Ingestion monitoring systems

1. An ingestion monitoring system comprising:

a wristband attachable on a wrist of a user, the wristband comprising

a housing,

a first acoustic transducer mounted on the housing so as to face the wrist to sense vibrations transmitted through the body of the user,

a second acoustic transducer mounted on the housing so as to be isolated from the wrist to sense airborne sounds,

a motion sensor, a heart rate monitor, a photoplethysmographic sensor and an analogue to digital converter arranged to convert electrical signals from the acoustic transducers, the motion sensor, the heart rate monitor and the photoplethysmographic sensor to digital signals; and

a processor arranged to process the digital signals to detect an ingestion event, wherein the processor is adapted to identify a movement of the user's arm detected by the motion sensor as a pattern which correlates with a given action, including cutlery use or eating by hand and physical exercise;

wherein food or drink consumption is inferred when there is an increase in heart rate and blood profusion in subcutaneous blood vessels without a corresponding motion of the arm that correlates to exercise.

2. The system of claim 1 , further comprising one or more other types of sensors which provide ancillary data for correlation with acoustic signals, wherein one sample from each of the sensor types is considered as a vector and an event is a time evolution of that vector.

3. The system of claim 2 , wherein raw signals generated by the sensors are transformed into a parameter space having a plurality of axis that coincide with indicators for given actions, wherein thresholding is applied directly to the indicators, whereby an action is deemed to have occurred when its threshold is exceeded.

4. The system of claim 2 , comprising a signal classification model used to classify detected signals from the acoustic transducers and the said one or more other types of sensors wherein the signal classification model employs a Gaussian Mixture Model technique, a Hidden Markov Model technique, or a hybrid technique employing both Gaussian Mixture Model and Hidden Markov Model techniques.

5. The system of claim 1 , wherein low frequency components are observed after low pass filtering and high frequency components are observed by transformation into a frequency domain over a sampling window.

6. The system of claim 1 , wherein the processor is configured to execute a statistically derived model which assigns probable causes to signal events and over a series of events infers an activity that is taking place.

7. The system of claim 1 , wherein the movement is classified as being vigorous, minimal, or still.

8. The system of claim 1 , comprising a temperature sensor.

9. The system of claim 1 , wherein the airborne sounds and the vibrations transmitted through the body of the user are correlated and combined to form part of a signature relating to an event.

10. The system of claim 1 , comprising a portable computing device, and one or more fixed location acoustic transducers.

11. The system of claim 10 wherein the portable computing device comprises other types of sensors which provide ancillary data for correlation with acoustic signals.

12. The system of claim 10 , wherein the computing device is adapted to identify sounds originating from one particular user based on the proximity of the portable computing device to one or more of the fixed location acoustic transducers.

13. The system of claim 1 , wherein the wristband comprises a filter for filtering out high frequency noise prior to analogue to digital conversion and a communication circuitry for transmitting sensor data to a remote location for identification.

14. The system of claim 13 , wherein the processor is comprised in a remote server and wherein the server is configured to run a statistically derived model or signal classification model.

15. The system of claim 1 , wherein the processor is comprised in the wristband.

16. The system of claim 1 , wherein the photoplethysmographic sensor is arranged to detect a period of apnoea related to a swallowing event, wherein a swallowing event is detected when a period of apnoea is correlated with an acoustic swallowing signature from airborne or intra-body vibrations.

17. A method of monitoring ingestion, comprising

providing a wristband attachable on a wrist of a user, the wristband comprising a housing, a first acoustic transducer mounted on the housing so as to face the wrist to sense vibrations transmitted through the body of the user and a second acoustic transducer mounted on the housing so as to be isolated from the wrist to sense airborne sounds, a motion sensor, a heart rate monitor, a photoplethysmographic sensor and an analogue to digital converter;

converting analogue electrical signals from the acoustic transducers, the motion sensor, the heart rate monitor and the photoplethysmographic sensor to digital signals;

providing a processor;

processing the digital signals to detect an ingestion event, wherein the processor is adapted to identify a movement of the user's arm detected by the motion sensor as a pattern which correlates with a given action, including cutlery use or eating by hand and physical exercise; and

inferring food or drink consumption when there is an increase in heart rate and blood profusion in subcutaneous blood vessels without a corresponding motion of the arm that correlates to exercise.

18. A wristband attachable on a wrist of a user, the wristband comprising a housing, a first acoustic transducer mounted on the housing so as to face the wrist to sense vibrations transmitted through the body of the user and a second acoustic transducer mounted on the housing so as to be isolated from the wrist to sense airborne sounds, a motion sensor, a heart rate monitor, a photoplethysmographic sensor, an analogue to digital converter arranged to convert electrical signals from the acoustic transducers, the motion sensor, the heart rate monitor and the a photoplethysmographic sensor to digital signals, and a communication circuitry for sending the digital signals to a remote location for processing to detect an ingestion event, to identify a movement of the user's arm detected by the motion sensor as a pattern which correlates with a given action, including cutlery use or eating by hand and physical exercise, and to infer food or drink consumption when there is an increase in heart rate and blood profusion in subcutaneous blood vessels without a corresponding motion of the arm that correlates to exercise.