It may become a usual event that we see doctors regularly using computer predictions in judging how to treat their cancer patients. This has come after scientists constructed mathematical formulas that have been able to outperform human experts in forecasting how sufferers will respond to treatment. This suggests that the future of medical diagnostics could be based around computer programs.
It was recently shown that a computer model of lung cancer made consistently better predictions of the future symptoms that would be suffered by a set of patients undergoing radiotherapy or chemotherapy than the doctors who were actually treating them. The authors argue that the study now demonstrates the increasingly important role of mathematics in cancer medicine.
This concept has been around for a while, but so far it has been difficult to find examples of computers outperforming their human counterparts. However, this study worked as personal medical details and the history of treatment for each patient were fed into the computer model, which then gave a better assessment than experienced radiation oncologists of how individuals were likely to respond to the treatment and how it could be altered to give a more positive outcome.
The authors stated that if models based on a patient, tumour and treatment characteristics already out-perform the doctors, then it may become unethical to make treatment decisions based solely on the doctors’ opinion when there is a more reliable source available. We believe models should be implemented in clinical practice to guide decisions.
However, this statement in itself brings up a number of ethical questions about the medical profession and how far we should rely on machines as a long-term diagnostic tool.