{"paper_id":"ae2f5c23-2622-43d5-9fec-8186d728e703","body_text":"Interv Pain Med Neuromod. 2022 December; 2(1):e128720.\nPublished online 2022 August 10.\ndoi: 10.5812/ipmn-128720.\nLetter\nIs Artificial Intelligence a New Diagnostic Approach for Patients with\nEndometriosis?\nBehnaz Nouri\n 1 and Siavash Roshandel 2, *\n1The Preventative Gynecology Research Center, Shahid Beheshti University of Medical Sciences, Tehran, Iran\n2Iran University of Science and Technology , Tehran, Iran\n*Corresponding author: Iran University of Science and Technology , Tehran, Iran. Email: s_roshandel@mecheng.iust.ac.ir\nReceived 2022 June 04; Accepted 2022 June 11.\nKeywords: Artificial Intelligence, Endometriosis, Laparoscopy\nDear Editor,\nNowadays, artificial intelligence (AI) is increasingly\nused in the medical industry , primarily machine learning\n(ML) as one element of AI (1). Specifically , AI has been sug-\ngested in the gynecological field, especially imaging and\nultrasound, to assess the uterus (2) and classify ovarian\ncysts. As far as we know , no studies have analyzed the pos-\nsible function of AI in diagnosing endometriosis. Artifi-\ncial intelligence technologies are promising in medicine\nto help physicians with diagnostic and therapeutic prob-\nlems.\nThe introduction of AI in midwifery and gynecology\nhas shown great potential. Artificial intelligence for in-\nterpreting cardiotocography (CTG) has increased its sen-\nsitivity from 60% to 94%, with a specificity of 91% (1). In\nbreast imaging, AI is part of everyday clinical practice. Arti-\nficial intelligence showed a 5.7% decrease in false-positive\nresults and a 4.9% decrease in false-negative results in\nmammography screening interpretation, according to US-\nlicensed radiologists participating in the McKinney et al.’s\nstudy (2). Hence, it is inevitable that AI will lead to non-\ninvasive diagnoses in women, especially endometriosis (2,\n3).\nThe gold standard for diagnosing endometriosis is\nminimally invasive laparoscopy . This procedure is often\ndifficult, costly , and time-consuming for even the most\nexperienced laparoscopic surgeons, as the patient’s con-\ndition and characteristics may prevent accurate assess-\nment of anatomical structures. Laparoscopic surgery can\ncause dangerous complications and health problems for\nwomen. The introduction of AI can be a significant step in\ndiagnosing endometriosis non-invasively (4).\nFuture AI research in diagnosing endometriosis will\nlikely focus on improving machine learning algorithms.\nThis will be achieved through managing datasets, properly\nusing convolutional neural networks and repetitive neu-\nral networks, and integrating several models into one end-\nto-end model. Finally , validating machine learning algo-\nrithms with large heterogeneous datasets ensures the gen-\neralization of results to different populations. Artificial in-\ntelligence seems to be the first step in the diagnosis of en-\ndometriosis. Inevitably , the use of artificial intelligence in\nthe diagnosis of endometriosis raises essential questions.\nDoes data interpretation require human supervision, or\ncan we rely on AI? Who is responsible for misdiagnosis?\nDoes relying on automation provided by AI destroy tech-\nnical skills and experience? Does AI reduce the workload\nof medical professionals or replace them altogether?\nThe application dimensions of AI in health care are\nonly recently becoming apparent, and the scientific com-\nmunity is only beginning to address many of the critical\nissues that arise. In the study of Guerriero et al., 106 women\nwith a final diagnosis of rectosigmoid endometriosis were\nstudied using AI (3). Regarding diagnostic accuracy , the\nneural network model was the best (accuracy 0.73, sensitiv-\nity 0.72, specificity 0.73, positive predictive value 0.52, and\nnegative predictive value 0.86). The accuracy of ultrasound\nmarkers in suspecting rectosigmoid endometriosis using\nAI models had similar results to the logistic model (3). The\nsensitivity and specificity of AI in the study of Bendifallah\net al. for the diagnosis of endometriosis ranged from 0.82\nto 1.0 and 0.91 to 0.95, respectively (5).\nData from a few studies worldwide show that AI can\nbe a good screening test for clinicians (6, 7). The introduc-\ntion of ML in these settings represents a significant change\nin medical evaluation because AI can replace diagnostic la-\nparoscopy .\nCopyright © 2022, Interventional Pain Medicine and Neuromodulation. This is an open-access article distributed under the terms of the Creative Commons\nAttribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/) which permits copy and redistribute the material just in\nnoncommercial usages, provided the original work is properly cited.\n\nNouri B and Roshandel S\nFootnotes\nAuthors’ Contribution: B. N. wrote the manuscript in\nconsultation with S. R. S. R. helped to draft the manuscript.\nConflict of Interests: The authors certify that they have\nno affiliations with or involvement in any organization or\nentity with any financial or non-financial interest in the\nsubject matter or materials discussed in the manuscript.\nFunding/Support: The author(s) received no financial\nsupport for the research, authorship, and/or publication of\nthis article.\nReferences\n1. Psarris A, Syndos M, Daskalakis G, Loutradis D. Fetal Ultrasonography:\nIs artificial intelligence the way forward? Letter to the Editor. Hell J\nObstet Gynecol. 2020;19(2):105–7. doi: 10.33574/hjog.1995.\n2. McKinney SM, Sieniek M, Godbole V, Godwin J, Antropova N, Ashrafian\nH, et al. International evaluation of an AI system for breast cancer\nscreening. Nature. 2020;577(7788):89–94. doi: 10.1038/s41586-019-1799-\n6. [PubMed: 31894144].\n3. Guerriero S, Pascual M, Ajossa S, Neri M, Musa E, Graupera B, et\nal. Artificial intelligence (AI) in the detection of rectosigmoid deep\nendometriosis. Eur J Obstet Gynecol Reprod Biol . 2021; 261:29–33. doi:\n10.1016/j.ejogrb.2021.04.012. [PubMed: 33873085].\n4. Sarbazi F, Akbari E, Karimi A, Nouri B, Noori Ardebili SH. The Clinical\nOutcome of Laparoscopic Surgery for Endometriosis on Pain, Ovar-\nian Reserve, and Cancer Antigen 125 (CA-125): A Cohort Study .Int J Fer-\ntil Steril. 2021;15(4):275–9. doi: 10.22074/IJFS.2021.137035.1018. [PubMed:\n34913296]. [PubMed Central: PMC8530215].\n5. Bendifallah S, Puchar A, Suisse S, Delbos L, Poilblanc M, Descamps P,\net al. Machine learning algorithms as new screening approach for pa-\ntients with endometriosis. Sci Rep. 2022;12(1):639. doi: 10.1038/s41598-\n021-04637-2. [PubMed: 35022502]. [PubMed Central: PMC8755739].\n6. Shaterian N, Soltani A, Samieefar N, Akhlaghdoust M. Artificial Intel-\nligence and Migraine: Insights and Applications. Interv Pain Med Neu-\nromod. 2022;2(1). e127675. doi: 10.5812/ipmn-127675.\n7. Sarbazi F, Akbari E, Nouri B. Pain Management in Endometriosis. In-\nterv Pain Med Neuromod. 2022;2(1). e128043. doi: 10.5812/ipmn-128043.\n2 Interv Pain Med Neuromod. 2022; 2(1):e128720.","source_license":"CC0","license_restricted":false}