The present invention relates to a method and apparatus for capturing an image of a luminescent sample, particularly, though not exclusively, a chemiluminescent sample.

Imaging of luminescent samples is common in the assaying of biomolecules.

Traditional assay imaging apparatus and methods use a light sensor and lens arrangement that projects an image of the whole sample onto the light sensor. This requires the sensor to be placed relatively far from the imaged sample in a large light tight darkroom enclosure. For these reasons traditional imaging systems are often bulky and expensive.

US4884890 discloses imaging apparatus for imaging florescent samples. In a first embodiment the imaging sensor comprises a slit 38 and photomultiplier tube 40 that is moved across the image plane. In a second embodiment the imaging apparatus comprises a static CCD sensor located on the image plane and sized to record the whole image. In both embodiments a further detector 56 is used to measure the overall brightness of the whole image and the output of the detector 56 used to normalise the output of the imaging sensor.

US2014/0206568 describes an alternative design of scanner that operates akin to a flatbed scanner. It includes an array of light sensors that are passed over the luminous sample to produce the image. This avoids the need for expensive optics and because the light sensor array can be positioned very close to the sample, the overall size and cost of the apparatus can be reduced.

Notwithstanding, the sensor array still has to move over the sample relatively slowly meaning that the brightness of the sample may change over the time required to pass the sensor array over the whole sample to form an image. As a result the scanned image may not accurately reproduce the relative brightness of different parts of the sample. Accurate representation of the relative brightness of different parts of a sample is important for certain analysis techniques.

To address this problem, the scanner of US2014/0206568 is arranged to sweep the sensor across the whole sample in a first direction to obtain first image data and then sweep the sensor across the whole sample in an opposite direction to obtain second image data. An average brightness is derived for each part of the image using the two sets of image data. This process relies on the rate of change of luminescence of the sample being the same during both the first and second sweeps. In practice this is often not the case.

The present invention was conceived to provide an improved scanner for imaging luminescent samples.

According to a first aspect of the invention there is provided a method of imaging of a luminescent sample; the method comprising: i) passing an electronic image sensor and/or luminescent sample across the other; ii) constructing image data of the luminescent sample using readings from the electronic image sensor taken during i); iii) measuring the luminance of a site of the luminescent sample multiple times to determine a change in luminance of the luminescent sample during i); and using the determined change in luminance to normalise the image data to compensate for the change.

According to a second aspect of the invention there is provided apparatus for imaging a luminescent sample; the apparatus comprising: at least one electronic image sensor; driving means for passing a first of the at least one electronic image sensor and/or the luminescent sample across the other in order to take readings from the first electronic image sensor of the luminescent sample; a controller to control the driving means in order to control the relative position of the first electronic image sensor relative the luminescent sample; processing means arranged to construct image data of the sample from the readings taken from the electronic image sensor; and wherein the at least one electronic image sensor is arranged to take readings of a site of the luminescent sample at different times, and the processing means arranged to use the readings of the site to determine a relative change in luminance of sample and to use the determined change in relative luminance to normalise the image data to compensate for the change.

By determining a change in luminance, as opposed to simply deriving an average luminance of the site, it is possible to compensate where a rate of change of luminance of the sample changes during imaging.

The following may apply to either aspect of the invention.

The luminance measurements can be used (e.g. by the processing means) to determine a profile of the change in relative luminance of the sample over the time taken to collect all the readings needed to complete the image data. Using the profile, pixel values of the constructed image data can be normalised based on the times at which the reading(s) used to determine the pixel values were taken.

The invention is thought to be of particular benefit for imaging chemiluminescent samples where changes in strength of luminescence often occur quickly relative to the time needed to scan the sample. Nevertheless, the method may also be applicable for imaging samples that luminesce through other processes, e.g. photoluminescence and fluorescence. In the case of fluorescence, the invention may also be used to compensate for variations in the luminance as a result of variation in the intensity of the light source used to illuminate the sample and/or a variation in the efficiency or yield of the fluorescence due to bleaching.

