GB2395594A - Bioassay reading system using a computer to locate and identify microlabels by identifying spatially sequential groups or identification codes - Google Patents

Abstract

There is provided a bioassay reading system (10) for interrogating microlabels (50). The system (10) comprises: <SL> <LI>(a) a platform (40) for receiving the microlabels (50); <LI>(b) a filament illuminator (90) and a episonic mercury illuminator (80) for illuminating the microlabels (50); <LI>(c) a microscope (20) and a camera (60) for generating at least one image of the illuminated microlabels (50) and converting said at least one image into corresponding image signals; <LI>(d) a motorized stage (30) for spatially actuating the platform (40) and its microlabels (50) relative to the microscope (20) for selecting preferred areas of the platform (40); and <LI>(e) a personal computer (70) for receiving the image signals and for processing the signals to determine identities of microlabels (50) present in said at least one image and their associated optical responses. </SL> The system (10) is distinguished in that the personal computer (70) is operable to locate and identify the microlabels (50) by identifying spatially sequential groups of identification codes present on the microlabels (50) depicted in said at least one image.

Description

BIOASSAY READING SYSTEM

Field of the invention

l'hc present invention relates to bioassay reading systems; in particular, the invention concerns a 5 bioassay reading system for reading bioassay micron-scale encoded substrates. T he substrates are preferably susceptible to fluorescence when interrogated and comprise identification bar-codes; for example, such substrates are known in the art as "Smart Beads". Moreover, the present invention also relates to a method of reading such bioassay micron-scale encoded substrates using the aforementioned bioassay reading system.

Background to the invention

In a published international PCT patent application no. PCTfUS00/27121 (WO 01/26038), there arc described methods of imaging colloidal rod particles, such particles functioning as nanobar 15 codes. The methods involve imaging or reading freestanding particles comprising a plurality of segments. Moreover, the particles are of length from I O rim to 50 1lm and of width from 5 rim to 50 m. The segments of the particles may be comprised of any material including a metal, allow, a metal allow, a metal nitride, a metal chalcogenide, a metal oxide, a metal sulfide, a metal selenide, a metal telluride, polymeric materials, crystalline and/or non-crystalline materials.

Moreover, in a published international PCT patent application no. PCT/GB99/03109 (WO 00/16893), there is described a system for performing parallel bioassays. In the system, microfabricated labels are made to each carry a biochemical test. When the system is in use, many different labels are mixed together with an analyte sample. The system additional 25 comprises a device that reads the individual labels are thereby isolates results of individual tests.

The microfabricated labels have a surface layer of anodized metal and are produced by utilizing anodizing, lithographic patterning and etching steps. Aluminium is a preferred metal for use in fabricating the labels, although other materials can be employed to fabricate the labels.

30 Moreover, in a published international PCT patent application no. PCT/GB02/00628 (WO 02/065123)7 there is disclosed a biochemical method of detecting one or more protein characteristics. The method utilizes supports, wherein each support has a largest dimension of less than 250 1lm and wherein each support incorporates sequential identifying means. Else method is distinguished in that it includes the steps of:

(a) attaching an information molecule to a main surface of a support, the information molecule being capable of interacting with at least one of said one or more protein characteristics to be detected; (b) suspending such supports comprising one or more different sequential identification means 5 and one or more different information molecules in a fluid; (c) adding a biological sample to be analysed to the fluid; (d) detecting interaction signals from supports in the fluid using signal detecting means; and (e) reading the sequential identification means of the supports which have an interaction signal using reading means, 10 thereby detecting at Icast one of said one or more protein characteristics. Furthermore, in the international application, there is also described apparatus susceptible for executing this method.

A similar approach for detecting genetic characteristics is described in a related international patent application no. PCT/GB02/00644 (WO 02/064829).

