For those diagnosed with brain cancer the statistics are grim, with less than a third surviving five years with the condition.
Part of this poor outlook is due to the nonspecific symptoms of this type of cancer, such as headaches and memory loss. This is further compounded by the lack of a quick, cost-effective test with which to triage those who display symptoms in primary care.
It’s this lack which a team of researchers from the University of Strathclyde, Glasgow are attempting to address using a new blood test which is analysed by a machine learning algorithm. Promising results of their clinical feasibility study, now published in Nature Communications, have brought hope that the test may soon be usable by health professionals.
The test uses attenuated total reflection-Fourier transform infrared (ATR-FTIR) spectroscopy to analyse the biochemical profile of a blood sample. A specially trained machine learning algorithm then compares the sample with the biochemical “fingerprint” of brain cancer.
The study findings revealed a sensitivity of 83.3% and a specificity of 87% for the new test in a cohort of 104 people. According to study co-author Dr Paul Brennan, these results indicate that “With this new test, we have shown that we can help doctors quickly identify which [people] with these nonspecific symptoms should be prioritized for urgent brain imaging.”
>> Read the original article here
Source: Medical News Today