Non-Invasive Measurement of Second Heart Sound Components

• US 20080091115A1
• Filed: 04/15/2005
• Published: 04/17/2008
• Est. Priority Date: 04/16/2004
• Status: Active Grant

• Current Assignee: Masimo Corporation
• Original Assignee: Masimo Corporation
• Inventors: Popov, Boris, Agarwal, Rajeev, Lanzo, Victor
• Granted Patent: US 7,909,772 B2
• Time in Patent Office: Days
• US Class Current: 600/528
• CPC Class Codes: A61B 5/021 Measuring pressure in heart or blood vessels A61B5/0205 takes precedence
• CPC Class Codes: A61B 5/352 Detecting R peaks, e.g. for synchronising diagnostic apparatus; Estimating R-R interval
• CPC Class Codes: A61B 5/369 Electroencephalography [EEG] devices for psychotechnics A61B5/16
• CPC Class Codes: A61B 7/04 Electric stethoscopes microphones, acoustic transducers therefor H04R

Abstract

A method and apparatus for estimating a location of pulmonary and aortic components of second heart sounds of a patient over an interval. The method comprises the steps of producing an electronic representation of heart sounds of the patient over the interval, identifying at least one second heart sound in the interval using the electronic representation, for each identified second heart sound generating an estimated value for a location of the aortic component and the pulmonary component. There is also included a method for using the estimated location of the aortic component and the pulmonary component for estimation of the blood pressure in the pulmonary artery of a patient.

Claims

1. A method for estimating a location of pulmonary and aortic components of second heart sounds of a patient over an interval, the method comprising the steps of:

producing an electronic representation of heart sounds of the patient over the interval;

identifying at least one second heart sound in the interval using said electronic representation;

for each identified second heart sound;

calculating a frequency weighted energy (FWE);

normalizing said FWE;

identifying peaks in said FWE;

determining a maximum peak from said identified peaks; and

retaining said maximum peak and peaks having an amplitude within a predetermined amount of an amplitude of said maximum peak;

wherein if two or more peaks are retained, two largest peaks are selected, a first peak as a candidate value for the aortic component and a second peak as a candidate value for the pulmonary component, wherein said first peak is prior to said second peak and wherein if only a single peak is retained, said single peak is selected as a candidate value for the aortic component; and

generating an estimated value for a location of the aortic component and the pulmonary component from said candidate values.

65. The method of claim 1, wherein said second heart sound identifying step comprises framing said second heart sound using a reference signal.

66. The method of claim 65, wherein said reference signal is an ECG of the patient.

67. The method of claim 1, wherein said FWE calculating step comprises generating a Non Linear Energy Operator (NLEO) function for each identified second heart sound.

68. The method of claim 67, wherein said NLEO function is

Ψ

[n]=x(n−

l)·

x(n−

m)−

x(n−

p)·

x(n−

q) for l+m=p+q, l≠

m and p≠

q.

69. The method of claim 68, wherein l=2, m=3, q=4 and p=1.

70. The method of claim 67, wherein in said peak identifying step, a maximum peak value is identified and any identified peaks having a value less than a predetermined amount of said maximum peak value are discarded.

71. The method of claim 70, wherein said predetermined amount is between 90% and 100%.

72. The method of claim 1, wherein said peak identifying step comprises identifying all local maxima which are greater than a predetermined distance apart, wherein each of said identified local maxima is a peak.

73. The method of claim 72, wherein said predetermined distance is at least 10 ms.

74. The method of claim 1, wherein said candidate value is within 0 and 10% of said maximum peak amplitude.

75. The method of claim 1, wherein said estimated value generating step comprises:

generating a first probability distribution of locations of said remaining aortic component candidate values relative to a predetermined point of their respective second heart sounds and a second probability distribution of locations of said remaining pulmonary component candidate values relative to a predetermined point of their respective second heart sounds; and

identifying a first local maximum in said first probability distribution and a second local maximum in said second probability distribution;

wherein a location of said first local maximum is the location of the aortic component and a location of said second local maximum is the location of the pulmonary component.

76. The method of claim 75, wherein said predetermined point is a start of said respective second heart sound.

77. The method of claim 1, further comprising, for each identified second heart sound and prior to said FWE calculating step, a signal to noise ratio (SNR) calculating step wherein when said calculated SNR is below a predetermined ratio, no candidate values for the aortic component or the pulmonary component are selected.

