The present invention relates to spirometry apparatus and in particular to spirometry apparatus for interpretation of flow-volume loops to provide diagnosis of a respiratory disease or condition.

Spirometry is the most common of the pulmonary function tests (PFTs). It measures lung function, specifically the amount (volume) and/or speed (flow) of air that can be inhaled and exhaled.

Spirometry is helpful in assessing breathing patterns that identify conditions such as asthma, pulmonary fibrosis, cystic fibrosis, and COPD. It is also helpful as part of a system of health surveillance, in which breathing patterns are measured over time.

The spirometry test is performed using a device called a spirometer, which comes in several different varieties. Spirometry generates graphs that plot the volume and flow of air coming in and out of the lungs from one inhalation and one exhalation.

Most spirometers display the following graphs, called spirograms:

• a volume-time curve, showing volume (litres) along the Y-axis and time (seconds) along the X-axis

• a flow-volume loop, which graphically depicts the rate of airflow on the Y-axis and the total volume inspired or expired on the X-axis

The most common parameters measured in spirometry are Vital capacity (VC), Forced vital capacity (FVC), Forced expiratory volume (FEV) at timed intervals of 0.5, 1.0 (FEV1), 2.0, and 3.0 seconds, forced expiratory flow 25-75% (FEF 25-75) and maximal voluntary ventilation (MVV), also known as Maximum breathing capacity.

Results are usually given in both raw data (litres, litres per second) and percent predicted— the test result as a percent of the "predicted values" for the patients of similar characteristics (height, age, sex).

Spirometers are becoming increasingly used in primary care as automated diagnosis of the common respiratory diseases is available using published algorithms. These algorithms use measurements derived from the exhalation profile. For example, FEV1 (Forced Expiratory Volume in the 1 ^st second) and FVC (Forced Vital Capacity) can be used to distinguish obstructive and restrictive disease.

Flowever, certain less common respiratory diseases or conditions can only currently be diagnosed by examination of a graph of flow vs volume by a specialist clinician. Automated diagnosis based on measurements derived from the exhalation profile may result in misdiagnosis of these less common respiratory diseases or conditions and consequently inadequate or inappropriate medical treatment.

If the forced expiration is immediately followed by a forced inspiration, then the flow vs volume graph becomes a loop as all the air expelled is returned to the lungs. Examination of the complete loop may can be used to more accurately distinguish respiratory diseases.

Flow-volume loops which typify certain respiratory disease or conditions are shown in Figure 1.

Spirometry guidelines require the flow-volume loop to be examined to confirm the diagnosis of obstruction and restriction and to eliminate other, less common, respiratory diseases and conditions ('Practical Handbook of Spirometry ¹ by Brendan Cooper, Jodie Flunt, Adrian H Kendrick, Vicky Moore T refor Watts, third edition, pp321-327 (Association for Respiratory Technology & Physiology, 2016); Miller M. R. et al (2005) SERIES "ATS/ERS TASK FORCE: STANDARDISATION OF LUNG FUNCTION TESTING, Ear. Respir. J., 26, pp319-338)

However, not all primary care clinicians have the required experience to recognise the different patterns of flow-volume loops.

According to a first aspect of the present invention, there is provided a spirometry apparatus comprising:

an air inlet;

an air outlet;

a sensor located between the air inlet and the air inlet which measures the instantaneous air flow during a forced expiration by a user from the air inlet to the air outlet and/or a forced inspiration by a user from the air outlet to the air inlet;

a transmitter to transmit flow data from the sensor to a processor, wherein the processor processes the flow data;

characterised in that the apparatus further includes a data storage memory; and the processor compares the flow data from the sensor to data stored on the data storage memory. In this context, the air inlet is an opening into which, in use, the user provides the forced expiration and/or inspiration. Thus, when the user provides the forced expiration, the air inet allows inflow of air. Conversely, when the user provides the forced inspiration, the air inlet allows outflow of air.

In this context, the air outlet is an opening which allows outflow and/or inflow of air when, in use, the user provides, respectively, the forced expiration and/or inspiration. Thus, when the user provides the forced expiration, the air outlet allows outflow of air. Conversely, when the user provides the forced inspiration, the air outlet allows inflow of air.

In one embodiment, the processor processes the flow data to generate a flow-volume loop and the generated flow-volume loop is compared to data stored on the data storage memory.

In one embodiment, the flow data comprises one or more parameters selected from Vital capacity (VC), Forced vital capacity (FVC), Forced expiratory volume (FEV) at timed intervals of 0.5, 1.0 (FEV1), 2.0, and 3.0 seconds, forced expiratory flow 25-75% (FEF 25-75) and maximal voluntary ventilation (MVV).

