Abstract
Breast cancer is the most common cancer in the UK, accounting for a third of cancers diagnosed annually. To better manage patients it is essential to diagnose early, with high accuracy, the type and grade of cancer to ensure correct and rapid treatment.Breast magnetic resonance imaging (MRI) provides excellent soft tissue contrast and is unaffected by fibroglandular tissue density. By dynamically imaging during the injection of a contrast agent, a sensitivity approaching 100% is realised and lesion vascularity is portrayed. This makes the examination not only useful in detection and staging of breast cancer, but also for monitoring response to neoadjuvant chemotherapy (NAC) treatment.
This work used an image processing technique known as texture analysis (TA), which analyses pixel intensity distributions on a pixel-by-pixel scale to identify patterns that may not be visually interpretable, to probe MRI images of women with proven breast cancer. Whilst previous works have demonstrated differentiation between normal, benign and malignant tissue, this work sought to extend this and look at classification of breast cancer subtypes, the utility in a clinical environment and to assess whether the technique could identify early response in patients undergoing NAC.
TA cannot only be applied using standard MRI set-ups, but the studies demonstrated preliminary promise in the classification of different cancer subtypes- both in terms of histological subtype and grading, as well as state-of-the-art molecular subtyping. While larger patient data sets are required to demonstrate this definitively, initial results show encouraging findings.
It has also been shown that TA can be used in patients undergoing NAC to indicate whether the patient will respond well or not, and of particular interest is that these results appear to correlate well with the final pathological outcome.
The research within this thesis has clearly demonstrated that TA is a useful research tool within the area of breast MRI and further investigation in this area is essential.
| Date of Award | 2014 |
|---|---|
| Original language | English |
| Supervisor | Richard Lerski (Supervisor) & Alastair Thompson (Supervisor) |