Page 38 - Read Online

Page 426 Liu et al. Art Int Surg. 2025;5:425-33 https://dx.doi.org/10.20517/ais.2025.01

INTRODUCTION
Ventral hernias develop in up to 15% of patients after laparotomy . Closing these hernias can be difficult in
[1]
the presence of large fascial defects or loss of domain - common sequelae of prolonged use of temporary
abdominal closure. Such situations often require complex abdominal wall reconstruction. Without well-
perfused tissue coverage, complications such as mesh exposure, hernia recurrence, and even enteric fistulae
may arise . Recurrent treatment failures and multiple reoperations impose significant financial burdens on
[2]
[3,4]
healthcare systems and impact patients’ quality of life .

Over the last decade, artificial intelligence (AI) has increasingly permeated everyday life and has been
[5]
[6-9]
adopted across various medical disciplines . Data play a key role in evidence-based medicine; however, the
growing volume of data generated through electronic medical records often exceeds the capacity of
traditional analytical methods to fully harness its potential . In plastic and reconstructive surgery, AI has
[10]
demonstrated value in assessing burn severity and treatment recommendations, wound and flap
monitoring, and surgical planning [11-13] . Applying machine learning to monitor wound healing and predict
recovery trajectories could significantly enhance clinical decision making in abdominal wall
reconstruction [10,12,13] . Despite these advances, clinical integration of AI in surgical fields remains in its
infancy compared to other specialties [14,15] .

This systematic literature review aims to highlight current trends in the use of AI in abdominal wall
reconstruction.

METHODS
A systematic literature search was conducted in accordance with the preferred reporting items for
systematic reviews and meta-analyses (PRISMA) guidelines . The electronic databases PubMed, Google
[16]
Scholar, EBSCO, Ovid, and Cochrane Library were searched for studies published between 2000 and 2024
that evaluated the use of AI in abdominal wall reconstruction. The search strategy utilized Boolean
operators, Medical Subject Headings (MeSH), and synonymous terms. Keywords included: abdominal wall
reconstruction, hernia, component separation, Rives-Stoppa, diastasis, abdominal wall defect, AI, machine
learning, and large language models. Results were limited to studies published in or translated into English
and involving adult patient populations. No restrictions were applied regarding the country of publication
or patient gender. Reference lists of relevant articles were manually screened to identify additional eligible
studies.

After removing duplicates, titles and abstracts were screened to select studies for full-text evaluation.
Included publications had to report on the application of AI or machine learning specifically in the context
of abdominal wall reconstruction. Letters to the editor, editorials, review articles, and conference
proceedings were excluded. Studies focusing on groin hernias without addressing ventral or incisional
hernias were also excluded. Full texts of the remaining articles were then assessed for eligibility.

RESULTS
A total of 142 publications were identified through database searches and manual screening of reference lists
[Figure 1]. After removing duplicates, 102 articles remained for abstract screening. Of these, 55 focused on
surgical procedures other than complex abdominal wall reconstruction and were excluded from the review;
among these, 14 addressed inguinal hernias exclusively, and 2 concerned breast reconstruction. Fourteen
studies did not involve the use of AI and were therefore excluded. Thirteen articles met the inclusion criteria
for full-text review. Of these, 12 were included in the final review, while one was an expert opinion and was