Sensors incorporating light guides have been proposed for use in biomedical and other applications, where information may be obtained optically from a remote position. In general, light from a source is transmitted by a light guide to a position from where information is required, and light modulated in some way to carry the information is transmitted back along the same or another light guide to a photo detector.

One kind of sensor is used as an optical dip-stick. It consists of a glass or plastics rod which is straight and has a bevelled or chiselled end. Light transmitted down one side of the rod is reflected twice at the bevelled end and returns up the other side of the rod when the tip is dry. When the tip is wet and the angle of incidence is less than the critical angle, light is lost and the returning optical signal is greatly reduced. A variant of that level sensor is called the ϋ-rod level sensor. It consists of a glass rod which is shaped into a *ϋ*, light being sent from one end to a detector at the other end. The rod acts as a light pipe when dry, but light leaves it when it is wet.

With both of these devices there is difficulty in detecting variation in fluid level, since neither device can discriminate accurately where along the length of the sensor wetting has occurred. To overcome some of these difficulties, an optical fibre, configured in a particular way, may be used.

An optical fibre consists of a light transmitting core and an outer cladding which, although -possibly translucent, ensures total internal reflection at the core/cladding interface. Optical fibres have been used to transmit light to the sensor

and to carry the signal back to a detector. In recently reported work, light from a suitable source travels along an optical fibre to its end, where reflection or scattering of the light returns it along the same or another fibre to a light measuring instrument, which interprets the returning light signal.

In accordance with the present invention, an optical sensor comprises a optic fibre in a continuous loop which extends, in use, from a light source to a photo detector, the optical fibre comprising an inner light-transmitting core and an outer cladding, which normally ensures total internal reflection of light transmitted along the optical fibre core from the source to the detector, a portion of the optical fibre being configured to allow light to escape from the light transmitting core at a sensing region in dependence on the optical properties of a surrounding medium at the sensing region.

The new sensor differs from the previous detectors described above in that it consists of a continuous loop of optical fibre, from light source to detector. An important feature is the configuration of the optical fibre at the sensing region. Most simply, the configuring of the optical fibre involves forming a, notch in, or otherwise stripping away, a portion of the cladding at the sensing region to expose an area of the optical fibre core. The exposed portion of the core will then be in contact with the surrounding medium and light loss at the sensing region will depend upon the relative refractive indices of the core and surrounding medium, or on the colour or other optical property of the medium. For example, if the sensor is wetted at the sensing region, much greater light is lost than when the sensing region is dry and the detector will

receive an optical signal of different magnitude. he presence, or characteristic, e.g. colour, of an appropriate medium may thus be determined at a precise position where the sensing region is situated. Sensitivity is increased if the optical fibre has a bend at, or immediately upstream of the sensing region, in the direction of light transmission along the optical fibre. This may be the result of a decrease in the angle of incidence of the internally reflected light on the surface of the light transmitting core at the sensing region, together with a shortening of the axial distance between successive internal reflections.

Furthermore, when a bend, curved to a radius less than the manufacturer's specified critical radius, is provided in a flexible optical fibre, the cladding fails to ensure total internal reflection of light transmitted along the core. Such a bend is hereafter referred to as a sharp bend. Again, this ma be the result of a decrease in the angle of incidence of the internally reflected light at the interface of the core and cladding, together with a shortening of the axial distance between successive internal reflections. The inventors have found that the effect is enhanced if there is a double sharp bend, preferably in more than one plane, at or adjacent to the sensing region. This leads to a different way of configuring the optical fibre, provided that the cladding is translucent. Thus the optical fibre will Be provided with a sharp bend or bends to form the sensing region and the light which then escapes from the light transmitting core into the cladding will either remain in the optical fibre and be transmitted to the detector, or be lost through the outer surface of the cladding, in dependence upon the relative refractive indices of the cladding and surrounding medium, or other optical

property of the medium. The effect is therefore substantially the same as when a portion of the cladding is stripped away to expose an area of the optical fibre core to the surrounding medium, at the sensing region. Such a sensor is particularly sensitive when the outer convex side of the sharp bend is exposed to the surrounding medium.

Fibre-optic guides, consisting of a light transmitting core of plastics material, such as poly ethyl methacr late, encased in a plastics cladding, of for example, a fluorinated polymer are available. Such so-called "polymer" optical fibres are flexible and this makes them suitable for many applications, in association with probes, catheters and other supporting apparatus. However, they are particularly useful for sensors in accordance with the invention, since if a polymer optical fibre is bent around a "sharp" bend of small radius, much less than that recommended by the manufacturers for normal use, the resulting configuration can influence the integrity of the cladding as described above.