78. The method of claim 77, wherein said SNR calculating step comprises detecting a start and finish of said identified second heart sound, calculating an average signal energy between said start and finish, calculating an average noise energy for a predetermined period before said start and after said finish, wherein said SNR is equal to said average signal energy divided by said average noise energy.

79. The method of claim 78, wherein said predetermined period is 50 ms.

80. The method of claim 77, wherein said predetermined ratio is 0.50.

2-64. -64. (canceled)

81. A method for estimating a location of pulmonary and aortic components of second heart sounds of a patient over an interval, the method comprising the steps of:

producing an electronic representation of heart sounds of the patient over the interval;

dividing said electronic representation into a plurality of sub-channels;

for each of said sub-channel representations;

identifying at least one second heart sound in the interval using said electronic representation; and

extracting an estimated location of a sub-channel aortic component and a sub-channel pulmonary component from said at least one second heart sound;

combining said estimated sub-channel aortic component locations to form the estimated aortic component location and said estimated sub-channel pulmonary component locations to form the estimated pulmonary component location.

82. The method of claim 81, wherein said extracting step comprises:

for each identified second heart sound;

calculating a FWE;

normalizing said FWE;

identifying peaks in said FWE;

determining a maximum peak from said identified peaks; and

retaining said maximum peak and peaks having an amplitude within a predetermined amount of an amplitude of said maximum peak;

wherein if two or more peaks are retained, two largest peaks are selected, a first peak as a candidate value for the aortic component and a second peak as a candidate value for the pulmonary component, wherein said first peak is prior to said second peak and wherein if only a single peak is retained, said single peak is selected as a candidate value for the aortic component; and

generating said estimated location of an aortic component and a pulmonary component from said candidate values.

83. The method of claim 81, wherein said electronic representation is divided into a low frequency channel, a medium frequency channel and a high frequency channel.

84. The method of claim 83, wherein said low frequency channel is from 30 to 50 Hz, said medium frequency channel from 50 to 120 Hz and said high frequency channel from 120 to 200 Hz.

85. A method for estimating a location of pulmonary and aortic components of second heart sounds a patient over an interval, the method comprising the steps of:

positioning a first transducer at a first position on the patient, said first transducer producing a first electronic representation of heart sounds of the patient over the interval;

positioning a second transducer at a second position on the patient, said second transducer producing a second electronic representation of heart sounds of the patient over the interval;

for said first electronic representation;

identifying at least one second heart sound in the interval;

for each identified second heart sound;

calculating a FWE;

normalizing said FWE;

identifying peaks in said FWE;

determining a maximum peak from said identified peaks; and

retaining said maximum peak and peaks having an amplitude within a predetermined amount of an amplitude of said maximum peak;

wherein if two or more peaks are retained, two largest peaks are selected, a first peak as a candidate value for the aortic component and a second peak as a candidate value for the pulmonary component, wherein said first peak is prior to said second peak and wherein if only a single peak is retained, said single peak is selected as a candidate value for the aortic component; and

generating a first estimated value for a location of an aortic component and a pulmonary component from said candidate values; and

for said each electronic representation;

identifying at least one second heart sound in the interval;

for each identified second heart sound;

calculating a FWE;

normalizing said FWE;

identifying peaks in said FWE;

determining a maximum peak from said identified peaks; and

retaining said maximum peak and peaks having an amplitude within a predetermined amount of an amplitude of said maximum peak;

wherein if two or more peaks are retained, two largest peaks are selected, a first peak as a candidate value for the aortic component and a second peak as a candidate value for the pulmonary component, wherein said first peak is prior to said second peak and wherein if only a single peak is retained, said single peak is selected as a candidate value for the aortic component; and

generating second estimated values for a location of the aortic component and the pulmonary component from said candidate values; and

combining said first and second estimated aortic location values and said first and second estimated pulmonary location values;

wherein the estimated location of the aortic component is said combined first and second estimated aortic location values and the estimated location of the pulmonary components is said combined first and second estimated pulmonary location values.

86. The method of claim 85, wherein said first position is at an apex of the patient'"'"'s heart.

87. The method of claim 85, wherein said second position is between the patient'"'"'s 3^rd and 4^th left intercostals space.

88. The method of claim 85, wherein said low frequency channel is from 30 to 50 Hz, said medium frequency channel is from 50 to 150 Hz and said high frequency channel is from 120 to 200 Hz.