15 Furthermore, in a published international PCT application no. PCT/US98/11224 (WO 98/53300), there are described an apparatus and a system for simultaneously analyzing a plurality of analyses anchored to microparticles. Microparticles each having a uniform population of a single kind of analyte are disposed as a substantially immobilized planar array inside of a flow chamber where steps of an analytical process are carried out by delivering a sequence of processing reagents to 20 the microparticles by a fluidic system under microprocessor control. In response to such process steps, an optical signal is generated at the surface of each microparticle which is characteristic of the interaction between the analyte carried by the microparticle and the delivered processing reagent. The plurality of analyses are simultaneously analyzed by collecting and recording images of the optical signals generated by all the microparticles in the planar array. A key feature 25 of the invention is the correlation of the sequence of optical signals generated by each microparticle in the planar array during the analytical process.

The inventors have appreciated that, in a first approach, microparticles described in the foregoing can be interrogated in a flow apparatus wherein the microparticles are transported serially in a 30 fluid flow relative to visual inspection equipment. Alternatively, in a second approach, the microparticles can be deposited onto a surface and then interrogated by moving the surface relative to a visual inspection system. In implementing the second approach, the inventors have appreciated that image data processing problems arise which renders such an approach difficult to

implement in practice. In order to address these problems, the inventors have devised an improved bioassay reading system.

5 Summary of the invention

According to a first aspect of the present invention, there is provided a bioassay reading system for interrogating microlabels, the system comprising: (a) a surface for receiving the microlabels; (b) illuminating means for illuminating the microlabels; 10 (c) imaging means for generating at least one image of the illuminated microlabels and converting said at least one image into corresponding image signals; (d) actuating means for spatially actuating the surface and its microlabels relative to the imaging means for selecting preferred areas of the surface; and (e) data processing means for receiving the image signals and for processing the signals to 15 determine identities of microlabels present in said at least one image and their associated optical responses, characterised in that the processing means is operable to locate and identify the microlabels by identifying spatially sequential groups of identification codes present on the microlabels depicted in said at least one image.

The invention is of advantage in that identifying the microlabels by searching for their corresponding spatially sequential groups of identification codes is capable of providing more reliable reading of the microlabels.

25 Preferably, the data processing means is arranged to process the image signals to identify spatial regions therein corresponding to groups of identification codes and hence their corresponding microlabcl positions, and then to process the image signals corresponding to a substantially central portion of each microlabel to determine its corresponding identification code. By utilizing information corresponding to only central portions of microlabels, the system is capable of being 30 rendered more resilient to misidentification of microlabels with damaged edge regions. This increased reliability allows a greater number of distinct codes to be used.

PrcDerably, the processing means is operable to subject the image signals to threshold detection to remove background artef'acts therein. Such threshold detection is capable of resulting in simpler

and less ambiguous processed image data in which to identify microlabel identification codes.

5 Preferably, the illuminating means is arranged to output radiation for use in interrogating the microlabels at a plurality of selectable wavelengths for determining identification codes of the microlabels and their corresponding optical responses. Using a plurality of wavelengths enables the microlabels to be interrogated using radiation of wavelengths where their identification codes are most easily distinguished, and their fluorescence determined using radiation of wavelengths 10 where their fluorescence is most easily measured.

Preferably, therefore, the illuminating means is operable to output radiation capable of exciting microlabel fluorescence for determining the optical responses. More preferably, the illuminating means includes a filament light source for illuminating the microlabels to determine their 15 identification codes, and at least one ultra-violet light source for illuminating the microlabels to determine their optical responses.

Preferably, in the system, the processing means is capable of identifying redundant identification code sequences at elongate ends of the microlabels, said redundant code sequences being 20 included in the microlabels for rendering them less susceptible to being misidentified in the system due to microlabel edge damage.

Preferably, the processing means is operable to compress intensity variations in the image signals for enhancing identification code detection when applying threshold detection to the compressed 25 image signals. Such compression is capable of improving the system's ability to determine microlabel identification codes correctly. More preferably, such compression is applied to favour identification codes corresponding to regions of relatively high light transmission and/or relatively low light reflectivity of each microlabel.

30 Preferably, the processing means is operable to determine the optical responses of the microlabels based on emission substantially from solid regions of the microlabels only. Such an approach renders the fluorescence response measurement less susceptible to modulation in relation to the microlabels' identification codes and to edge effects resulting from infinite image resolution and variation in the geometry of individual microlabels.

