1 [cited 2021 Dec 11];3(1):42–52. Available from: https://pubmed.ncbi.nlm.nih.gov/25466337/
Delahanty RJ, Kaufman D, Jones SS. Development and Evaluation of an Automated Machine Learning Algorithm for In-Hospital Mortality Risk Adjustment Among Critical Care Patients. Crit Care Med [Internet]. 2018 Jun 1 [cited 2021 Dec 11];46(6):e481–8. Available from: https://pubmed.ncbi.nlm.nih.gov/29419557/
Parreco J, Hidalgo A, Parks JJ, Kozol R, Rattan R. Using artificial intelligence to predict prolonged mechanical ventilation and tracheostomy placement. J Surg Res [Internet]. 2018 Aug 1 [cited 2021 Dec 11];228:179–87. Available from: https://pubmed.ncbi.nlm.nih.gov/29907209/
Parreco JP, Hidalgo AE, Badilla AD, Ilyas O, Rattan R. Predicting central line-associated bloodstream infections and mortality using supervised machine learning. J Crit Care [Internet]. 2018 Jun 1 [cited 2021 Dec 11];45:156–62. Available from: https://pubmed.ncbi.nlm.nih.gov/29486341/
Tomašev N, Glorot X, Rae JW, Zielinski M, Askham H, Saraiva A, et al. A clinically applicable approach to continuous prediction of future acute kidney injury. Nature [Internet]. 2019 Aug 1 [cited 2021 Dec 11];572(7767):116–9. Available from: https://pubmed.ncbi.nlm.nih.gov/31367026/
Alderden J, Pepper GA, Wilson A, Whitney JD, Richardson S, Butcher R, et al. Predicting Pressure Injury in Critical Care Patients: A Machine-Learning Model. Am J Crit Care [Internet]. 2018 Nov 1 [cited 2021 Dec 12];27(6):461–8. Available from: https://pubmed.ncbi.nlm.nih.gov/30385537/
Cuendet GL, Schoettker P, Yüce A, Sorci M, Gao H, Perruchoud C, et al. Facial Image Analysis for Fully Automatic Prediction of Difficult Endotracheal Intubation. IEEE Trans Biomed Eng [Internet]. 2016 Feb 1 [cited 2021 Dec 14];63(2):328–9. Available from: https://pubmed.ncbi.nlm.nih.gov/26186767/
Connor CW, Segal S. Accurate classification of difficult intubation by computerized facial analysis. Anesth Analg [Internet]. 2011 [cited 2021 Dec 14];112(1):84–93. Available from: https://pubmed.ncbi.nlm.nih.gov/21081769/
Matava C, Pankiv E, Raisbeck S, Caldeira M, Alam F. IMAGE & SIGNAL PROCESSING A Convolutional Neural Network for Real Time Classification, Identification, and Labelling of Vocal Cord and Tracheal Using Laryngoscopy and Bronchoscopy Video. Available from: https://doi.org/10.1007/s10916-019-1481-4
Smistad E, Løvstakken L. Vessel Detection in Ultrasound Images Using Deep Convolutional Neural Networks. undefined. 2016;10008 LNCS:30–8.
Hetherington J, Lessoway V, Gunka V, Abolmaesumi P, Rohling R. SLIDE: automatic spine level identification system using a deep convolutional neural network. Int J Comput Assist Radiol Surg 2017 127 [Internet]. 2017 Mar 30 [cited 2021 Dec 14];12(7):1189–98. Available from: https://link.springer.com/article/10.1007/s11548-017-1575-8
Pesteie M, Lessoway V, Abolmaesumi P, Rohling RN. Automatic Localization of the Needle Target for Ultrasound-Guided Epidural Injections. IEEE Trans Med Imaging. 2018 Jan 1;37(1):81–92.
Smistad E, Johansen KF, Iversen DH, Reinertsen I. Highlighting nerves and blood vessels for ultrasound-guided axillary nerve block procedures using neural networks. J Med imaging (Bellingham, Wash) [Internet]. 2018 Nov 10 [cited 2021 Dec 14];5(4):1. Available from: https://pubmed.ncbi.nlm.nih.gov/30840734/
Huang C, Zhou Y, Tan W, Qiu Z, Zhou H, Song Y, et al. Applying deep learning in recognizing the femoral nerve block region on ultrasound images. Ann Transl Med [Internet]. 2019 Sep [cited 2021 Dec 14];7(18):453–453. Available from: /pmc/articles/PMC6803209/
Hemmerling TM. Robots Will Perform Anesthesia in the Near Future. Anesthesiology [Internet]. 2020 Feb 1 [cited 2021 Dec 14];132(2):219–20. Available from: http://pubs.asahq.org/anesthesiology/article-pdf/132/2/219/516421/20200200_0-00007.pdf
Adadi A, Berrada M. Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI). IEEE Access. 2018;6:52138–60.
Shaw J, Rudzicz F, Jamieson T, Goldfarb A. Artificial Intelligence and the Implementation Challenge. J Med Internet Res 2019;21(7)e13659 https//www.jmir.org/2019/7/e13659 [Internet]. 2019 Jul 10 [cited 2021 Dec 16];21(7):e13659. Available from: https://www.jmir.org/2019/7/e13659
Proposed Regulatory Framework for Modifications to Artificial Intelligence/Machine Learning (AI/ML)-Based Software as a Medical Device (SaMD)-Discussion Paper and Request for Feedback. [cited 2021 Dec 16]; Available from: https://www.fda.gov/downloads/medicaldevices/deviceregulationandguidance/guidancedocuments/ucm514737.pdf.
Weber G, Mandl K, Jama IK-, 2014 undefined. Finding the missing link for big biomedical data. jamanetwork.com [Internet]. 2014 [cited 2021 Dec 15]; Available from: https://jamanetwork.com/journals/jama/article-abstract/1883026