In some applications it is desirable for the sensing region to be on the inner concave side of the optical fibre where the fibre passes around a bend. Even if the cladding is stripped away on this side of the optical fibre, sensitivity is not great, owing probably to the small number of internal reflections incident on the concave side of the bend. Furthermore, if the bend is "sharp" it may breach the integrity of the cladding on the outer convex side of the bend, thus reducing sensitivity on the inner concave side. . We find that this problem can be overcome by stripping away a portion of the cladding on the concave side of the bend and providing the cladding at the convex side of the bend with an inwardly facing rough reflecting surface. This has the effect of avoiding loss of light from the convex

side of the bend, and increasing the number of reflections incident on the concave side of the bend. The rough reflecting surface is preferably provided by roughening the outer surface of the cladding on the convex side of the bend, and covering the roughening with a reflective paint or other coating.

The sensors have a wide variety of applications but in most cases, the detector will be a photoelectric transducer, such as a photodiode, which may be connected into a suitable electronic circuit to produce an electrical signal corresponding to the intensity of the optical signal detected. The electrical signal may be fed to a display, or to recording apparatus, or as a control signal to a pump, valve or motor associated with the liquid or medium being sensed, possibly via a discrimination circuit which responds to whether the signal level is above or below a certain threshold: A typical use for the sensor in a non medical field is as a level sensor, bubble detector, colour sensor or interface detector in mixtures of gas, liquid and/or solid. For example the optical fibre of the sensor may be embedded in a wall of a liquid container, with the sensing region exposed at the inner surface of the container wall. The output signal from the sensor then may be used to control equipment, for example, for supplying liquid to the container. Two very important uses for these sensors in the medical field are (1) in association with a liquid sucker, such as a catheter for removal of blood or other body fluids in vascular and cardiac surgery, or a dentist's mouth piece for removing saliva from a patient's mouth, and (2) as level sensors in reservoirs for blood or liquids for infusion. Conventionally, such suckers operate

continuously and if they are to have sufficient capacity to remove all liquid as it accumulates, it is inevitable that they will continually aspirate air and liquid. This is irritatingly noisy and, if the liquid is blood, the resulting shear stresses in the blood are a major cause of blood trauma and the formation of gas icrobubbles and fat globules. These problems can be overcome if the liquid sucker has a tubular body provided with a sensor in accordance with the present invention with the sensing region positioned adjacent to the suction tip of the body. The signal received by the detector may then be used automatically to switch on a pump or other source of suction to the sucker when the active area is wetted by the liquid, and to switch off the source of suction when the liquid level has dropped below the sensing region.

By appropriately positioning the suction tube relatively to a body cavity, a constant level of fluid may be maintained, for example in the pericardium during topical hypothermic myocardial protection.

By using light of a particular wavelength, information about the state of the liquid adjacent to the sensing region may be obtained. For example, if the light transmitted down the optical fibre is red, conveniently supplied by a red light emitting diode, and the liquid is blood, the loss of light from the sensing region, and hence the optical signal received by the detector, varies with the colour of the blood. The variation in the colour is correlated with the oxygen saturation of the blood. Such a sensor, which may be mounted in a blood conduit or other probe, may provide a valuable clinical guide to the performance of an artificial lung during open heart surgery.

A further application of the sensor is for

sensing the pH of blood or other liquid, where the sensor is coated with immobilised material which changes colour with the pH of the medium with which it is in contact. The change in colour of the immobilised material will produce a corresponding change in the level of the detected signal.

Some examples of sensors constructed in accordance with the present invention are illustrated diagrammatically in the accompanying drawings in which;

Figure 1 is a partially sectioned front view of a first liquid sucker fitted with one sensor;

Figure 2 is a side view corresponding to Figure

1; Figure 3 is a section taken on the line III-III in Figure 2;

Figure 4 is a detail of a optical fibre used in a sensor;

Figure 5 is an axial section through part of a second liquid sucker provided with a second type of sensor;

Figure 6 is a detail of the second type of sensor;

Figure 7 is a section taken on the line VII-VII in Figure 6; and.

Figure 8 is a block diagram of a photoelectric circuit associated with the sensors.

