Imagine a future where cancer patients can sidestep unnecessary chemotherapy, thanks to artificial intelligence. But is this a blessing or a controversial development? AI is stepping in where human eyes might fail, and it's sparking a revolution in cancer care.
For years, doctors have relied on microscopes to examine cancer biopsies, but this method isn't foolproof. It might overlook intricate patterns that indicate a tumour's potential threat. This is where AI comes into play, and a Norwegian start-up, DoMore Diagnostics, is leading the charge.
Their mission? To revolutionize colorectal cancer diagnosis and treatment. By harnessing AI, they aim to provide more precise assessments, potentially saving patients from harmful and unnecessary chemotherapy.
But here's where it gets intriguing: their AI technology scrutinizes tissue samples with a level of detail that surpasses the human eye. Torbjørn Furuseth, CEO of DoMore Diagnostics, confidently asserts, "We personalize cancer treatment by harnessing AI's power." With the alarming statistics from the World Health Organization, showing colon cancer as the third most common and second most deadly cancer globally, this innovation couldn't be more timely.
The AI algorithm is trained on a vast dataset of images, making it exceptionally adept at identifying high-risk features associated with cancer recurrence and mortality. This is a significant improvement over traditional methods, as it can predict patient outcomes more accurately than human pathologists, according to Andreas Kleppe, a research director at Oslo University Hospital Research.
The AI tool's ability to analyze digital images of cancer tissue samples allows it to estimate the cancer's growth rate and risk level. This is crucial in determining which patients require aggressive treatments like chemotherapy and which can safely avoid them. After surgery, when the tumour is removed, this prognostic analysis is vital, as some patients may still have small metastases, secondary cancer growths that can spread.
Chemotherapy is often administered as a blanket treatment, but DoMore Diagnostics argues that this approach exposes most patients to side effects without any real benefit. They claim that their AI-based test is currently being used in hospitals across Europe, the United States, Japan, and Mexico to validate prognostic analyses, potentially sparing patients from unnecessary chemotherapy.
And this is the part most people miss: the AI's decision-making process is a bit of a black box. Kleppe admits, "We don't really know what the AI is looking for." Yet, when correlated with pathologist evaluations, the AI's decisions make sense. It considers features pathologists look for and combines them with other factors, providing a more comprehensive analysis.
As this technology advances, it raises questions about the future of cancer treatment. Will AI eventually replace human pathologists? Or will it always be a tool to assist them? What are your thoughts on this AI-driven approach to cancer care? Is it a welcome advancement or a step into uncharted ethical territory?