For ease of constructing the imaging apparatus the electronic image sensor (or electronic image sensors in embodiments in which more than one is used) is favourably passed across (e.g. over) the sample in order to acquire the readings used to construct the image. For simplicity this will be the relative movement described in the remainder of the description. Nevertheless, it should be borne in mind that where this relative movement is mentioned or implied, other relative motions, e.g. movement of sample relative to the sensor, or of both could be employed instead.

The luminance of the sample may be measured during i), i.e. before all light readings needed to complete the image data have been taken, and preferably before the electronic image sensor has completed a full pass across the sample; additionally measurements may also be taken before and after i). Alternatively, measurements may be taken instead only before and after i) though this less preferred as it will not provide as accurate a profile of the change of luminance of the sample.

The change in luminance of the sample during i) may be determined, e.g. by the processing means, by measuring the luminance of a single site of the sample multiple times, e.g. at and/or near the start of the i) at and/or near the end of i) and one or more times there between.

Alternatively the luminance of each of the several different sites may be measured multiple times and the change in luminance at each of the several different sites used to determine relative changes in luminance of the sample in order to normalise the image data. Favourably the multiple readings from each of the several sites are taken during i).

The electronic image sensor is favourably used to take a reading of the site in order to measure the luminance of the site. In order that the electronic image sensor can take light readings to construct image data of an un-imaged portion of the sample between taking luminance measurements, the method favourably includes moving the electronic image sensor relative to the sample between taking luminance measurements.

The electronic image sensor may be passed relative to the sample in a first (forward) direction to capture light readings for constructing image data of the sample (optionally including capturing a reading at the site in order to measure the luminance of the site a first time), and (optionally at intervals during said pass) the electronic image sensor is moved relative to the sample in an opposite (backwards) direction to the site in order to measure the luminance of the site a further time. In one arrangement the electronic image sensor may be moved in the opposite direction to measure the luminance of the same site each time.

The electronic image sensor may move in the first direction to a first transition position in order to capture light readings for constructing image data of a first portion of the sample; move back in the opposite direction from the first transition position to the first site to take a further luminance measurement at the first site and then move forward towards and beyond the first transition position to a second transition position in order to capture light readings for constructing image data of a second different portion of the sample. The electronic image sensor may then move in the opposite direction a second time in order to image the first site a further time before moving in the first direction again towards and beyond the second transition position to capture light readings for constructing image data of a further portion of the sample.

The electronic image sensor may move in the first direction at a first speed (scanning speed) in order to take readings of the sample to construct image data. When moving in the opposition direction between the transition position and the site to take a further reading, the electronic image sensor may move at a speed that is faster than the first speed. Similarly, when moving from the site back in the first direction to the transition position the electronic image sensor may move at a speed faster than the first speed. Once reaching the transition position the electronic image sensor may continue to move in the first direction at the first speed so that readings can be taken to continue constructing image data of the sample. This reduces the time taken to take multiple measurements at the site when using only a single electronic image sensor.

Rather than repeatedly moving back to the same site to take a luminance measurement, the electronic image sensor may be moved in the second direction to measure the luminance of a different site of the sample a further time, e.g. a site that is closer to the transition position of the electronic image sensor. This reduces the distance that the electronic image sensor needs to travel to obtain a further luminance measure. In a variant, the electronic sensor may be moved in the first direction over substantially the whole sample at scanning speed to take readings to construct the image data and take a first luminosity measurement at one or more sites of the sample, and then move back over substantially the whole sample at a relatively fast speed compared with the scanning speed to take a second measurement at the one or more sites.

In a variant, the method may comprise passing two electronic image sensors across the sample, the electronic image sensors spaced apart in a direction of travel of the sensors; each of the sensors passing over the site of the sample, and each measuring the luminance of the site in order to determine a relative change in luminance of the site. The two electronic image sensors may be arranged to move synchronously over the sample.

This arrangement has the advantage of avoiding backtracking of the electronic image sensor and thus reduces the time required to capture an image of the whole sample. The drawback of this arrangement is a cost disadvantage of using an additional electronic image sensor.

