According to the World Health Organization, spinal fractures affect millions of people worldwide. With early detection and diagnosis, patients can be managed and treated with more optimal neurovascular outcomes, such as in spinal cord compression. Therefore, detecting, predicting, and diagnosing spinal fractures early is critical for patient outcome optimization and improving neurovascular prognosis.  

Imaging has been found to aid in the efficient and accurate detection of spinal fractures. In this research project, a machine learning model is developed using artificial intelligence, imaging, and clinical patient data to aid the detection and prediction of spinal fractures and increase its diagnostic accuracy to optimize patient neurovascular outcomes.

The machine learning model was built using deep network algorithms to detect and diagnose spinal fractures in patients. Clinical data, including over 3,230 patients from hospitals, was used to develop, train, and test the machine learning model. To train the model, 55% of the patient data was randomly selected while the remaining 45% of the data was utilized for testing fracture diagnosis capabilities of the model.

In diagnosing spinal fractures in patients, the machine learning model was able to achieve an overall accuracy of 89.8% using imaging in diagnosis and prognosis of patient neurovascular outcomes.

Thus, machine learning algorithms can be used to help medical professionals and underserved areas globally for enhancing early detection and diagnostic accuracy of spinal fractures, neurovascular outcomes, and patient care.