The sucker shown in Figures 1 to 3 comprises a tubular body consisting of a probe 10 and a suction tip 11, which are bonded together at a spigot and socket connection. The tip 11 terminates in a suction opening 12. In use the other end of the probe 10 is connected to a pump or other source of vacuum. The probe 10 has in its wall a duct 13 for accommodating legs 14 and 15 of a continuous loop of optical fibre the duct 13 opening into a slot 16 in one of two bulbous portions 17 and 18 of the tip 11.

Between the portions 17 and 18, the external surface of the tip 11 is provided with a groove 19, in which a multiply bent portion 23 of the optical fibre seats, with about half of the cross-section of the fibre proud of the groove. The portions 17 and 18 protect the sensing region from contact with moist surfaces when the tip 11 is inserted into, or moved around within, a body cavity.

The optical fibre consists of a light transmitting core 20 surrounded by a cladding 21. Where the conductor passes around the groove 19, the cladding 21 is stripped away at a portion 22 to expose the core 20, and define a sensing region, as shown in Figure 4. Stripping away a portion of the cladding 21, may not be necessary if the cladding 21 is translucent and the fibre is sharply bent in such a way that light can escape from the fibre core. The segment of the fibre which is conformed in this way constitutes the sensing region. This may happen when the conductor is of the polymer type and the optical fibre is bent round a much smaller radius than that recommended for normal light transmission. For example, in the case of an optical fibre having a polymethyl methacrylate core 20 and a cladding 21 of a fluorinated polymer and of 1 mm outside diameter, the fibre should not normally be bent around a radius of less than 15 mm, if it is to retain its proper light transmitting properties. In the case of a sucker of the kind illustrated in Figures 1 to 3, the optical fibre is typically bent around a radius of 2 mm, resulting in light escaping from the core 20 and into the cladding 21 and from the cladding when wet.

An alternative construction is shown in Figures 5 to 7. As before the sucker has a tubular body comprising a probe 10A and tip 11A which are fitted together by a spigot and socket connection. Legs 14A

and 15A of a continuous loop optical fibre are accommodated in a duct in the wall of the probe 10A, the legs merging through bends 23A into a central bend 23B which extends around within a groove 19A in the inner wall of the tubular body, at the junction of the probe and tip. As shown in Figures 6 and 7, the outer surface of the cladding 21A at the outer convex side of the central bend 23B in the optical fibre has been roughened and the roughness coated with a silver paint 24. This ensures that no light will escape from the core 20A and out through the cladding 21A at the outer convex portion of the central sharp bend. upon assembly, the probe 10A and tip 11A are fitted together with an interposed adhesive and with the central bend 23B of the optical fibre in the groove 19A. Excess bonding agent is provided in the groove 19A securely to fix and pot the fibre within the groove. At this time the inner concave side of the optical fibre projects from the groove slightly into the passageway through the tubular body. A reaming tool, substantially of the size of the passageway through the tubular body is then passed through the tubular body to abrade the bent portion of the fibre thereby stripping off the cladding 21A and exposing the core 20A at a sensing region at the inner concave portion 25 of the fibre.

In both examples, the legs 14A and 15A terminate in connectors 26 which, in use, are coupled via connectors 27 to a light emitting diode 28 and a photodiode 29 respectively. The electrical output from the photodiode 29 is passed to a conditioning and discriminating circuit 30, an output of which controls a vacuum pump 31, which provides the suction through the tubular body. When the sensing region is in contact with air, there is substantially total internal reflection as the light from the LED 28

passes around the optical fibre to the detector 29, the output of which is therefore at a maximum level. When the sensing region is contacted by blood or other liquid, there is significant loss of light into the surrounding medium whereby the output of the detector 29 is significantly reduced. The circuit 30 discriminates between the signal levels such that when the level is below a threshold the pump 31 is started and when it is above the threshold the pump is stopped.

Spurious readings could be produced if the optical fibre is contacted at the sensing region by material, such as tissue, fibrin or droplets of blood held by surface tension. A conformation of the sensing region shown in Figures 1 and 2 makes tissue contact by the whole sensing region less likely and permits the detection of different fluid levels owing to the change of signal as the fluid level passes the sensing region when the sucker is used with its axis substantially vertical.

It would be possible for the second example, illustrated in the Figures 5 to 7, to be modified so that the bent portion of the optical fibre extends diametrically across the centre of the passageway through the tubular body, rather than around the inner periphery of the passage. However, this would be less satisfactory for some medical applications, owing to the possibility of fouling of the optical fibre with tissue or other matter at the sensing region.

It would be apparent that the sensor could be mounted in an analagous manner in the walls of liquid containers, in order to sense the advance or retraction of a liquid/air interface past the sensing region.