- 4

According to a second aspect of the present invention, there is provided a method of interrogating microlabels in a bioassay reading system, the system comprising: (a) a surface for receiving the microlabels; 5 (b) illuminating means for illuminating the microlabels; (c) imaging means for generating at least one image of the illuminated microlabels; (d) actuating means for spatially actuating the surface and its microlabels relative to the imaging means for selecting preferred areas of the surface; and (e) data processing means for receiving the image signals; the method comprising the steps of: (f) illuminating the microlabels using the illuminating means to generate said at least one image of the illuminated microlabels; (g) converting said at least one image into corresponding image signals; 15 (h) processing the signals to determine identities of microlabels present in said at least one image and their associated optical responses, characterized in that the method further comprises in step (h) the step of locating and identifying the microlabels by identifying spatially sequential groups of identification codes present on the microlabels depicted in said at least one image.

Preferably, the method further comprising in step (h) the steps of: (a) processing the image signals in the processing means to identify spatial regions therein corresponding to groups of identification codes and hence their corresponding microlabel positions, and then 25 (b) processing the image signals corresponding to a substantially central portion of each microlabel to determine its corresponding identification code.

Prcicrably, the method further comprising in step (h) the step of subjecting the image signals in the processing means to threshold detection to remove background artefacts therein.

Preferably, the method further comprising in step (h) the step of arranging for the illuminating means to output radiation to interrogating the microlabels at a plurality of selectable wavelengths for determining identification codes of the microlabels and their corresponding optical responses.

- 5

Preferably, the illuminating means is operable to output radiation capable of exciting microlabel fluorescence for determining the optical responses.

Preferably, in the method, the illuminating means includes a filament light source for illuminating 5 the microlabels to determine their identification codes, and a ultra-violet light source for illuminating the microlabels to determine their optical response.

Preferably, in the method, the processing means is capable of identifying redundant identification code sequences at elongate ends of the microlabels, said redundant code sequences being 10 included in the microlabels for rendering them less susceptible to being misidentified in the system due to microlabel edge damage.

Preferably, in the method, the processing means is operable to compress intensity variations in the image signals for enhancing identification code detection when applying threshold detection to 15 the compressed image signals.

Preferably, in the method, compression is applied to favour identification codes corresponding to regions of relatively high light transmission and/or relatively low light reflectivity of each microlabel. Preferably, the processing means is operable to determine the optical responses of the microlabels based on emission substantially from solid regions of the microlabels only.

It will be appreciated that features of the invention can be combined in any combination without 25 departing from the scope of the invention.

Description of the drawings

Embodiments of the invention will now be described, by way of example only, with reference to 30 the following diagrams wherein: Figure I is an illustration of a bioassay reading system according to the invention; - 6

Figure 2 is an illustration of snake-like motion of a stage of the system of Figure 1 relative to a microscope thereof; Figure 3 is an example of a pixel image depicting a microlabel, the image generated from a 5 camera of the system of Figure 1; Figure 4 is an illustration of an inverse version of the image in Figure 3; Figure S is an illustration of threshold detection applied to detect the presence of microlabels 10 in the inverse image of Figure 4; Figure 6 is an illustration of central strip detection to render microlabel barcode detection more reliable; and 15 Figures 7a to 7d depict a flow chart of steps executed to read microlabels using the system as depicted in Figure 1.

Description of embodiments of the invention

20 Referring to Figure 1, an embodiment of a bioassay reading system according to the invention is shown and indicated generally by 10. In overview, the system 10 comprises an optical microscope 20. The microscope 20 is provided with a motorized stage 30 for actuating a platform 40 bearing microlabels 50 relative to the microscope 20 for enabling the microscope 20 to inspect preferred regions of the platform 40. Moreover, the microscope 20 is provided with a 25 monochrome charge- coupled-device (CCD) camera 60 for receiving magnified images of the preferred regions. It will be apparent to the person skilled in the art that a different type of camera could be used for receiving these images. The camera 60 is coupled in communication with a personal computer (PC) 70 which is arranged to control movement of the platform 40 relative to the microscope 20. The PC 70 is also programmed to perform image analysis on 30 signals generated by the camera 60 when inspecting the platform 40 and thereby recognize and collate data pertaining to inspected microlabels 50 present in the preferred regions.

