Ensemble of deep learning models to predict platinum resistance in high grade serous ovarian cancer

Han, Kyullhee; Ham, Hyeonjung; Song, Yong Sang; Park, Taesung; Ahn, Tae-Jin

doi:10.1504/ijdmb.2020.112851

Cited by 1 publication

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Han et al used a deep learning model to predict platinum resistance in patients with HGSOC and achieved a classification accuracy of 74% [17]. Liu et al predicted platinum resistance in patients based on histopathological images using the Inception V3 deep learning model and achieved an AUC of 0.846 for classification [18].…”

Section: Discussionmentioning

confidence: 99%

An Attention-Based Deep Learning Network for Predicting Platinum Resistance in Ovarian Cancer

Zhuang,

Li,

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Ovarian cancer is one of the three most common types of gynecological cancer globally, with high-grade serous ovarian cancer being the most common and aggressive histological type. Guided treatment of high-grade serous ovarian cancer typically involves platinum-based combination chemotherapy, necessitating the assessment of whether the patient is platinum resistant. This study proposes a deep learningbased method to determine whether a patient is platinum resistant using multimodal positron emission tomography/computed tomography images. In total, 289 patients with high-grade serous ovarian cancer were included in this study. An end-to-end Squeeze-Excitation-Spatial Pyramid Pooling-Dense Convolutional Network model was built by adding a Squeeze-Excitation Block and Spatial Pyramid Pooling Layer to a Dense Convolutional Network. Multimodal data from positron emission tomography/computed tomography images of regions of interest were used to predict platinum resistance in patients. Through five-fold crossvalidation, the Squeeze-Excitation-Spatial Pyramid Pooling-Dense Convolutional Network achieved a high accuracy rate and area under the curve of 92.6% and 0.93, respectively, for predicting platinum resistance in patients. The importance of incorporating the Squeeze-Excitation Block and Spatial Pyramid Pooling Layer into the deep learning model and considering multimodal data was substantiated by performing ablation studies and experiments with single-modality data. The classification results indicate that our proposed deep learning framework performs better in predicting platinum resistance in patients, which can help gynecologists make more appropriate treatment decisions.

show abstract

Section: Discussionmentioning

confidence: 99%