File Download
Supplementary

Conference Paper: Deep convolution neural network model for automatic risk assessment of patients with non-metastatic nasopharyngeal carcinoma

TitleDeep convolution neural network model for automatic risk assessment of patients with non-metastatic nasopharyngeal carcinoma
Authors
KeywordsNasopharyngeal Carcinoma
Convolution neural network
Progression-free survival
Segmentation
Classification
Issue Date2019
Citation
International Conference on Medical Imaging with Deep Learning (MIDL), London, UK, 8‑10 July 2019 How to Cite?
AbstractNasopharyngeal Carcinoma (NPC) is endemic cancer in the south-east Asia with approximately a third of the world incidences. With the advent of intensity-modulated radiotherapy excellent locoregional control are being achieved. Consequently, this had led to pretreatment clinical staging classification to be less prognostic of outcomes such as recurrence after treatment. Alternative pretreatment strategies for prognosis of NPC after treatment are needed to provide better risk stratification for NPC. In this study we proposed a deep convolution neural network model based on contrast-enhanced T1 (T1C) and T2 weighted (T2) MRI scan to predict 3-year disease progression of NPC patient after primary treatment. For development, we retrospective obtained 596 non-metastatic NPC patients from four independent centres in Hong Kong and China. Our proposed model first performs a segmentation of the primary NPC tumour to localise the tumour, and then uses the segmentation mask as prior knowledge along with the T1C and T2 scan to classify 3-year disease progression. For segmentation, we adapted and modified a VNet to encode both T1C and T2 scan and also encoding to classify T and overall stage classification. Our results shows that the modified network performed better than baseline VNet with T1C only and also network with no T and overall classification. The classification result for 3-year disease progression achieved an AUC of 0.828 in the validation set but did not generalised well for the test set which consist of 146 patients from a different centre to the training data (AUC = 0.69). Despite the low multicentre performance, our preliminary results show that deep learning may offer prognostication of disease progression of NPC patients after treatment. One advantage of our model is that it does not require manual segmentation of the region of interest, hence reducing clinician's burden compared to other methods such as radiomics. However further development in generalising multicentre data set are needed before clinical application of deep learning models in assessment of NPC.
DescriptionMIDL 2019 Conference Abstract Paper 46
Persistent Identifierhttp://hdl.handle.net/10722/276351

 

DC FieldValueLanguage
dc.contributor.authorDu, R-
dc.contributor.authorCao, P-
dc.contributor.authorHan, L-
dc.contributor.authorAi, Q-
dc.contributor.authorKing, AD-
dc.contributor.authorVardhanabhuti, V-
dc.date.accessioned2019-09-10T03:01:23Z-
dc.date.available2019-09-10T03:01:23Z-
dc.date.issued2019-
dc.identifier.citationInternational Conference on Medical Imaging with Deep Learning (MIDL), London, UK, 8‑10 July 2019-
dc.identifier.urihttp://hdl.handle.net/10722/276351-
dc.descriptionMIDL 2019 Conference Abstract Paper 46-
dc.description.abstractNasopharyngeal Carcinoma (NPC) is endemic cancer in the south-east Asia with approximately a third of the world incidences. With the advent of intensity-modulated radiotherapy excellent locoregional control are being achieved. Consequently, this had led to pretreatment clinical staging classification to be less prognostic of outcomes such as recurrence after treatment. Alternative pretreatment strategies for prognosis of NPC after treatment are needed to provide better risk stratification for NPC. In this study we proposed a deep convolution neural network model based on contrast-enhanced T1 (T1C) and T2 weighted (T2) MRI scan to predict 3-year disease progression of NPC patient after primary treatment. For development, we retrospective obtained 596 non-metastatic NPC patients from four independent centres in Hong Kong and China. Our proposed model first performs a segmentation of the primary NPC tumour to localise the tumour, and then uses the segmentation mask as prior knowledge along with the T1C and T2 scan to classify 3-year disease progression. For segmentation, we adapted and modified a VNet to encode both T1C and T2 scan and also encoding to classify T and overall stage classification. Our results shows that the modified network performed better than baseline VNet with T1C only and also network with no T and overall classification. The classification result for 3-year disease progression achieved an AUC of 0.828 in the validation set but did not generalised well for the test set which consist of 146 patients from a different centre to the training data (AUC = 0.69). Despite the low multicentre performance, our preliminary results show that deep learning may offer prognostication of disease progression of NPC patients after treatment. One advantage of our model is that it does not require manual segmentation of the region of interest, hence reducing clinician's burden compared to other methods such as radiomics. However further development in generalising multicentre data set are needed before clinical application of deep learning models in assessment of NPC.-
dc.languageeng-
dc.relation.ispartofMedical Imaging with Deep Learning (MIDL) 2019 International Conference-
dc.subjectNasopharyngeal Carcinoma-
dc.subjectConvolution neural network-
dc.subjectProgression-free survival-
dc.subjectSegmentation-
dc.subjectClassification-
dc.titleDeep convolution neural network model for automatic risk assessment of patients with non-metastatic nasopharyngeal carcinoma-
dc.typeConference_Paper-
dc.identifier.emailCao, P: caopeng1@hku.hk-
dc.identifier.emailVardhanabhuti, V: varv@hku.hk-
dc.identifier.authorityCao, P=rp02474-
dc.identifier.authorityVardhanabhuti, V=rp01900-
dc.description.naturepublished_or_final_version-
dc.identifier.hkuros302999-
dc.publisher.placeUnited Kingdom-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats