How AI is Changing the Future of Ovarian Cancer Research
Every two hours, a woman in the UK dies from ovarian cancer. That’s a shocking number, and it hasn’t changed much in the last decade. But now, new technology is helping us make real progress. Artificial intelligence, or AI, is one of the most exciting tools in the fight against ovarian cancer.
AI is already changing the face of cancer research, driving innovation in breast, lung, prostate, brain, skin, and thyroid cancers. It’s shown incredible promise in early detection and treatment, two things ovarian cancer researchers are very interested in.
In this blog, we’ll explain AI – what it is, how it works in ovarian cancer research, and how it could help save lives.
What is AI?
Artificial Intelligence (AI) is when computers are trained to do tasks that would usually need human thinking – things like learning from patterns, making decisions, or spotting something unusual in a scan.
AI works through something called machine learning. This means we ‘teach’ the computer by feeding it lots of data like images, numbers, or text. Over time, the computer learns to make predictions or suggestions based on that data.
Can AI tell us more about how effective treatment can be?
How AI can help diagnose ovarian cancer
Ovarian cancer isn’t just one disease. There are multiple types and subtypes, and they all respond differently to treatment. Because of this, it’s essential to know what type someone has so they get the proper care.
That’s where digital pathology and AI come in.
What is digital pathology?
It’s a way of using computers to look at scans of tissue samples. These images can show what kind of ovarian cancer someone has - but it's usually hard work, and can take a long time for doctors to analyse.
What does AI do?
AI can look at these scans and learn to spot patterns faster than the human eye. It uses a special kind of machine learning called a deep convolutional neural network (basically, a very smart pattern-spotting system).
This means it could help doctors:
- Predict what type of cancer someone has before a full diagnosis.
- Work out who needs which tests or treatments next.
- Improve early detection, which is key to saving lives.
Making AI smarter with new tools
To make AI better, scientists are exploring smart ways to improve how it learns:
- Federated learning – Instead of sending all data to one place, we train AI locally in different places and share the results.
- Transfer learning – Start by training AI with general images, then fine-tune it with cancer-specific scans.
- Synthetic data – Creating ‘fake’ but realistic images based on real ones to give the AI more to learn from. (And don’t worry - even the best experts can’t tell the fake images from the real ones!)
Can AI help decide who needs surgery or chemo?
For a common type of ovarian cancer called High-Grade Serous Ovarian Carcinoma (HGSOC), doctors often start with chemotherapy. But what if surgery were actually better?
AI can help answer that.
By looking at CT scans and using a method called radiomics, AI can spot tiny patterns and numbers in the scan that predict whether a tumour will shrink with chemo. This helps doctors:
- Choose the best treatment plan
- Avoid giving chemo when surgery would be more effective
- Personalise care to each patient’s unique cancer.
Bridging AI and precision medicine
Some patients have run out of treatment options - but AI might help them, too.
Right now, doctors often go through lots of steps to match a patient’s tumour to the best treatment. But what if AI could do that faster?
Scientists are using large language models (LLMS), such as ChatGPT, to read medical papers, match patterns in the data, and suggest treatment options. One tool is already being tested using made-up patient cases, and results are promising.
What about privacy?
Using real patient data can be tricky because of privacy laws. But there’s a clever solution: using data and medical images that don’t include personal information. These can help AI learn without risking privacy.
Why this matters
97% of data in hospitals goes unused. That’s a huge waste of knowledge. AI can help us dig into that hidden data and unlock answers we didn’t even know existed.
From earlier diagnoses to smarter treatments, AI in ovarian cancer research could save lives. And with global collaboration and support from big players like Microsoft AI for Good Lab, we’re finally seeing momentum.
A collaboration of change-makers: AI for Ovarian Cancer Research Grant
For the first time ever, four global ovarian cancer charities are joining forces to fund innovative, AI-driven research that will improve ovarian cancer survival rates, in partnership with Microsoft’s AI for Good Lab.
AI is already changing the face of cancer research, driving innovation in breast, lung, prostate, brain, skin, and thyroid cancers. It has shown amazing promise in ovarian cancer problem areas such as early detection and treatment.
By harnessing the power of technology, this grant will unlock new possibilities - changing how we understand, diagnose and treat ovarian cancer.
Find out more about the global partnership and how this collaboration hopes to change the future of ovarian cancer diagnosis and treatment.
