AI In Cancer Research and Prevention
The hype around artificial intelligence has reached a fever pitch, with its acolytes celebrating rapid advancements in computer power and use in novel applications. AI’s role in healthcare, particularly in cancer research, presents one of the most exciting possibilities. Most of AI’s success has been in accelerating the process or reducing the time needed to come through vast reams of data or run through complex scenarios.
That’s changing.
Researchers at Uppsala University used AI to identify tissue changes that signaled early-stage tumor growth in otherwise healthy adult men. The ability to detect tissue changes indicative of cancer months, potentially years earlier than through normal pathological assessment, could dramatically improve health outcomes. This is one step beyond number-crunching, and a perfect example of AI’s ability to both extrapolate and isolate patterns with extremely complex images.
The “Missed Study”
Researchers are specifically looking for ways to identify fluctuations or abnormalities in tissue before they would otherwise be spotted by pathologists or are missed during cancer screenings. The study found AI was capable of spotting signs of prostate cancer missed by pathologists in more than 80% of samples from men who would later be diagnosed with cancer. Examining samples from 232 men over several years, half of the men developed prostate cancer, with the rest of the sample pool cancer-free after eight years.
Read More: Updated Research Shows Firefighters Face Increased Cancer Risk
From Negative to Positive
All samples were initially negative after an assessment by a professional pathologist. The AI team trained AI models on biopsy images in incremental batches, with the intent being for the models to identify patterns within the minute abnormalities in eventual tumors. In the near-term, this technique may impact how soon men with negative tests should be tested again, with subtle signs indicating varying degrees of risk.
More Ways AI Is Used In Cancer Research
The “Missed Study” is one of many projects that use AI to improve efficiency by aggregating and contextualizing massive, often impossibly large amounts of data. These tools, often hand-built using models created by OpenAI, Google, and other niche AI healthcare companies like Tempus and PathAI, could lead to significant improvements in cancer prevention as well. The ability to process reams of complex data sets and identify patterns in biopsies, taking into account factors like age, race, socioeconomic backgrounds, regional risks, and other variables, could dramatically improve screening value. It could indicate who really needs screening, when to start, and how frequently.
Committed to Prevention
One of the biggest risks is that AI will be funded and therefore focused on drug and treatment development. There is more money to be made treating cancer than preventing it, and it will be critical for governments, regulatory bodies, drug companies, and the public to ensure prevention has its place at the table. Less Cancer knows cancer prevention saves lives, lowers the emotional and economic toll of diagnosis and treatment, and supports healthier lives. Support our work to keep prevention a priority with a donation today.