US 11,903,685 B2
Apparatus and method for estimating bio-information
Jin Woo Choi, Suwon-si (KR); Joon-Hyuk Jang, Seoul (KR); Youn Ho Kim, Hwaseong-si (KR); Chang Mok Choi, Suwon-si (KR); Tae-Jun Park, Seoul (KR); and Kwangsub Song, Seoul (KR)
Assigned to SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR); and Industry-University Cooperation Foundation Hanyang University, Seoul (KR)
Filed by SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR); and Industry-University Cooperation Foundation Hanyang University, Seoul (KR)
Filed on Mar. 12, 2020, as Appl. No. 16/817,021.
Claims priority of application No. 10-2019-0092188 (KR), filed on Jul. 30, 2019.
Prior Publication US 2021/0030278 A1, Feb. 4, 2021
Int. Cl. A61B 5/022 (2006.01); A61B 5/021 (2006.01); A61B 5/00 (2006.01)
CPC A61B 5/022 (2013.01) [A61B 5/02116 (2013.01); A61B 5/681 (2013.01); A61B 5/6803 (2013.01); A61B 5/6843 (2013.01); A61B 5/6879 (2013.01); A61B 5/7267 (2013.01)] 7 Claims
OG exemplary drawing
 
1. An apparatus for estimating bio-information of a subject, the apparatus comprising:
a sensor part configured to measure a pulse wave signal of a body surface of the subject and configured to measure a contact force of the body surface, while the body surface is in contact with the sensor part,
wherein the sensor part comprises:
a pulse wave sensor comprising:
a light source configured to emit light onto the body surface; and
a detector configured to detect light that is reflected or scattered from the body surface; and
a force sensor configured to measure the contact force when the body surface comes into contact with the pulse wave sensor and the body surface applies pressure to the pulse wave sensor, and
a contact area sensor,
wherein the contact area sensor is configured to measure a contact area of the sensor part with which the body surface is in contact, while the body surface is in contact with the sensor part and the body surface applies pressure to the sensor part; and
a processor configured to:
extract a first feature from the measured pulse wave signal and the measured contact force;
select an estimation model from a plurality of estimation models, based on a second feature of the subject; and
estimate the bio-information, by inputting the extracted first feature in the selected estimation model, wherein the bio-information is a physiological feature of the subject,
wherein the estimation of the bio-information is performed upon the subject activating a request for bio-information in a wearable device,
wherein a first estimation model of a first group corresponding to the second feature when the second feature is greater than or equal to a predetermined threshold, is obtained by training a first deep neural network (DNN), using the first group, the first estimation model being one of the plurality of estimation models and
a second estimation model of a second group corresponding to the second feature when the second feature is less than the predetermined threshold, is obtained by training a second DNN different from the first DNN, using the second group, the second estimation model being one of the plurality of estimation models,
wherein the processor is configured to extract the first feature further based on the measured contact area.