Researchers studied how wearable devices, for example, smartwatches and fitness trackers, combined with machine learning can be used in medical research and patient care, focusing specifically on people with spinal cord injuries. The authors reviewed current research and created a guideline to help understand and when how these technologies should be used. They found that current wearable devices have major problems measuring things like sleep patterns and autonomic nervous system function in people with spinal cord injuries, and that standard machine learning methods don't work well because they can't handle missing data, signal errors, and individual differences between patients. The authors created a guide that emphasizes matching what devices can actually measure with what doctors need to know, while also addressing data privacy and making sure the results are clinically useful. They conclude that making these technologies work properly requires doctors, engineers, and data scientists to work together and be honest about what the technology can and cannot do.