The microlabels 50 preferably each comprise a substantially elongate substrate bearing identification features and one or more coatings of active substances susceptible to interacting - 7

with biological samples and materials. The coatings are preferably susceptible to exhibiting fluorescent behaviour, either directly and/or after suitable processing using other substances.

In order to stimulate fluorescence, the microscope 20 incorporates at least one episcopic super S high pressure mercury illuminator 80. Moreover, for inspecting identification features of the microlabels, at least one diascopic filament bulb illuminator 90 is also included within the microscope 20. Alternatively, this identification could also be achieved by making use of an additional episcopic illuminator of suitable wavelength. The filament bulb illuminator 90 may of course alternatively be another light source such as light-emitting diode array or other visible 10 light source.

The system 10 is susceptible to being operated in either a manual mode or an automatic mode.

When functioning in the manual mode, a user 100 inputs instructions into the PC 70.

15 The manual mode of operation will now be described in overview.

In the manual mode, the system 10 prompts the user 100 to actuate the stage 30 until sample microlabels 50 are positioned within a field of view imaged by the microscope 20 onto the

camera 60. The sample microlabels 50 are then illuminated in silhouette by energizing the 20 aforesaid diascopic filament illuminator 90; a corresponding magnified silhouette image is thereby imaged through the microscope 20 to the camera 60. The frequencies used to obtain the identification image of the microlabels 50 should be chosen to have minimal overlap with the excitation spectrum of any fluorophores used to detect reaction on a microlabel 50. This has the advantage of maximizing the contrast of the identification image aiding decoding, reducing the 25 rate of photobleaching of the sample, and increasing the accuracy of fluorescence quantification.

The camera 60 generates a digital silhouette image signal corresponding to the magnified silhouette image, the silhouette signal being transmitted to the PC 70. The PC 70 receives the silhouette signal and presents it as an image on its display screen 110 to the user 100. The PC 70 prompts the user 100 to make some adjustments, if required, to the microscope 20, for example a 30 focus adjustment and/or an illumination adjustment. Thereafter, the PC 70 and/or the user make changes to the optical path as necessary to subject the microlabels 50 to ultraviolet radiation by the high pressure mercury illuminator 80 and to detect the resulting image by means of the camera 60. The microlabels 50 are thereby excited into fluorescence resulting in the projection of a corresponding fluorescence magnified image at the camera 60. A corresponding fluorescence

image signal is output from the camera 60 to the PC 70 which stores the fluorescence signal together with its corresponding silhouette signal in memory of the PC 70, for example preferably on its volatile solid state memory. It would be less likely that the storage of the fluorescence image and corresponding silhouette is stored on the hard disk or non- volatile solid state memory.

5 Subsequently, the l'C 70 is able to recall data from its memory corresponding to the signals and perform data processing thereon to determine identities of the sample microlabels and associate these identities with corresponding degrees of fluorescence. If required, when illuminated by the filament illuminator 90, the microlabels 60 can, either alternatively or additionally, be viewed in renection mode.

In the automatic mode, the system 10 scans for the microlabels 50 by actuating the platform 40 in a snake-like manner relative to the microscope 20; such a snake-like motion is depicted in Figure 2. It will be appreciated that other types of motion are also possible, for example a spiral-like motion, which enable the system 10 to inspect the platform 40 in a most efficient manner without 15 unnecessarily presenting the microscope 20 with a same region of the platform 40 multiple times.

However, some overlap in such platform 40 motion is permissible and advantageous to reduce a probability of microlabels 50 being missed from inspection. Setting the overlap of the platform motion to equal or exceed the maximum dimension of any used microlabel would allow all microlabel present to be read. The use of positional coordinates of each microlabel may then be 20 used to verify that each microlabel is only read once. The overlap may of course also be smaller than microlabel size if less accurate reading is performed or image matching is used to verify correct reading.

