The increasing clinical demands on radiology departments worldwide has challenged current service delivery models, particularly in publicly-funded healthcare systems. It is no longer feasible for many Radiology departments with their current staffing level to report all acquired images promptly, leading to large backlogs of unreported studies. It is estimated that, at any time, 330 000 patients are waiting more than 30 days for their medical reports in the United Kingdom. For this reason, alternative models should be explored.
Artificial intelligence has been proposed as an automated means to reduce this backlog and identify exams that merit immediate attention. Integrating AI with innovative platforms would improve department workflow and workforce efficiency, while also creating more accurate reports, offering enormous potential benefits to patients, clinicians and hospitals.
In our lab, we have develop anomaly detection and Natural Language Processing (NLP) methods to aid triage. By using these technologies, it is possible to identify anomalies in the exams and reports, predicts the priority of it, and then automated generate sentences describing the radiologic abnormalities seen.
Wood, D.A., Lynch, J., Kafiabadi, S., Guilhem, E., Busaidi, A.A., Montvila, A., Varsavsky, T., Siddiqui, J., Gadapa, N., Townend, M. and Kiik, M., 2020. arXiv preprint arXiv:2002.06588.
Decision fusion of 3D convolutional neural networks to triage patients with suspected prostate cancer using volumetric biparametric MRI
Mehta, P., Antonelli, M., Ahmed, H., Emberton, M., Punwani, S. and Ourselin, S., 2020, In Medical Imaging 2020: Computer-Aided Diagnosis (Vol. 11314, p. 1131433). International Society for Optics and Photonics.