According to a study, artificial intelligence appears to be almost twice as effective as the current method in grading the aggressiveness of a rare type of cancer from scans. Scientists have made this finding.
Unlike the naked eye, AI was able to detect the details that a lab would only be 44% correct in identifying.
This could help improve the treatment, leading to better services that would benefit thousands each year, say researchers from the Royal Marsden Hospital and the Institute of Cancer Research.
It also excites them by its ability to identify breast cancer at an early stage.
AI has proven itself to be very promising in terms of breast cancer diagnosis and reduction of treatment periods.
The computers can be fed with a lot of information on which they will be able to identify patterns and be capable of making predictive actions, solving problems or even learning from their mistakes.
The recognition of details invisible to the unaided eye, gave a 82 percent accuracy rate, which is better than that of lab analysis (at 44 percent).
According to the researchers from the Royal Marsden Hospital and Institute of Cancer Research, it could enhance the therapy provided to thousands of men each year.
Additionally, it makes other types of cancer detection at an early stage possible.
Indeed with the development of artificial intelligence there are promising signs in diagnosing breast cancer and shortening treatment durations.
The computers can receive huge amount with patterns that are used for forecasting, problem-solving, as well making mistakes.
“We are very enthusiastic about the possibilities that this cutting-edge technology offers,” commented Professor Christina Messiou, who serves as a consultant radiologist at The Royal Marsden NHS Foundation Trust and holds the position of professor in imaging for personalized oncology at The Institute of Cancer Research in London.
“We believe it has the potential to improve patient outcomes significantly, by enabling quicker diagnoses and more precisely tailored treatment options,” she added.
Writing in Lancet Oncology, the reseachers employed radioonics technology that revealed previously undetectable manifestations of the pathology such as soft-tissue thickening and bone destruction for staging of retroperitoneal sarcoma.
Such data helped the algorithms to rate the aggressiveness of 89 other European and US hospital patients’ tumors, from scans, more reliably than biopsies which involved a portion of tissues analyzed under the microscopes.
Writing in Lancet Oncology, the researchers used radiomics for identification of subclinical features of retroperitoneal sarcoma on MRI and CT scans of 165 cases.
By using these details, the artificial intelligence machine could assess the aggressiveness of another eighty-nine tumours from scan photos in European and American hospitals more precisely as compared with biopsies where only a little sample of malignant tissue is examined by microscope.
Tina McLaughlan, a dental nurse from Bedfordshire, received her diagnosis of a sarcoma in her abdomen last June due to stomach pain. Doctors used computerized-tomography (CT) scan images to identify the issue and decided against a needle biopsy due to potential risks.
Following the removal of the tumor, the 65-year-old now undergoes scans at the Royal Marsden every three months. While she wasn’t part of the AI trial, she expressed her belief to BBC News that it could greatly benefit other patients.
“When you have your first scan, and they can’t give you a definite diagnosis – they didn’t inform me throughout my treatment until the post-op histology report, so having immediate results would be incredibly helpful,” said Ms. McLaughlan. “Hopefully, it would lead to faster diagnoses.”
Each year, around 4300 people are diagnosed with this kind of cancer in England.
The founder of this company believes that one can day use it across globe on high-risk patients whose individual treatment should be personalized as compared to those who have low risk that should not undergo other unnecessary treatment and scanning.
Dr Paul Huang, from the Institute of Cancer Research, London, said: This type of technology can transform lives of patients with sarcoma, making it possible to develop personalised treatment protocol directed to the molecular profile of sarcoma.
Such promising results – it’s great.
Approximately four thousand three hundred people get diagnosed with this type of cancer every year in England.
Prof Messiou thinks that technology will one day be applicable globally, and high risks patients will receive different treatment, while avoiding excess treatments and follow-ups scans for low risks patients.
Dr Paul Huang, from the Institute of Cancer Research, London, said: Such technology can change the lives of sarcoma patients and tailor the treatment according to the specific cancer biology.
“These results are certainly very promising.”
Leave a Reply