During such motion, the samples 50 on the platform 40 are illuminated and monitored, via the 25 camera 60 and the microscope 20, at the PC 70. When microlabels 50 enter into the magnified field of view of the camera 60, actuation of the stage 30 is halted. The user 100 may then be

prompted by the PC 70 to make adjustments from initial settings to the microscope 20, for example to adjust focus adjustment and/or illumination. Thereafter, silhouette and fluorescence images of the microlabels 50 in the field of view are generated using illumination from the

30 filament illuminator 90 and the mercury illuminator 80 respectively. The PC 70 controls the alteration of the optical path of the filament illuminator 90 and the mercury illuminator 80 using automated shutters and flanges arranged to break their optical path onto the microlabels when required. When corresponding microlabel 50 silhouette image data and fluorescent image data have been recorded in memory of the PC 70, the PC 70 returns adjustment of microscope 20 to

aforesaid initial settings and then proceeds to actuate the platform 40 in the aforementioned snake-like manner relative to the microscope 20 until further microlabels 50 are identified and so on until the entire platform 40 has been scanned.

5 There are also applications such as the analysis of Single Nucleotide Polymorphisms (SNP) or cellular staining where a filter wheel is used to generate a plurality of illumination wavelengths from the mercury illuminator 80. This filter wheel may be automated and controlled by the PC 70. 10 In the automatic mode, it will further be appreciated that adjustment of the microscope 20 can be made automatic under control of the PC 70. For example, focus adjustment can be achieved by employing an algorithm executing on the PC 70 which attempts to maximize the amplitude of spatial components in the image signal corresponding to fine features in the magnified image received at the camera 60; such automatic focus adjustment corresponds to a maximization of 15 'high frequency spatial image components'. Alternatively, or additionally, automatic illumination adjustment and/or camera 60 iris adjustment can also be performed by the PC 70. Another likely approach of achieving the automatic focusing would be via direct maximization of the high frequency spatial components via the discrete Fourier transform of the image.

20 Moreover, in automatic mode, the user 100 can terminate scanning of the platform 40 before an entire area of the platform 40 has been scanned when a sufficient number of microlabels 50 have been interrogated to generate sufficiently reliable statistical bioassay results. It is also possible for the user 100 to stipulate the termination conditions of the interrogation process when for example sufficient number of microlabels 50 has been interrogated. Alternatively the reader could be set to 25 only interrogate those microlabels 50 which have reacted in e.g. a multiplexed bioassay reaction where reaction is indicated through the fluorescent signal. This would speed up the reading and would be useful if there was no need to read the identification of the non reacted microlabels 50.

The microlabels 50 are preferably of a type as described in the aforementioned international PCT 30 applications nos. PCT/GB02/00628, PCT/GB99/03109 and PCT/GB02/00644 which are hereby incorporated by rcterence. In these patent applications the microlabels are generally referred to as beads, smart beads, particle, microparticles or supports. The microlabels are preferable manufactured from metal, such as aluminum, copper, silver or gold, using microfabrication methods. However, alternative types of microlabel fabricated from various diverse materials can - 10

be used in conjunction with the system 10; such diverse materials include, but are not limited to metals, plastics materials, ceramic-like materials, biological materials and other organic materials or a combination of any such materials. The shape of the microlabels may also be tailored to different applications. When large number of identification codes are needed while the microlabel 5 surface area is kept low the shape is preferably elongated with its identification code e. g. a tramsmissive barcode arranged along its largest dimension, see figure 4 which shows this example. For other planar reading applications the microlabels 50, which are substantially planar may be substantially circular, elliptical, squares, rectangles or any polygon in shape. The microlabel has preferably a largest dimension which is less or equal to ca 2501lm, more 10 preferably less or equal to ca IOOIlm and most preferably less or equal to 50pm. It is important to miniaturize the particles to decrease the use of reagents while still maintaining adequate sensitivity. For applications that require extreme miniaturization the particles may have dimensions in the range of ca O.l- lOIlm. During the interrogation the microlabels 50 may be located through their shape, surface area, and/or from recognizable features of their identification 15 code. It is also possible to have more than one type of microlabels 50 in any one experiment analyzed by the system 10. This would for example mean that a look up table (LUT) with multiple alternatives would be used to identify different microlabels 50.

