As more and more data is being collected by the country’s healthcare system, researchers are exploring the option of using a subfield of artificial intelligence (AI), namely machine learning to help improve the quality of medical care and patient outcomes. An overview of some of the ways in which machine learning has the potential to contribute to advancements in plastic surgery were published in Plastic and Reconstructive Surgery recently.
Potentially a Powerful Tool
Dr. Jonathan Kanevsky from the McGill University in Montreal and his colleagues wrote, “Machine learning has the potential to become a powerful tool in plastic surgery, allowing surgeons to harness complex clinical data to help guide key clinical decision-making.” A few key areas were highlighted, all of which showed how machine learning and “Big Data” would be able to contribute to progress in the fields of reconstructive and plastic surgery.
Machine learning analyzes historical data to develop algorithms that are capable of knowledge acquisition. Dr. Kanevsky and his co-authors wrote, “Machine learning has already been applied, with great success, to process large amounts of complex data in medicine and surgery.”
The doctor has his colleagues believe that plastic surgery will be able to benefit from similar approaches, especially with the availability of the American Society of Plastic Surgeons ‘Tracking Operations and Outcomes for Plastic Surgeons’ (TOPS) database. Five specific areas that would be able to benefit from machine learning include:
A post-operative microsurgery application has been created to monitor blood perfusion of tissue flaps, and it works by analyzing information on smartphone photos. In the future though, algorithms may be developed to help provide suggestions regarding what the best approach would be regarding reconstructive surgery for a patient.
Peripheral Nerve and Hand Surgery
Machine learning could be useful in helping to predict how successful tissue-engineered nerve grafts will be, as well as for developing automated controllers that could be used by patients suffering from neuro-prostheses.
Machine learning has already been used to help predict healing times of burns, while also providing an effective tool for assessing the depth of them. In time, algorithms could also be developed to allow for faster prediction of percentage of a patient’s body area that has been burned, which is critical during patient resuscitation and surgical planning.
This technology could also be used in cosmetic surgery, such as when predicting and stimulating potential outcomes of cosmetic facial surgery and reconstructive breast surgery.
Machine learning has already been implemented for automated diagnosis of infant skull growth defects. In future though, the technology could be enhanced to identify the genes responsible for cleft lip and palate.
While this technology is extremely useful, Kanevsky stresses the point that measures would need to be implemented to ensure the safety and clinical relevance of results obtained by these means. He also mentioned that computer-generated algorithms are not yet able to replace the trained human eye, which is why human intervention would still be required to ensure the best possible outcomes.