You are here

  • Home
  • Archive
  • Volume 15, Issue 10
  • Response to ‘Application of deep learning models for detection of subdural hematoma: a systematic review and meta-analysis’

Article Text

Article menu
Download PDFPDF
Letter
Response to ‘Application of deep learning models for detection of subdural hematoma: a systematic review and meta-analysis’
  1. Saeed Abdollahifard1,2,
  2. Amirmohammad Farrokhi1,2,
  3. Ashkan Mowla3
  1. 1 School of Medicine, Shiraz University of Medical Sciences, Shiraz, Iran
  2. 2 Research Center for Neuromodulation and Pain, Shiraz, Iran
  3. 3 Division of Stroke and Endovascular Neurosurgery, Department of Neurological Surgery, Keck School of Medicine University of Southern California (USC), Los Angeles, California, USA
  1. Correspondence to Dr Ashkan Mowla, Department of Neurological Surgery, Keck School of Medicine University of Southern California (USC), Los Angeles, CA 90033, USA; mowla{at}usc.edu

https://doi.org/10.1136/jnis-2023-020804

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    We truly appreciate readers' concerns1 about our article ‘Application of deep learning models for detection of subdural hematoma: a systematic review and meta-analysis’.2 While our objective differed from what readers might have expected, we have taken their feedback into account.

    The authors are aware of the differences between segmentation and detection accuracies, but in this study, we pooled the sensitivity and specificity of the models to highlight the effectiveness of artificial intelligence (AI) models in detecting subdural hematomas (SDH) in order to primarily explore the use of such models both in screening and detection. The main goal of the article was not to evaluate the performance of various AI models individually but to increase awareness about how they can assist physicians as a relatively new concept introduced. The authors emphasize that considering the accuracy rates of these AI tools, they might have the potential to revolutionize the future of neurosurgery and neurology.3 4 It is noteworthy that pooling the results on this matter could be useful to further expand research and development in the application of AI. The authors agree with the technical comment made by the readers, and future reviews should distinguish between the segmentation and detection tasks and pool the accuracy rate …

    View Full Text

    Footnotes

    • Contributors All authors contributed to developing the idea, writing the draft, and finalizing the manuscript.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests AM: Speakers bureau/consultant to Cerenovus, Stryker, Wallaby Medical, RapidAI, BALT USA, LLC. Others have no disclosure.

    • Provenance and peer review Not commissioned; internally peer reviewed.

    Linked Articles