Two NHS trusts in the UK will begin using the AI blood test PinPoint for women suspected of having uterine cancer. In a study involving 3,313 participants, the machine learning model demonstrated a 99% accuracy in detecting and ruling out gynecological cancers, according to The Guardian.
Each year, around 90,000 women in England are referred for examination due to heavy bleeding after menopause. Approximately 10,000 of these receive a diagnosis of uterine (or endometrial) cancer, which results in 2,700 deaths annually.
The standard procedure involves a transvaginal ultrasound to measure the thickness of the endometrium. If concerns persist, a biopsy and hysteroscopy are performed. Many patients find these procedures painful.
However, 20% of those referred do not have cancer. Developers estimate that the test will allow for the exclusion of the disease at the initial consultation stage, sparing about 18,000 women each year from unnecessary invasive procedures.
How the Test Works
The PinPoint algorithm analyzes 30 markers in a blood sample. Based on these markers, the model assesses the risk level as low, elevated, or high.
This technology was developed by PinPoint Data Science, a Leeds-based company specializing in statistical analysis of medical data.
The test is being implemented following trials involving 16,481 patients. General practitioners from 170 clinics in Yorkshire referred patients for screening for nine types of cancer. All participants, including 3,313 women with bleeding and suspected uterine cancer, provided samples.
"A 99% accuracy is outstanding by any clinical standard. But equally important is that the test reliably rules out cancer in low-risk women," said Sean Duffy, Chief Medical Officer at PinPoint Data Science.
Mid Yorkshire NHS Teaching Trust plans to use the test for six types of gynecological cancers and cancers of the upper gastrointestinal tract. Leeds Teaching Hospitals NHS Trust will implement it for gynecological cancers.
Reactions from Doctors
General practitioner Jacinta Walsh from West Yorkshire noted that currently, ruling out cancer requires up to six visits to a doctor. She stated that PinPoint will expedite the process and free up time for other patients.
Consultant gynecologist Tracy Jackson from Leeds Trust remarked that most women referred turn out to be healthy. The new technology will allow for more efficient triaging.
"Women who actually have cancer will get to see a doctor faster, receive a diagnosis, and start treatment. That should be our focus," she emphasized.
Cancer Research UK described the results of the PinPoint test as "promising." However, representative Samantha Harrison stressed the need for further research to assess the benefits for patients and the healthcare system.
In April, a group of researchers reported that they had trained AI to detect pancreatic cancer long before it becomes visible on scans.