Machine Learning Can Help Detect Abdominal Hernia Surgery Complications
A new machine-learning model can help clinicians detect complications with 85 percent accuracy following abdominal hernia surgery, a new study shows.
Source: Getty Images
By - Surgeons from the University of Texas MD Anderson Cancer Center in Houston have developed machine-learning models that can calculate risks for hernia recurrence and other complications.
A study that assessed the models was published in the Journal of American College of Surgeons. Researchers reviewed 725 patients who had an operation relating to ventral hernia repair between March 1, 2005, and June 30, 2019.
Researchers used these patients’ data to create nine machine-learning algorithms to determine complications that may arise after the surgery. The models took into account patient demographics and characteristics, like smoking status, as well as patient outcomes and characteristics of the procedure, such as surgical technique.
They found that the machine-learning models were 85 percent accurate for predicting hernia recurrence, 72 percent accurate for predicting surgical site occurrence, and 84 percent accurate for predicting 30-day hospital readmission.
“It’s really important for surgeons to understand what the risk factors are to abdominal wall reconstruction,” said Charles E. Butler, MD, FACS, professor and chairman at MD Anderson Cancer Center, in the press release. “It is such a common problem that surgeons have to deal with in just about every subspecialty of surgery. It puts tremendous financial, emotional, and physical strains on the healthcare system and on the patients that are affected as well as the surgeons dealing with these problems.”
Butler said that the main goal for the future would be to train the machine-learning models on a broader set of data, enabling them to widen their repertoire of complications.
“We believe the models can be improved and made to be more generalizable in subsequent iterations, and we’re currently embarking on a multicenter study to validate the models and develop a first-of-its-kind integrated tool that uses these models and clinical data and imaging data to provide a robust prediction tool,” said lead study author Abbas M. Hassan, MD, a postdoctoral fellow at the department of plastic and reconstructive surgery at MD Anderson Cancer Center.
Various other studies have also described how machine learning can be beneficial in monitoring surgical procedures.
In August 2019, Texas A&M researchers created a machine-learning system that could predict the likelihood of bleeding following heart surgery. Maintaining control of patient reactions following procedures is critical as it can help to prevent hospital readmissions. Providers hoped that this type of technology would apply risk scores more accurately and lead to higher-level care delivery.
Another study from November 2021 described an effort to improve machine learning, thereby enhancing joint damage assessments in patients with rheumatoid arthritis. Presented at the American College of Rheumatology conference last year, this effort describes the benefits that machine learning can potentially provide to scoring joint damage, leading to more advanced diagnoses.
