Artificial pancreas with neural signal input

1. Apparatus for controlling the delivery of insulin to a subject and comprising:

a first input for receiving neural information recorded from the subject's nervous system and indicative of prospective or actual food and/or drink intake;

a second input for receiving a signal indicative of the subject's blood glucose level;

a processor or processors configured to determine food and/or drink characteristics based on the received neural information, determine an amount of insulin to be delivered based on the measured blood glucose levels and the determined food and/or drink characteristics, and output an insulin pump control signal indicative of the determined amount of insulin; and

a third input for receiving a neural signal recorded from the subject's nerves innervating the liver and/or brown adipose tissue and/or nuclei of the hypothalamus indicative of a subject's insulin sensitivity;

said processor(s) being configured to determine a characteristic of a neural stimulation signal to be applied by a neural stimulation unit using said signal indicative of a subject's insulin sensitivity, and the apparatus comprising a second output for providing neural stimulation unit control signal indicative of said characteristic; and

said neural stimulation unit being configured to apply said neural stimulation signal to efferent sympathetic nerves to the liver and/or brown adipose tissue and/or hypothalamic areas in order to increase or decrease the subject's insulin sensitivity during delivery of the determined amount of insulin.

2. Apparatus according to claim 1 , wherein the processor(s) are configured to determine an amount of insulin to be delivered as one or both of a basal insulin dose or a meal bolus insulin dose.

3. Apparatus according to claim 1 , said processor(s) being configured to implement a learning mode in which signals received via said first input are classified according to the food and/or drink intake to which they relate and to store the results in a memory of the apparatus, the processor(s) being configured to determine food and/or drink characteristics based on a received neural information by analysing the signal and performing a look-up in said memory.

4. Apparatus according to claim 1 and comprising one or more of:

an insulin pump connected or connectable to receive the insulin pump control signal;

a body worn or implantable blood glucose sensor connected or connectable to said second input; and

one or more body worn or implantable neural sensors connected or connectable to said first input.

5. Apparatus according to claim 4 wherein said neural stimulation unit comprises a body worn or implantable neural stimulation unit connected or connectable to said output.

6. Apparatus according to claim 1 , said processor(s) being configured to implement a learning mode in which signals received via said third input are mapped to said characteristic of the neural stimulation signal to be applied by said one or more body worn or implantable electrodes, and the characteristic of the neural stimulation signal being at least one of an output current or voltage, pulse width, frequency, or on time and off time.

7. A method of controlling the delivery of insulin to a subject and comprising:

recording, using one or more body worn or implantable electrodes, neural information from the subject's nervous system indicative of prospective or actual food and/or drink intake;

determining food and/or drink characteristics based on the recorded neural information;

measuring the subject's blood glucose level using a body worn or implantable blood plasma glucose sensor;

determining an amount of insulin to be delivered based on the measured blood glucose level and the determined food and/or drink characteristics; and

providing a control signal indicative of the determined amount of insulin to an insulin pump to thereby deliver a required amount of insulin from the insulin pump to the subject;

recording, using one or more body worn or implantable electrodes, a neural signal from nerves innervating the liver and/or brown adipose tissue and/or nuclei of the hypothalamus indicative of the subject's insulin sensitivity;

determining a characteristic of a neural stimulation signal to be applied to the subject using said neural signal indicative of the subject's insulin sensitivity; and

applying that stimulation signal to the subject's efferent sympathetic nerves to the liver and/or brown adipose tissue and/or hypothalamic areas via one or more body worn or implantable electrodes in order to increase or decrease the subject's insulin sensitivity during delivery of the determined amount of insulin.

8. A method according to claim 7 , wherein said blood glucose sensor is an implanted, subcutaneous blood glucose sensor.

9. A method according to claim 7 , wherein said recording is performed using one or more body worn electrodes configured to record neural signals.

10. A method according to claim 7 , wherein said recording of neural information from the subject's nervous system indicative of prospective or actual food and/or drink intake is performed using one or more implanted electrodes configured to detect cephalic signals in efferent fibres in the vagus nerve to the pancreas and/or in afferent fibres in the vagus nerve from the stomach and/or to detect signals indicative of insulin sensitivity from afferent sympathetic nerves innervating a liver and/or brown adipose tissue and/or nuclei of hypothalamic areas.

11. A method according to claim 7 , wherein said recording of neural information from the subject's nervous system indicative of prospective or actual food and/or drink intake is performed using one or more electrodes configured to monitor neural signals in the orbitofrontal cortex.