In contradistinction to conventional bar-code reading apparatus, microlabel 50 bar-code reading is 20 not performed by edge detection processes executed in the PC 70. The inventors have found that the system 20 is susceptible to being rendered more fault tolerant by employing an alternative image processing method; the alternative method comprises a first simple imaging processing step followed by a second averaging step. In the second step, weighted averages of image pixel groups provided from the camera 60 in its output signal are computed, and pixel groups 25 corresponding to individual barcode bits identified.

The aforesaid alternative image processing method will now be elucidated in more detail.

In the first imaging step (STEP A; Figure 7), the microlabels 50 are illuminated in transmission 30 mode, namely to generate a magnified silhouette image thereof at the camera 60. The camera 60 transmits the image as a pixel data set to the PC 70; such a pixel image data set is shown in Figure 3, the pixel image including n x m pixels; in the image, the microlabel 50 appears black in contrast to its illuminated background. The PC 70 then proceeds to invert the image data set

(STEP B; Figure 7) to generate a corresponding inverse image as illustrated in Figure 4.

- 11

The PC 70 then applies threshold detection (STEP D; Figure 7) to eliminate spurious artifacts from the inverse image in a process as depicted in Figure 5. Such artifacts can arise from small specks of biological material or even biological coating material which has become detached 5 from the microlabels 50. In Figure 5, a graph indicate by 200 corresponds to a line of pixels across the image of Figure 4 intersecting the microlabel 50 along its longitudinal axis. 'I'he graph 200 comprises an abscissa axis 210 corresponding to spatial position within the inverse image and an ordinate axis 220. One or more thresholds, for example upper and lower thresholds UT, LT respectively, can be applied to generate a corresponding processed inverse image indicated by 10 250. If required, only a single threshold value can be employed. Moreover, with varying illumination, the threshold can be made spatially variant within the image presented to the camera 60. In the second averaging step, the inverse image as a pixel data set is subjected to pattern 15 recognition filtering (STEP E; Figure 7), namely groups of pixels having approximately a spatial extent corresponding to an expected size for the microlabels 50 are identified. Subsequently for the groups, the PC 70 then proceeds to try to identify barcodcs within the groups and thereby determined whether or notthe groups actually correspond to microlabels. If the PC 70 is unable to identify a barcode for an identified pixel group, the pixel group is ignored in subsequent data 20 processing. In manual mode it is also possible for the user 100 to obtain fluorescence data even for objects which cannot be decoded. This feature give the advantage of allowing accurate calibration of the reader using standard calibration particles prior to commencing interrogation of microlabels 50.

25 In manufacture and subsequent use in bioassays, the microlabels 50 are prone to physical damage.

Being preferably elongate in nature, ends of the microlabels 50 are especially susceptible to fracture. In order to render the microlabels 50 more robust when used in the system 10, the microlabels 50 each include at their elongate ends barcode redundancy. Such redundancy is preferably provided by including binary digits of a value 101 in the barcode ends of each 30 microlabel 50. Analysis software executing on the PC 70 is arranged to assume that beginning and end digits of each barcode are of a value binary value I adjacent to an associated binary 0; the software uses these end digits to partition microlabels 50 in the processed inverse image into bits of mutually equal width (STEP F; Figure 7).

- 12

Once each microlabel 50 in the processed inverse image stored in the PC 70 has been partitioned into its individual digits (STEP G; Figure 7), a central elongate strip of each microlabel 50 image present in the processed inverse image is selected by the software (STEP H; Figure 7) to further eliminate uncertainty arising from lateral edges of each microlabel 50 and from microlabel 50 5 curvature; such a central strip is illustrated in Figure 6.

