Abstract 14P
Background
Classification of molecular subtypes of breast cancer is widely used in clinical decision-making, leading to different treatment responses and clinical outcomes. We classified molecular subtypes using a novel deep learning algorithm in whole-slide histopathological images (WSIs) with invasive ductal carcinoma of the breast.
Methods
We obtained 1,094 breast cancer cases with available hematoxylin and eosin-stained WSIs from the TCGA database. We applied a new deep learning algorithm for artificial neural networks (ANNs) that is completely different from the back-propagation method developed in previous studies.
Results
Our model based on the ANN algorithm had an accuracy of 67.8% for all datasets (training and testing), and the area under the receiver operating characteristic curve was 0.819 when classifying molecular subtypes of breast cancer. In approximately 30% of cases, the molecular subtype did not reflect the unique histological subtype, which lowered the accuracy. The set revealed relatively high sensitivity (70.5%) and specificity (84.4%).
Conclusions
Our approach involving this ANN model has favorable diagnostic performance for molecular classification of breast cancer based on WSIs and could provide reliable results for planning treatment strategies.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.
Resources from the same session
228P - Real-world data on dose adjustment of cabozantinib in advanced renal cell carcinoma
Presenter: Hemavathi Baskarane
Session: Poster Display
Resources:
Abstract
229P - The application of diffusion kurtosis imaging in predicting muscle invasion of bladder cancer: A comparison with conventional DWI
Presenter: Shuai Jiang
Session: Poster Display
Resources:
Abstract
230P - Oncological outcomes between partial cystectomy and radical cystectomy in solitary muscle invasive bladder cancer with downgraded T stage
Presenter: Ming Wei Hsu
Session: Poster Display
Resources:
Abstract
231P - BMI-predicted progression-free survival after pembrolizumab therapy for urothelial cancer: Asian version of BMI classification is suitable for Asian patients
Presenter: mirii harada
Session: Poster Display
Resources:
Abstract
232P - The immunosuppressive features of the 20S Proteasome β-subunit gene family in von Hippel-Lindau (VHL)-mutated clear cell renal cell carcinoma (ccRCC): A TCGA-based bioinformatics study
Presenter: Saja Alzghoul
Session: Poster Display
Resources:
Abstract
233P - The crosstalk between PBRM1 loss and tumor immune microenvironment (TIME) of clear cell renal cell carcinoma (ccRCC): A possible interconnection to immunotherapy response
Presenter: Ahmed Al Sharie
Session: Poster Display
Resources:
Abstract
235P - Do FGFR2 and 3 proteins have a role in the prognosis of urothelial bladder carcinoma?
Presenter: Alshimaa Al Hanafy
Session: Poster Display
Resources:
Abstract
236P - The effects of chemotherapy on body composition in patients with advanced urothelial carcinoma
Presenter: KOSUKE KITAMURA
Session: Poster Display
Resources:
Abstract
237P - Real-world analysis of adjuvant nivolumab in resected urothelial cancer: A single institute study in Taiwanese patients
Presenter: Mu-Hsin Chang
Session: Poster Display
Resources:
Abstract
238P - Enfortumab-vedotin for metastatic urothelial carcinoma refractory to platinum-based chemotherapy and immune checkpoint inhibitors: A single institution experience
Presenter: Yuki Endo
Session: Poster Display
Resources:
Abstract