Convent Artificial Intelligence: A Glance at the Future of Diagnosis

Artificial Intelligence (AI) is intelligence demonstrated by machines, as opposed to natural intelligence displayed by human beings. It is intended to mimic and display "human" cognitive skills, such as "learning" and "problem-solving."

Deep learning, a subset of AI has transformed both the accuracy and speed of clinical image analysis, particularly in orthopaedics. Being trained by a relevant data set, deep learning algorithms can determine abnormalities on imaging specifically radiographs, with high precision in seconds.

An early demonstration of this potential in orthopedics is in wrist fractures in x-ray images. The artificial intelligence algorithm was able to identify fractures with accuracy similar to the interpretations of human radiologists. This resulted in a positive FDA approval.

AI in foot and ankle imaging

“Deep Learning Algorithms Improve the Detection of Subtle Lisfranc Malalignments on Weightbearing Radiographs,” a retrospective comparative study by Sohell Ashkani-Esfahani published in Foot & Ankle International, presented the results of 640 feet with/without Lisfranc malalignment on radiographs, assessed by two surgeons versus two deep learning models –trained, validated, and tested.

AI is here to stay, at least in imaging

No significant differences were observed between the patients and the controls regarding age, gender, race, and body mass index, according to the study. However, the best deep learning algorithm outperformed our human interpreters (<1% vs ~10% misdiagnosis), 94.8% sensitivity, 96.9% specificity, 98.6% accuracy, 95.8% F score, and 99.4% AUC.