Complications research in neurosurgery using artificial intelligence

Quantitative analysis of complications in neurosurgery is a permanently topical issue limited by the lack of a unified «complication» definition in the professional community, approaches to registering, surveillance, and classification of complications. These drawbacks, in turn, disable quantitative safety indicators comparison between different neurosurgical facilities, as well as large-scale epidemiological studies.

PURPOSE OF THE STUDY: to determine the actual tasks of studying complications in neurosurgery, taking into account the potential of modern artificial intelligence technologies and its usage examples in the research of the FSAI «NMRC named after ac. N. N. Burdenko» of the ministry of health of Russia.

MATERIALS AND METHODS: In the present work, we analyzed text medical records that were initially typed in using keyboard by doctors into the electronic health records system “E-med” of the FSAI «NMRC named after ac. N. N. Burdenko» of the ministry of health of Russia in the period between 2000 and 2017. Textual information was analyzed via natural language processing.

RESULTS. We outlined the main goals in complication analysis, highlighted the power of artificial intelligence to ground the definition of «complications in neurosurgery» scientifically, to study the spectrum of complications, to identify, and predict them, and to define related risk factors.

CONCLUSION. Modern artificial intelligence technologies have a significant potential for application in complications research in neurosurgery.

1. Ziewacz JE, Mcgirt MJ, Chewning SJJ. Adverse events in neurosurgery and their relationship to quality improvement. Neurosurg Clin N Am. 2015;26(2):157–165, vii. doi:10.1016/j.nec.2014.11.014

2. Brock S, Saleh C, Zekaj E, Servello D. How to compare clinical results of different neurosurgical centers? Is a classification of complications in neurosurgery necessary for this purpose? Surg Neurol Int. 2016;7(Suppl 20): S 565–9. doi:10.4103/2152–7806.188471

3. Danilov GV, Shifrin MA, Kotik KV, et al. Artificial intelligence in neurosurgery: a systematic review using topic modeling. part I: Major research areas. Sovrem Tehnol v Med. 2020;12(5). doi:10.17691/stm2020.12.5.12

4. Danilov G, Board Secretary S, Orlov Y, Kulikov A, Anesthesiologist S, Potapov AS. Artificial Intelligence Technologies in Neurosurgery: a Systematic literature review Using Topic Modeling. part II: research objectives and perspectives. 2020;12(6): reviews. doi:10.17691/stm2020.12.6.12

5. Danilov G, Shifrin M, Strunina U, Pronkina T, Potapov A. An Information extraction algorithm for detecting adverse eventsin Neurosurgery Using documents written in a Natural rich-in-Morphology language. Stud Health Technol Inform. 2019;262:194–197. doi:10.3233/SHTI190051

6. Danilov G. V., Shifrin M. A., Potapov A. A., et al. Method for extracting information from unstructured texts written in natural language (patent rU 2751993C 1). published online july 21, 2021. accessed december 27, 2021. https://www.fips.ru/registers-doc-view/fips_servlet?dB=rUpaT&docNumber=2751993&TypeFile=html (In russ.).

7. Danilov G, Kosyrkova A, Shults M, et al. Inter-rater reliability of Unstructured Text labeling: artificially vs. Naturally Intelligent approaches. Stud Health Technol Inform. 2021;281:118–122. doi:10.3233/SHTI210132