Prof. Dehua Chen Profile

Prof. Dehua Chen

at Donghua Univ

SPIE Involvement:

Author

Publications (3)

This will count as one of your downloads.

You will have access to both the presentation and article (if available).

DOWNLOAD NOW

This content is available for download via your institution's subscription. To access this item, please sign in to your personal account.

Email or Username Forgot your username?

Password Forgot your password?

Show

Keep me signed in

No SPIE account? Create an account

Proceedings Article | 25 May 2023 Paper

Prediction of ACS mortality risk based on feature interaction and re-weighting

Wen Fang, Qiao Pan, Dehua Chen, Mei Wang

Proceedings Volume 12636, 126364T (2023) https://doi.org/10.1117/12.2675114

KEYWORDS: Machine learning, Data modeling, Transformers, Feature extraction, Diseases and disorders, Performance modeling, Cardiovascular disorders

Read Abstract +

Acute coronary syndrome (ACS) is a serious coronary heart disease with a high morbidity and mortality rate. The prognosis and the occurrence of adverse events vary according to the degree of vascular disease in patients with ACS. Early identification of high-risk groups therefore plays an important role in clinical management. Risk scoring tools are widely used in clinical practice, focus on converting specific prognostic factors into risk indices and then combining them with logistic regression to make predictions. These methods focus well on statistical correlations in the data, but ignore the instability associated with differences in the distribution of data in real data sets and also lack a focus on treatment information in the context of the disease. To address these issues, this paper proposes a model for predicting the risk of death in ACS based on feature interaction and re-weighting (i.e. FIFR Stable-Net). The model is able to take into account differences in the distribution of data features, and enhances feature learning by re-weighting risk features to reduce the impact of differences in the distribution of patient feature, thereby enhancing model stability. In addition, the model incorporates treatment features such as patient medication and applies an attention-based mechanism of feature interaction to obtain a more informative feature representation for the ACS mortality risk prediction task. To illustrate the effectiveness of the FIFR Stable-Net, we evaluated our method on a real dataset of ACS patients and compared it experimentally with the baseline models. The results showed that our model outperformed other models in terms of AUC and other assessment metrics, and AUC can reach up to 91.57%.

Proceedings Article | 6 December 2022 Paper

Breast tumor diagnosis via phrase level self-attention mechanism

Dehua Chen, Orlando Mayugi

Proceedings Volume 12458, 124583F (2022) https://doi.org/10.1117/12.2660676

KEYWORDS: Tumors, Breast, Associative arrays, Mammography, Breast cancer, Tumor growth modeling, Tissues, Neural networks, Statistical modeling, Software development

Read Abstract +

In recent years there has been a surge in female Breast Cancer patients leading to an increase in the administration of Electronic Health Records (EHR) data during the treatment process. So far, mammogram examinations are the main means of breast cancer detection. The data obtained from such medical reports remains an important source for constructing AI diagnostic models for breast cancer. With a purpose of contributing and assisting doctors in clinical decision-making through provision of high-quality and efficient tumor diagnosis using AI, this paper explores and trains the breast tumor X-ray examination report, and constructs the breast tumor classification neural network (BTDNN) model to understand the auxiliary diagnosis of breast tumor. The proposed model (BTDNN) builds and uses a semantic network to define and structure the standard for breast cancer X-ray report medical entities and entity annotations which act as the standardized input of the model. Based on the semantic network, we also built a breast cancer diagnosis model based on Phrase level self-attention mechanism that uses Phrase level context technology to train the model for analysing medical reports in the same concept as a health care professional with original, local and global method context, in order to improve the prediction accuracy. As result, we compared our proposed model to several other models and concluded that combining the semantic and contextual technology could improve the final prediction accuracy of a medical diagnosis.

Proceedings Article | 9 October 2022 Paper

An interpretable classification model of breast tumors with tabular mammography data

Xiaomeng Gao, Dehua Chen, Qiao Pan

Proceedings Volume 12246, 122461M (2022) https://doi.org/10.1117/12.2643654

KEYWORDS: Breast, Mammography, Classification systems, Breast cancer, Tumors

Read Abstract +

Breast cancer is the most prevalent malignant disease among cancers, and its mammography and examination reports are an important basis for clinicians to judge the nature of breast tumours. Current research based on mammography reports has focused on the natural language classification models, ignoring the practical problems of the semantic hierarchy in examination reports and the imbalance of benign and malignant categories. To address this issue, we propose an interpretable classification model of breast tumors with tabular mammography data (STaNet). The model is based on a semantic tree to structure the image presentation part of the mammography report, allowing it to focus on the disease information carried at the semantic level of the report better. Meanwhile, we dynamically conbine the focus loss function parameters with the category loss, allowing the model to dynamically adjust its focus category as it learns. We evaluated our model using real-world mammogram report data and compared the model with other excellent models. The experimental results show that the accuracy of our proposed model is better than other models and its interpretable results are in line with the diagnostic criteria of pathologists.

My Library

You currently do not have any folders to save your paper to! Create a new folder below.

Folder Name

Folder Description

View contact details

UPDATE YOUR PROFILE

Is this your profile? Update it now.

Sign into your SPIE.org account

Don’t have a profile and want one?

Create an account on SPIE.org

Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks. You are receiving this notice because your organization may not have SPIE eBooks access.*

*Shibboleth/Open Athens users─please sign in to access your institution's subscriptions.

To obtain this item, you may purchase the complete book in print or electronic format on SPIE.org.

ORGANIZATIONAL
Sign in with credentials provided by your organization.

Organizational Username

Organizational Password

Show/Hide Password

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members:

Non-members: ADD TO CART

Keywords/Phrases

Search In:

Publication Years