Scanning for signs of cancer is always a crucial first step in diagnosing the disease, with earlier diagnosis often proving crucial to the eventual outcomes and impact of treatment.
Artificial intelligence (AI) has been postulated in many quarters as holding the keys to earlier diagnosis in many cases, as its learning capacity can help it make predictions based on limited data concerning early signs of disease.
Research published in the journal Radiology has indicated that this appears to be the case with scans of lung nodules, which uses machine learning to detect whether small spots are the beginnings of cancerous tumours.
While scanning for tumours has gone on since the 1960s, the use of AI learning enables the past data to be used as a guide to the future, with the system being able to use information on spots from old scans that later turned out to be tumours to identify them at earlier stages in present-day scans.
This has helped improve the success rate of clinicians studying the scans in spotting cancerous growths. The study stated: “In this multireader multicase study, the performance of radiologists and pulmonologists in estimation of IPN malignancy risk significantly improved with the assistance of an artificial intelligence-based CAD tool.”
The study was undertaken with the involvement of the University of Pennsylvania, the University of Oxford, the Medical University of South Carolina, Wake Forest School of Medicine, Henry Ford Health System, Vanderbilt Ingram Cancer Center, and Oxford University Hospitals.
AI technology may not just help in scans, as it can also directly help to develop new medications. Biotech start-up Insilico Medicine has used the technique to develop a new candidate drug for treating Covid-19. This has a molecular structure that inhibits the 3CL protease enzyme involved in the replication of the virus.
Nonetheless, it may be in the area of scanning and early diagnosis that AI can most help those working with existing CT technology to achieve better results than ever before.
If you need medical imaging systems in the UK, contact us today.