University of Twente Student Theses
Harnessing AI to Predict a Pathologic Response in Breast Cancer Patients Undergoing Neoadjuvant Chemotherapy
Snoeijink, J.A.N. (2025) Harnessing AI to Predict a Pathologic Response in Breast Cancer Patients Undergoing Neoadjuvant Chemotherapy.
|
PDF
7MB |
| Abstract: | Breast cancer can be treated with neoadjuvant chemotherapy, which involves the use of chemotherapy before surgery. The main goal of this type of chemotherapy is to shrink the tumor and enhance the likelihood of achieving a pathologic complete response (pCR) following surgery. Despite the improvement of neoadjuvant chemotherapy in recent years, which has led to increased pCR rates in especially the Her2+ and triple-negative subgroups, reliable non-invasive biomarkers or imaging methods for the prediction of pCR are currently lacking. This study included 291 patients from Deventer Hospital and utilized both clinical data and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI). Radiological features were extracted from the DCE-MRI data for analysis. The results demonstrated that the MAMA-MIA network achieved the highest accuracy in segmenting breast tumors on DCE-MRI. Nevertheless, manual correction of the MAMA-MIA network's segmentations was necessary to ensure suitability for both clinical and research applications. The optimal balance between sensitivity and specificity for predicting pCR and Residual Cancer Burden (RCB)-0 classification was achieved using data from the first two MRIs during neoadjuvant chemotherapy, combined with clinical and radiological features. For pCR and RCB-0 predictions, the model achieved sensitivities of 0.75 and 0.81, and specificities of 0.83 and 0.80, respectively. Furthermore, the multi-class classification model for RCB demonstrated an accuracy of 0.69 and a Cohen’s kappa of 0.51. |
| Item Type: | Essay (Master) |
| Clients: | Deventer Hospital, Deventer, Netherlands |
| Faculty: | TNW: Science and Technology |
| Subject: | 44 medicine, 50 technical science in general |
| Programme: | Technical Medicine MSc (60033) |
| Link to this item: | https://purl.utwente.nl/essays/104926 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page