Table 2.

Different examples of artificial intelligence (AI) use across several orthopaedic sub-specialties

Scope	Examples
1. Fractures detection and prediction [41]	- Evaluate the accuracy of deep neural networks to diagnose neck femur fractures in comparison to perceptual training of medically naïve individuals [1]. - Predict hip fractures and estimate predictor importance in Dual-energy X-ray absorptiometry (DXA)-scanned individuals [30] - Evaluate the ability of convolutional neural network to detect distal radius fracture on an antero-posterior view of the wrist [19]. - Incorporate diverse measurements of bone density and geometry from central QCT imaging and of bone microstructure from high-resolution peripheral QCT imaging, can improve fracture prediction [6].
2. Osteoarthritis and arthroplasty	- Compare different gait patterns in patients with uni-compartment knee arthroplasty versus total knee arthroplasty [22]. - Early prediction of symptomatic knee osteoarthritis using MRI images [5, 43]. - Develop machine-learning based implant recognition system for hip arthroplasty designs [24]. - Measures of knee cartilage thickness can predict future loss of knee cartilage [23]. - Investigate the quantification of osteoarthritis and prediction of tibial cartilage loss by analysis of the tibia trabecular bone from magnetic resonance images of knees [32]. - Knee cartilage segmentation using a tri-planar convolutional network [44]. - ML tool demonstrates clinical utility with early prediction of patients who are most at risk of developing poor postoperative functional outcomes and PROMs after primary total knee arthroplasty [10]. - Predict length of stay, discharge disposition, and inpatient charges for primary anatomic, reverse, and hemishoulder arthroplasty [26].
3. Spine surgery	- Classification of scoliosis curves [2]. - Detection of lumbar spine compression fractures [3]. - Using a handgrip device and target tracking test to detect impairments of hand motor function in patients with cervical spondylotic myelopathy [31]. - Detection of spinal metastasis using a multi-resolution approach [56].
4. Foot and Ankle surgery	- Using automated segmentation to study distance and coverage mapping in Chopart joints in patients with progressive collapsing foot deformity (PCFD) [7, 8]. - Advanced semi-automated segmentation to evaluate hallux rigidus [15]. - Objective Computational technique to classify ankle osteoarthritis on weight bearing computation tomography (WBCT) [51]. - Semi-automated assessment of different hallux valgus parameters on (WBCT) of the hallux valgus [16]
5. Miscellaneous	- Switching neural networks used to classify multiple osteochondromas [37]. - Develop a machine learning algorithm to predict the prolonged opioid use after total hip arthroplasty (THA) [25]. - Online image messaging platform for remote monitoring of surgical incision sites [58]. - Ensemble learning techniques to study skeletal maturity [12]. - Classify pathological gait patterns using 3D ground reaction force (GRFs) data [4].