A new system to diagnose breast cancer 30 times faster than doctors
Researchers have developed automatic learning software that diagnoses patient’s breast cancer accurately and 30 times faster than doctors, based on the mammogram results and the patient’s medical history.
According to the website Science Alert, The new system could help doctors give a better diagnosis from the first time, which means fewer mammogram scans and fewer false positives.
Steven Wong, the researcher from Huston Methodist Research Institute, said: ‘’ this program intelligently reviews millions of records in a short amount of time, which efficiently enables us to determine breast cancer risks using patient’s mammogram’’, and he added: ‘’this has the ability to decrease the number of tissue samples or unnecessary biopsies’’.
A mammogram is a breast X-Ray that records any potentially cancerous cells before the appearance of the symptoms.
In many countries, Women over the age of 50 are advised to make a preventive screening every two years, but despite the goodness of the system, 50% of the tests in the united states results a false positives, which means that one in two healthy women are wrongly told that they might have Cancer.
Furthermore, a big grey area is found in the mammogram Rays, which falls between a 3 and 95% cancer risk. These patients are recommended to conduct follow-up biopsies.
According to researchers, 20% of biopsies taken in the United States nowadays are unnecessary, and the new artificial intelligence program was designed to significantly reduce that number by making more accurate diagnosis than the first time.
What distinguishes the new system?
Computer software is already used by doctors to analyse the mammogram results, but, this new system takes things a step further by considering the doctor’s reports on these scans, and by the patient’s full medical history , to accurately determine breast cancer risk.
In the most recent attempt to prove the new system’s validity, the researchers tested the artificial intelligent program on a 500 breast cancer patient’s mammogram results and pathology reports.
Within a few hours, the software showed the diagnostic information and identified the breast cancer subtype for each woman.
The doctors then compared the results between the artificial intelligent diagnosis and the clinical one, which showed that the software was 99% accurate, while the same analysis would have taken more than 500 hours with the doctors.
To be clear, the artificial intelligent system hasn’t been tested yet in the real world setting – the analyses in this study resulted from existing breast cancer patients – but the researchers now provided enough evidences enough to try real examinations.
Plus, it won’t be able to prevent all the false positives or the suspicious mammogram results, sometimes there’s just not enough available information to make a diagnosis. But it should help doctors to make a more accurate conclusion.