This evaluation is a complex, multidisciplinary process requiring the analysis of various data sources, including radiological images, tumor tissue characteristics, and clinical data. This collaboration between specialists takes time, and since neoadjuvant therapy is not always effective, some patients endure significant side effects without benefiting from the treatment.
In a new study published in Nature Communications, researchers from the NKI, led by Ritse Mann, developed an AI model called Multi-Modal Response Prediction (MRP). This model can help physicians predict how breast cancer patients will respond to neoadjuvant therapy.
The research team tested MRP using data from 2,436 breast cancer patients treated at NKI between 2004 and 2020. Unlike traditional AI models that rely on a single data type, MRP combines multiple sources, such as radiological images, tumor tissue characteristics, and clinical data. This approach improves accuracy and provides insight into the reasoning behind predictions, making cancer care more effective and personalized.
Researcher Yuan Gao explains: “Our MRP system shows potential in assessing neoadjuvant therapy response in breast cancer, particularly before treatment begins. It helps identify patients most likely to achieve a complete response, offering valuable guidance for personalized treatment strategies.”
What makes MRP unique is its ability to predict outcomes and explain why it reaches those conclusions. This transparency increases trust among physicians and enables them to use the model at multiple stages of treatment:
- Before therapy: MRP can identify patients unlikely to respond well to neoadjuvant therapy, sparing unnecessary side effects, and explore alternative treatments.
- During treatment: The model tracks changes in patient data, allowing physicians to adjust treatment plans as needed.
- After treatment: MRP can help determine which patients no longer need surgery because no cancer cells remain, potentially avoiding unnecessary procedures.
By integrating imaging and pathological data over time and handling variations in data availability across multiple centers, MRP can reduce physicians’ workload, accelerate precise treatment decisions, and ultimately improve patient outcomes.
A step towards personalized care
The MRP system represents an important step toward more personalized breast cancer treatment, personalizing therapy to each patient's specific needs. By leveraging artificial intelligence and combining different types of data, this model has the potential to enhance treatment effectiveness and improve patients’ quality of life. Researchers are now working to translate this innovation into a practical tool for clinical use.
Mann emphasizes: “Our goal is to integrate multi-modality and multi-timepoint data to replicate and automate the multidisciplinary team meetings, enabling faster and more accurate treatment response evaluations and paving the way for future clinical trials.”
Artificial intelligence in healthcare creates new possibilities for breast cancer treatment and highlights AI’s broader potential to improve medical care.