89. A method for estimating pulmonary artery pressure of a patient over an interval, the method comprising the steps of:

producing an electronic representation of heart sounds of the patient over the interval;

identifying at least one second heart sound in the interval using said electronic representation;

for each identified second heart sound;

calculating a FWE;

normalizing said FWE;

identifying peaks in said FWE;

determining a maximum peak from said identified peaks; and

retaining said maximum peak and peaks having an amplitude within a predetermined amount of an amplitude of said maximum peak;

wherein if two or more peaks are retained, two largest peaks are selected, a first peak as a candidate value for the aortic component and a second peak as a candidate value for the pulmonary component, wherein said first peak is prior to said second peak and wherein if only a single peak is retained, said single peak is selected as a candidate value for the aortic component; and

generating an estimated value for a location of an aortic component and a location of pulmonary component from said candidate values;

determining a splitting interval as a time between said aortic component location and said pulmonary component location;

normalizing said splitting interval; and

estimating the systolic pulmonary artery pressure using a predetermined function which describes a relationship between said normalized splitting interval and the systolic and diastolic pulmonary artery pressures.

90. A device for estimating a location of pulmonary and aortic components of second heart sounds of a patient over an interval, the device comprising:

at least one transducer for sensing heart sounds of the patient, said heart sounds comprising, during motion of the patient, a first sound portion indicative of second heart sounds and a second sound portion indicative of motion induced noise; and

a second heart sound processor to identify second heart sounds from said sensed heart sounds and compute an aortic component location and an estimated pulmonary component location from said identified second heart sounds without significant interference in said identification and computation from the motioned induced noise portion of said sensed heart sounds,

91. The device of claim 90, wherein said second heart sound processor comprises:

a means for identifying at least one second heart sound in the interval using said sensed heart sounds;

a means for calculating a frequency weighted energy (FWE) of each identified second heart sound;

a means for normalizing said calculated FWE of each identified second heart sound;

a means for identifying peaks in said calculated FWE of each identified second heart sound and determining a maximum peak from said identified peaks; and

a means for retaining said maximum peak and peaks having an amplitude within a predetermined amount of an amplitude of said maximum peak wherein if two or more peaks are retained, two largest peaks are selected, a first peak as a candidate value for the aortic component and a second peak as a candidate value for the pulmonary component, wherein said first peak is prior to said second peak and wherein if only a single peak is retained, said single peak is selected as a candidate value for the aortic component; and

a means for generating an estimated value for a location of the aortic component and the pulmonary component from said candidate values.

92. The device of claim 91, wherein said second heart sound identifying means frames said second heart using a reference signal.

93. The device of claim 92, wherein said reference signal is an ECG of the patient.

94. The device of claim 91, wherein said FWE calculating means generates a Non Linear Energy Operator (NLEO) function for each identified second heart sound.

95. The device of claim 94, wherein said NLEO function is:

Ψ

[n]=x(n−

l)·

x(n−

m)−

x(n−

p)·

x(n−

q) for l+m=p+q, l≠

m and p≠

q.

96. The device of claim 95, wherein l=2, m=3, q=4 and p=1.

97. The device of claim 94, wherein said peak identifying means identifies a maximum peak value and discards any identified peak having a value less than a predetermined amount of said maximum peak value.

98. The device of claim 97, wherein said predetermined amount is between 90% and 100%.

99. The device of claim 91, wherein said peak identifying means identifies all local maxima which are greater than a predetermined distance apart, wherein each of said identified local maxima is a peak.

100. The device of claim 99, wherein said predetermined distance is at least 10 ms.

101. The device of claim 91, wherein said candidate value is within 0 and 10% of said maximum peak amplitude.

102. A device for estimating the systolic and diastolic pulmonary artery pressure of a patient over an interval, the device comprising:

at least one transducer for sensing heart sounds of the patient, said heart sounds comprising, during motion of the patient, a first sound portion indicative of second heart sounds and a second sound portion indicative of motion induced noise; and

a processor, said processor;

computing an aortic component location and a pulmonary component location from said identified second heart sounds without significant interference in said computation from the motioned induced noise portion of said sensed heart sounds;

calculating a normalized splitting interval from said aortic component location and said pulmonary component location; and

estimating the systolic and diastolic pulmonary artery pressure using a predetermined function which describes a relationship between said normalized splitting interval and the systolic and diastolic pulmonary artery pressures.

103. The device of claim 102, further comprising a display means for displaying the systolic and diastolic pulmonary artery pressures.