Abstract
BackgroundCancer has become the second leading cause of death globally. Most cancer cases are due to genetic mutations, which affect metabolism and result in facial changes.ObjectiveIn this study, we aimed to identify the facial features of patients with cancer using the deep learning technique.MethodsImages of faces of patients with cancer were collected to build the cancer face image data set. A face image data set of people without cancer was built by randomly selecting images from the publicly available MegaAge data set according to the sex and age distribution of the cancer face image data set. Each face image was preprocessed to obtain an upright centered face chip, following which the background was filtered out to exclude the effects of nonrelative factors. A residual neural network was constructed to classify cancer and noncancer cases. Transfer learning, minibatches, few epochs, L2 regulation, and random dropout training strategies were used to prevent overfitting. Moreover, guided gradient-weighted class activation mapping was used to reveal the relevant features.ResultsA total of 8124 face images of patients with cancer (men: n=3851, 47.4%; women: n=4273, 52.6%) were collected from January 2018 to January 2019. The ages of the patients ranged from 1 year to 70 years (median age 52 years). The average faces of both male and female patients with cancer displayed more obvious facial adiposity than the average faces of people without cancer, which was supported by a landmark comparison. When testing the data set, the training process was terminated after 5 epochs. The area under the receiver operating characteristic curve was 0.94, and the accuracy rate was 0.82. The main relative feature of cancer cases was facial skin, while the relative features of noncancer cases were extracted from the complementary face region.ConclusionsIn this study, we built a face data set of patients with cancer and constructed a deep learning model to classify the faces of people with and those without cancer. We found that facial skin and adiposity were closely related to the presence of cancer.
Highlights
In daily social activities, people tend to judge the health status of others based on their facial appearance
We compared the average faces of the cancer and noncancer data sets and presented the training and validation results; we showed the distinguishable facial features of people with and those without cancer, captured by the grad-class activation mapping conventional neural network (CNN) (CAM) method
We built a face data set of patients with cancer and constructed a deep learning model to classify the faces of people with and those without cancer
Summary
People tend to judge the health status of others based on their facial appearance. Skin color is malleable and may change rapidly in response to health status It provides more relevant information regarding an individual’s current physical condition [5,6,7,8,9,10]. Facial symmetry is considered to be an indicator of health because it corresponds with the ability to maintain developmental stability and resist asymmetric growth (eg, pathogens or mutation rate) [4,12,13,14,15] These studies attempt to reveal the association of perceptual facial features with physical condition, which is helpful in understanding the underlying psychological and biological reasons for an individual’s behavior and social activities. Conclusions: In this study, we built a face data set of patients with cancer and constructed a deep learning model to classify the faces of people with and those without cancer. We found that facial skin and adiposity were closely related to the presence of cancer
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