Once the aforesaid central strip has been identified for each microlabel 50, the software applies averaging (STEP I; Figure 7) over an area corresponding to each binary bit of each barcode of each microlabel 50 within the processed inverse image. Thereafter, averaged values for each bit 10 are compared against a threshold value (STEP J; Figure 7) to determine whether the bits correspond to a binary 0 or a binary I value.

If after averaging each binary bit and subsequent threshold detection a valid barcode is detected by the software, for example by way of parity bit validation, the barcode is assumed by the 15 software to pertain to the microlabel 50. Conversely, if one or more barcodes are found not to be valid, further image signal processing is undertaken. In this further processing, high intensity pixels in the aforesaid central strip have their corresponding intensity values mapped (STEP L; Figure 7) by a mathematical function or look up table (LUT) towards, or to, a particular value in favour of identifying parts of the barcode as a binary 0 value. This further processing is 20 undertaken to improve the decoding rate of microlabels with improperly etched and/or otherwise obscured apertures. Subsequent threshold detection (STEP M; Figure 7) is then performed to determine corresponding barcodes.

After identification of the microlabels 50 and their corresponding barcodes, the software 25 executing on the PC 70 identifies from the processed inverse image those parts of each microlabel 50 which corresponding to solid material, namely not holes therethrough (S l EPs N and O; Iiigure 7). The PC 70 then processes the fluorescent image to determine fluorescence intensity values corresponding to each solid region of each microlabel 50, averages the fluorescence intensity for each microlabel 50 for the solid region (STEP P; Figure 7) and then records decoded barcodes 30 with corresponding averaged fluorescence value (STEP Q; Figure 7).

The PC 70 proceeds to repeat (STEP R; Figure 7) the automatic mode of operation until sufficient microlabels 50 have been interrogated to obtain statistically reliable results.

- 13

Figure 7 provides a schematic illustration of processing steps performed in software executing in the PC 70 to implement the automatic mode of operation.

It will be appreciated that the system 10 can be modified without departing from the scope of the 5 invention. Where for example the fluorescent signal of the microlabels may be replaced of equivalent reporter technology such as radioactivity or enzyme linked systems.

- 14

Claims (22)

1. A bioassay reading system (10) for interrogating microlabels (50), the system (10) comprising: (a) a surface (40) for receiving the microlabels (50); (b) illuminating means (80, 90) for illuminating the microlabels (50); (c) imaging means (20) for generating at least one image of the illuminated microlabels (50) and converting said at least one image into corresponding image signals; (d) actuating means (30) for spatially actuating the surface (40) and its microlabels (50) relative to the imaging means (20) for selecting preferred areas of the surface (40); and (e) data processing means (70) for receiving the image signals and for processing the signals to determine identities of microlabels (50) present in said at least one image and their associated optical responses, characterized in that the processing means (70) is operable to locate and identify the microlabels (50) by identifying spatially sequential groups of identification codes present on the microlabels (50) depicted in said at least one image.

2. A system (10) according to Claim I, wherein the data processing means (70) is arranged to

process the image signals to identify spatial regions therein corresponding to groups of identification codes and hence their corresponding microlabel (50) positions, and then to process the image signals corresponding to a substantially central portion of each microlabel (50) to determine its corresponding identification code.

3. A system (10) according to Claim 1 or 2, wherein the processing means (70) is operable to subject the image signals to threshold detection to remove background artefacts therein.

4. A system (10) according to Claim 1, 2 or 3, wherein the illuminating means (80, 90) is arranged to output radiation for use in interrogating the microlabels (50) at a plurality of selectable wavelengths for determining identification codes of the microlabels (50) and their corresponding optical responses.

5. A system (10) according to Claim 4, wherein the illuminating means (80, 90) is operable to output radiation capable of exciting microlabel fluorescence for determining the optical responses. -

6. A system (10) according to Claim 4 or 5, wherein the illuminating means (80, 90) includes a filament light source (90) for illuminating the microlabels (50) to detemmine their identification codes, and a ultraviolet light source (80) for illuminating the microlabels (50) to determine their optical response.

