US 11,918,372 B2
Cognitive function evaluation system, method, and storage medium for dementia by analyzing voice of evaluatee for recognition, remembering or jugment capabilities
Sadayuki Sumi, Hyogo (JP); Ryosuke Nagumo, Osaka (JP); Kengo Abe, Nara (JP); Yoshihiro Matsumura, Osaka (JP); Takashi Nishiyama, Hyogo (JP); Hirobumi Nakajima, Kyoto (JP); Kohji Sasabe, Osaka (JP); Makoto Kariyasu, Kyoto (JP); Takako Yoshimura, Hyogo (JP); and Minoru Toyama, Oita (JP)
Assigned to Panasonic Intellectual Property Management Co., Ltd., Osaka (JP)
Appl. No. 16/759,573
Filed by Panasonic Intellectual Property Management Co., Ltd., Osaka (JP)
PCT Filed Oct. 15, 2018, PCT No. PCT/JP2018/038346
§ 371(c)(1), (2) Date Apr. 27, 2020,
PCT Pub. No. WO2019/087758, PCT Pub. Date May 9, 2019.
Claims priority of application No. 2017-213157 (JP), filed on Nov. 2, 2017.
Prior Publication US 2021/0177340 A1, Jun. 17, 2021
Int. Cl. A61B 5/00 (2006.01)
CPC A61B 5/4803 (2013.01) [A61B 5/4088 (2013.01); A61B 5/72 (2013.01)] 13 Claims
OG exemplary drawing
 
1. A cognitive function evaluation system, comprising:
a display device configured to display an image that indicates sentence data indicating a sentence to be uttered by an evaluatee, wherein the sentence to be uttered by the evaluatee is a sentence containing five or more characters, each character consisting of a stop consonant and a subsequent vowel;
a microphone configured to obtain utterance data indicating a voice of the evaluatee uttering the sentence as instructed by the image displayed by the display device;
computer configured to calculate, from the utterance data obtained by the microphone, a feature based on the utterance data, wherein the computer is further configured to compare the calculated feature to reference data indicating a relationship between voice data indicating a voice of a person and a cognitive function of the person to evaluate the cognitive function of the evaluatee; and
the display device is configured to output the sentence to be uttered by the evaluatee and output a result of evaluation by the computer, wherein
the sentence contains a character string of consecutive syllables, each including a vowel, and
wherein the sentence further contains at least three syllables, each including a vowel different from vowels of other syllables,
the computer is configured to calculate, as the feature, at least one of amounts of changes in a first formant frequency and a second formant frequency of a vowel, a number of times required for the changes in the first formant frequency and the second formant frequency of the vowel, or rates of changes that are ratios of the amounts of changes to the number of times required, and
further wherein computer calculates, as the feature, at least one of a shape or an area of a polygon defined by plotting a ratio of a second formant frequency to a first formant frequency calculated from the vowel of each of the at least three syllables in a coordinate space defined by the second formant frequency of the vowel with respect to the first formant frequency of the vowel; and
the display device further outputs an instruction for causing the evaluatee to utter the sentence a plurality of times,
the computer is further configured to calculate, as the feature, an amount of change in reading time calculated from reading times obtained for the plurality of times the evaluatee uttered the sentence, and further wherein the computer is configured to identify the vowel in the voice data and calculate the feature based on a formant obtained from a spectrum of the vowel identified; and
the computer further configured to evaluate that the cognitive function of the evaluatee decreases as a variation in the formant increases.