AI can help determine surgical candidacy of LSS patients to comparable level as panel of spine experts

6550
Raphael Mourad

Artificial intelligence (AI) can be used to bring efficiency and automation to the decision-making process for determining surgical candidacy for those with lumbar spinal stenosis (LSS), with performance comparable to a multidisciplinary panel of physicians. This is the key finding of new research published in the European Spine Journal by Raphael Mourad (Remedy Logic, New York, USA) et al.

The study also found that imaging combined with certain clinical variables such as motor deficit and pain are the key predictors of surgery candidacy.

According to the research team, “considering physicians and other health care providers must obtain advance approval from a health insurance company before a specific service, our model could fill in as a significant instrument for fast and efficient decisions at limited cost”.

The researchers proposed a novel hybrid AI model which computes the probability of spinal surgical recommendations for LSS, based on patient demographic factors, clinical symptom manifestations, and magnetic resonance imaging (MRI) findings.

The hybrid model combines a random forest model trained from medical vignette data reviewed by surgeons, with an expert Bayesian network model built from peer-reviewed literature and the expert opinions of a multidisciplinary team in spinal surgery, rehabilitation medicine, interventional and diagnostic radiology. Sets of 400 and 100 medical vignettes reviewed by surgeons were used for training and testing.

An independent panel of five spinal surgeons (fellowship trained spinal surgeons with more than five years of experience in practice) was set up. The panel reviewed the 500 medical vignettes to determine the surgical recommendation probability for each vignette.

They found that the model demonstrated high predictive accuracy, with a root mean square error (RMSE) between model predictions and ground truth of 0.0964, while the average RMSE between individual doctor’s recommendations and ground truth was 0.1940.

For dichotomous classification, the area under the receiver operating characteristic (AUROC) and Cohen’s kappa were 0.9266 and 0.6298, while the corresponding average metrics based on individual doctor’s recommendations were 0.8412 and 0.5659, respectively.

Speaking to Spinal News International, Andrej Rusakov, CEO of Remedy Logic, said: “We believe that democratising access to the world’s best clinical expertise will make receiving high quality medical advice more equitable globally. We have packaged the knowledge of the leading US orthopaedic and neurosurgeons into an AI that we plan to make available to anyone with an internet connection at a fraction of the cost. We are working to make independent, unbiased, and highly accurate medical advice available to everyone around the world.”


LEAVE A REPLY

Please enter your comment!
Please enter your name here