7. A system (10) according to any one of the preceding claims, wherein the processing means (70) is capable of identifying redundant identification code sequences at elongate ends of the microlabels (50), said redundant code sequences being included in the microlabels (50) for rendering them less susceptible to being misidentified in the system (10) due to microlabel (50) edge damage.

8. A system ( 10) according to any one of the preceding claims, wherein the processing means (70) is operable to compress intensity variations in the image signals for enhancing identification code detection when applying threshold detection to the compressed image signals.

9. A system (10) according to Claim 8, wherein compression is applied to favour identification codes corresponding to regions of relatively high light transmission and/or relatively low light reflectivity of each microlabel (50).

10. A system (10) according to any one of the preceding claims, wherein the processing means (70) is operable to detemlinc the optical responses of the microlabels (50) based on emission substantially from solid regions of the microlabels (50) only.

11. A system (lo) substantially as hereinbefore described with reference to one or more of Figures I to 7.

12. A method of interrogating microlabels (50) in a bioassay reading system (10), the system (10) comprising: (a) a surface (40) for receiving the microlabels (50); (b) illuminating means (80, 90) for illuminating the microlabels (50); (c) imaging means (20) for generating at least one image of the illuminated microlabels (50); (d) actuating means (30) for spatially actuating the surface (40) and its microlabels (50) relative to the imaging means (20) for selecting preferred areas of the surface (40); and - 16

(e) data processing means (70) for receiving the image signals; the method comprising the steps of: (f) illuminating the microlabels (50) using the illuminating means (80, 90) to generate said at least one image of the illuminated microlabels (50); (g) converting said at least one image into corresponding image signals; (h) processing the signals to determine identities of microlabels (50) present in said at least one image and their associated optical responses, characterized in that the method further comprises in step (h) the step of locating and identifying the microlabels (50) by identifying spatially sequential groups of identification codes present on the microlabels (50) depicted in said at least one image.

13. A method according to Claim 12, the method further comprising in step (h) the steps of: (a) processing the image signals in the processing means to identify spatial regions therein corresponding to groups of identification codes and hence their corresponding microlabel (50) positions, and then (b) processing the image signals corresponding to a substantially central portion of each microlabel (50) to determine its corresponding identification code.

14. A method according to Claim 12 or 13, the method further comprising in step (h) the step of subjecting the image signals in the processing means to threshold detection to remove background artefacts therein.

15. A method according to Claim 12, 13 or 14, the method further comprising in step (h) the step of arranging for the illuminating means (80, 90) to output radiation to interrogating the microlabels (50) at a plurality of selectable wavelengths for determining identification codes of the microlabels (50) and their corresponding optical responses.

16. A method according to Claim 15, wherein the illuminating means (80, 90) is operable to output radiation capable of exciting microlabel fluorescence for determining the optical responses.

17. A method according to Claim 16, wherein the illuminating means (80, 90) includes a filament light source (90) for illuminating the microlabels (50) to detemmine their identification - 17

codes, and a ultra-violet light source (80) for illuminating the microlabels (50) to determine their optical response.

18. A method according to any one of Claims 12 to 17, wherein the processing means (70) is capable of identifying redundant identification code sequences at elongate ends of the microlabels (50), said redundant code sequences being included in the microlabels (50) for rendering them less susceptible to being misidentified in the system (10) due to microlabel (50) edge damage.

19. A method according to any one of Claims 12 to 18, wherein the processing means (70) is operable to compress intensity variations in the image signals for enhancing identification code detection when applying threshold detection to the compressed image signals.

20. A method according to Claim 19, wherein compression is applied to favour identification codes corresponding to regions of relatively high light transmission and/or relatively low light reflectivity of each microlabel (50).

21. A method according to any one of Claims 12 to 20, wherein the processing means (70) is operable to determine the optical responses of the microlabels (50) based on emission substantially from solid regions of the microlabels (50) only.

22. A method of interrogating microlabels (50) in a bioassay reading system (10) substantially as hcreinbefore described with reference to one or more of Figures 1 to 7.

- 18