In one embodiment, the data storage memory includes flow data related to respiratory disorders or conditions, and the processor includes an algorithm which compares the flow data from the sensor to the data stored on the data storage memory to identify an associated respiratory disorder or condition. The flow data suitably includes data relating to flow-volume loops.

In a further embodiment, the data storage memory includes flow data related to respiratory disorders or conditions, and the processor includes an algorithm which compares the generated flow-volume loop to the data stored on the data storage memory to identify an associated respiratory disorder or condition.

In a yet further embodiment, the data storage memory includes flow-volume loop data related to respiratory disorders or conditions, and the processor includes an algorithm which compares the generated flow-volume loop to the data stored on the data storage memory to identify an associated respiratory disorder or condition. In one embodiment, the apparatus further includes a display and an output from the processor is displayed on the display.

The present invention utilises an algorithm derived from machine learning in order to replicate the expertise of an expert respiratory clinician in identifying a respiratory disease from the characteristics of the flow-volume loop. Machine learning techniques generate an algorithm that can replicate the clinician's interpretation using data obtained by analysis of observed flow-volume loops that represent a range of respiratory conditions by an expert respiratory clinician. The observed flow- volume loops represent anonymised patient data. The machine learning may use statistical techniques such as support vector machine or neural network learning.

In one embodiment, the algorithm is derived from machine learning using data obtained by analysis of observed flow-volume loops that represent a range of respiratory conditions by an expert respiratory clinician.

The algorithm may be incorporated into the computer spirometry software code, enabling an expert interpretation of the flow-volume loop to be displayed.

In one embodiment, the sensor is a turbine flow transducer.

According to a second aspect of the invention, there is provided a method of determining at least one pulmonary measurement comprising:

i) measuring the instantaneous air flow during a forced expiration by a user and/or a forced inspiration by a user using the spirometry apparatus of the present invention;

ii) transmitting flow data from the sensor to the processor, wherein the processor processes the flow data to generate a flow-volume loop;

iii) comparing the generated flow-volume loop to data stored on the data storage memory.

In one embodiment, the data storage memory includes flow data related to respiratory disorders or conditions, and the processor includes an algorithm which compares the flow data from the sensor to the data stored on the data storage memory to identify an associated respiratory disorder or condition. The skilled person will appreciate that the features specified above in connection with embodiments of the invention may be combined with each other and any of the aspects of the invention as defined. Thus, the present invention includes within its scope an aspect of the invention combined with two or more of the features described anywhere herein as optional features. All such combinations of features described herein are considered to be made available to the skilled person.

An Embodiment of the invention will now be described in detail, by way of example only, with reference to the accompanying drawings in which:

Figure 1 shows flow-volume loops which typify certain respiratory disease or conditions;

Figure 2 is a schematic representation of a spirometry apparatus of the invention; and

Figure 3 is a representation of part of an embodiment of spirometry apparatus of the invention wherein the sensor is a turbine flow transducer.

For the avoidance of doubt, the skilled person will appreciate that in this specification, the terms "up", "down", "front", "rear", "upper", "lower", "width", etc. refer to the orientation of the components as found in the game when arranged for normal use as shown in the Figures.

An embodiment of the spirometry apparatus 100 of the invention is shown in Figure 2. The spirometry apparatus 100 comprises an air inlet 101a and an air outlet 101b. In use, the user provides a forced expiration and/or a forced inspiration via the air inlet 101a. A sensor 102 located between the air inlet 101a and the air outlet 101b measures the instantaneous air flow during a forced expiration and/or a forced inspiration by a user from the air inlet to the air outlet; a transmitter 103 transmits flow data measured by the sensor 102 from the sensor 102 to a processor 104. The processor 104 processes the flow data to generate a flow-volume loop. The spirometry apparatus 100 further includes a data storage memory 105 and the processor 104 includes an algorithm which compares the flow data from the sensor 102 to data stored on the data storage memory 105 to identify an associated respiratory disorder or condition.

Part of an embodiment of the invention wherein the sensor 102 is a turbine flow transducer is shown in Figure 3. The turbine transducer comprises two fixed swirl plates (1) in a tube (2) which divert the flow of the exhaled air going through the tube into a vortex. A flat vane (3) affixed to a pivot (4) is held between two bearings (5) fixed to the centres of the swirl plates such that the flat vane (3) is free to rotate. The rate of rotation is proportional to the flow through the turbine and the number of rotations is proportional to the volume.

The turbine is constructed from a clear plastic and an infra-red emitter (6) produces a beam of light that passes through the turbine. This beam is interrupted by the rotating vane (3) resulting in pulses of light exiting the turbine. An infra-red sensor (7) receives the pulses.

The output of the infra-red sensor is transmitted via a transmitter 103 to a processor 104 that converts the number of pulses during an exhalation or inhalation to volume and the rate of pulses to flow. The processor generates a flow-volume loop based upon the flow and volume of the exhalation and inhalation.