Researchers Mauricio E. Perez-Pachon of the Mayo Clinic in Rochester, Minnesota, and Dr. Jose T. Santayla of the Sima Clinic in Loja, Ecuador, stated that "the use of AI-powered predictive models enables surgeons to tailor their interventions to the individual needs of each patient, reducing the risk of excessive blood loss and ensuring optimal outcomes."
Liposuction is one of the most common cosmetic procedures worldwide, performed on more than 2.3 million patients annually. While generally considered a safe procedure, excessive blood loss is a potentially serious complication, especially when large amounts of fat are removed. Artificial intelligence-based tools have been developed to minimize blood loss in various medical fields, such as spinal surgery, orthopedics, and trauma.
The researchers and their colleagues analyzed data from 721 patients who underwent large-volume liposuction, in which the total amount of fat and fluid removed exceeded 4,000 milliliters (four liters). The procedures were performed at two clinics, one in Colombia and the other in Ecuador, following identical protocols.
The team used data from 621 patients to develop a model for predicting blood loss, taking into account a wide range of demographic, clinical, and surgical data. They then tested the model's performance on data from the remaining 100 patients, achieving 94% accuracy and excellent agreement between predicted and actual blood volume, with a standard deviation of 26 milliliters. The maximum difference between predicted and actual volume was 188 milliliters, while the minimum was only 0.22 milliliters.
The researchers emphasized that this model can help surgeons make informed decisions about perioperative management, including the need for blood transfusions, fluid management, and other critical care procedures. They added that the model's proactive approach can reduce complications, accelerate recovery, and improve patient education and informed consent processes.
The team plans to conduct further studies to expand the model's database with data from global surgeons, asserting that "future research in artificial intelligence holds enormous potential to enhance patient safety."
The study was published in the Journal of Plastic and Reconstructive Surgery.
