By
Doctors are using artificial intelligence to get the jump on a wide range of diagnoses. In addition to accelerating the detection of common conditions such as cancer and cardiac disease, AI can flag rare disorders that clinicians typically aren’t looking for. It’s beginning to prevent amputations by identifying patients with an elevated risk of limb loss long before their symptoms become acute.
And before long, it may become a key tool in determining lower-limb amputees’ K-level.
A recent paper in the Journal of Clinical Medicine described an AI model that accurately predicts walking ability by analyzing nine data points from a patient’s chart. In addition to physical indicators such as age, body mass index, and amputation level, the model examines psychosocial factors such as education, family support, and depression. The model “could uncover more efficient predictors that might be missed by using traditional methods,” the authors assert, “thereby improving clinical decision-making and patient outcomes.”
The study encompassed 104 lower-limb amputees with an average age of 62 years. The vast majority were males (77 percent) who lost limbs due to vascular causes (88 percent); 62 percent had above-knee amputations. Participants received standard K-level evaluations using traditional methods; they also performed two functional mobility tests (Timed Up and Go, and Two-Minute Walking).
Independently, the AI model examined 20 data points to predict each patient’s K-level and functional mobility results. In 93 percent of the cases, the model found the same K-level as the human clinician. Perhaps more impressive, it forecast performance on the Up and Go and Two-Minute tests with 83 percent accuracy.
“The primary motivation for applying prediction methods in medicine is to reduce the risk of incorrect diagnoses or treatments,” the authors conclude. “Accurate prediction of walking ability after lower limb amputation plays a critical role in patient rehabilitation and forms the basis for appropriate prosthesis prescription.” By using AI, “clinicians can better tailor rehabilitation interventions to address specific deficits and optimize functional outcomes for individuals undergoing pros-thetic rehabilitation following lower limb amputation.”
The full paper is available online at mdpi.com/journal/jcm.
