Abstract
Depression is the most common mood disorder in the world which is also the leading cause of suicide. Depression might be diagnosed using varying modalities, as it affects many aspects of the patient, not only the patient will appear melancholic, but the patient's sleep quality and life quality might also be disrupted. In this study, we aim to extract facial landmark features unique to depression patients and build a machine learning model to classify the depression severity. The facial landmark activities considered in this study include: speed statistics of the landmarks, the standard deviation of eye pupils, and statistical features of mouth area. We found several facial landmark activity features with significant differences between healthy volunteers and depressed patients. We also successfully built machine learning models for automatic depression severity prediction.
| Original language | English |
|---|---|
| Title of host publication | 2019 58th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2019 |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 1376-1381 |
| Number of pages | 6 |
| ISBN (Electronic) | 9784907764678 |
| DOIs | |
| Publication status | Published - 2019 Sep |
| Event | 58th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2019 - Hiroshima, Japan Duration: 2019 Sep 10 → 2019 Sep 13 |
Publication series
| Name | 2019 58th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2019 |
|---|
Conference
| Conference | 58th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2019 |
|---|---|
| Country | Japan |
| City | Hiroshima |
| Period | 19/9/10 → 19/9/13 |
Fingerprint
Keywords
- Automatic diagnosis
- Depression
- Facial features
- Machine learning
ASJC Scopus subject areas
- Artificial Intelligence
- Industrial and Manufacturing Engineering
- Safety, Risk, Reliability and Quality
- Control and Optimization
- Instrumentation
Cite this
Facial Landmark Activity Features for Depression Screening. / Sumali, Brian; Mitsukura, Yasue; Tazawa, Yuuki; Kishimoto, Taishiro.
2019 58th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2019. Institute of Electrical and Electronics Engineers Inc., 2019. p. 1376-1381 8859798 (2019 58th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2019).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Facial Landmark Activity Features for Depression Screening
AU - Sumali, Brian
AU - Mitsukura, Yasue
AU - Tazawa, Yuuki
AU - Kishimoto, Taishiro
PY - 2019/9
Y1 - 2019/9
N2 - Depression is the most common mood disorder in the world which is also the leading cause of suicide. Depression might be diagnosed using varying modalities, as it affects many aspects of the patient, not only the patient will appear melancholic, but the patient's sleep quality and life quality might also be disrupted. In this study, we aim to extract facial landmark features unique to depression patients and build a machine learning model to classify the depression severity. The facial landmark activities considered in this study include: speed statistics of the landmarks, the standard deviation of eye pupils, and statistical features of mouth area. We found several facial landmark activity features with significant differences between healthy volunteers and depressed patients. We also successfully built machine learning models for automatic depression severity prediction.
AB - Depression is the most common mood disorder in the world which is also the leading cause of suicide. Depression might be diagnosed using varying modalities, as it affects many aspects of the patient, not only the patient will appear melancholic, but the patient's sleep quality and life quality might also be disrupted. In this study, we aim to extract facial landmark features unique to depression patients and build a machine learning model to classify the depression severity. The facial landmark activities considered in this study include: speed statistics of the landmarks, the standard deviation of eye pupils, and statistical features of mouth area. We found several facial landmark activity features with significant differences between healthy volunteers and depressed patients. We also successfully built machine learning models for automatic depression severity prediction.
KW - Automatic diagnosis
KW - Depression
KW - Facial features
KW - Machine learning
UR - http://www.scopus.com/inward/record.url?scp=85073875328&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85073875328&partnerID=8YFLogxK
U2 - 10.23919/SICE.2019.8859798
DO - 10.23919/SICE.2019.8859798
M3 - Conference contribution
AN - SCOPUS:85073875328
T3 - 2019 58th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2019
SP - 1376
EP - 1381
BT - 2019 58th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2019
PB - Institute of Electrical and Electronics Engineers Inc.
ER -