The image data may comprise pixel values and the method may include combining pixel values of contiguous pixels of the image data, wherein the pixel values are derived from light readings taken from the same light sensor element of the electronic image sensor at two or more different positions of the electronic image sensor on its pass across the sample. This technique may allow a part of the sample to be used in an assay even if it exhibits luminance below the noise floor.

Before the electronic image sensor has completed its pass across the sample, the site of the sample may be selected from within one or more portions of the sample from which readings have been taken during the pass. The site may be selected based on readings that indicate luminance above a threshold. The threshold is favourably chosen to be sufficiently above the noise floor to ensure that a satisfactory signal/noise ratio is achieved. The luminance readings provide an indication of the relative luminance of the sample at the times the readings were taken. From this it is possible to determine an indication of the relative luminance of the sample at other times during imaging of the sample through extrapolation and/or interpolation. As such the method may include extrapolating and/or interpolating to determine an indicator of a relative luminance of the sample at a time t _x during i) that does not coincide with (e.g. is before, after or between) a luminance measurement taken at one of the sites. The determined indicator of relative luminance may be applied to normalise the image data derived from readings taken at and/or near time t _x . The process of extrapolation and/or interpolation may be linear, e.g. linear extrapolation; alternatively polynomial or other forms of extrapolation and interpolation may be used.

In order to do this, time information is associated with each pixel value, the time information being indicative of the time during the imaging process that the reading(s) were taken from which the pixel value is derived.

The electronic image sensor favourably comprises an array of light sensors. For example the electronic image sensor may comprise a linear 1D sensor array (i.e. a line of light sensor elements of a single pixel width). Alternatively a TDI (time delay and integration) area sensor could be used.

The sensor may comprise a CCD or CMOS image sensor.

The image sensor may be sensitive for imaging one or more of visible, infra-red and ultraviolet emissions from a luminescent sample.

The processing means may comprise one or more suitable programmed processors.

The invention will now be described by example with reference to the following Figures in which:

Figure 1 is a schematic of the components of imaging apparatus for imaging a luminescent sample; Figure 2 is a schematic plan view of the bed and detector of the imaging apparatus of Fig 1 illustrating movement of the detector relative the bed according to a first variant mode in order to allow for normalisation of image data;

Figure 3 is a chart plotting relative luminance against time for an example luminescent sample;

Figure 4 is a schematic plan view of the bed and detector illustrating movement of the detector relative the bed according to a second variant mode; and

Figure 5 is a schematic plan view of a variant imaging apparatus comprising two detectors;

Fig 1 shows a schematic of apparatus for imaging a luminescent sample such as a chemiluminescent sample. The apparatus may be used for assaying to determine, for example, one or more of the presence, amount or functional activity of an analyte such as a biomolecule. The sample may take the form, for example, of a gel or blot assay.

The apparatus comprises a bed 1 for supporting the luminescent sample 2, a moveable detector 3 comprising an array of light sensing elements 3A. An example detector 3 may comprise a CMOS array sensitive to one or more of visible, infra-red and ultra violet light depending on the frequency of emission expected from the sample.

The apparatus also includes a drive system to move the detector 3 across (typically over) the sample 2 whilst light readings are taken from the light sensing elements 3A, and a process and control system 4 arranged to construct and normalise image data of the sample 2 using the readings outputted from the detector 3.

The specific form of the drive system is not important so long as it provides for accurate control and positioning of the detector 3 relative the sample 2. A suitable drive system may comprise a stepper motor 5 in combination with a linear actuator 6, such as a lead screw or rack and pinion mechanism that operates under control of a controller 4A provided as a function of the process and control system 4. The process and control system 4 can be implemented using conventional computer hardware, e.g. one or more processors communicatively coupled with memory, suitably programmed using techniques known to those skilled in the art.

The apparatus includes a cover or housing 7 that, at least during scanning of the sample, conceals the detector 3 from light derived from sources other than the luminescent sample 2.

In use, the luminescent sample 2 is supported on the bed 1 and the detector 3 caused to pass in close proximity across the sample 2, taking light readings from the light sensing elements 3A as it does so. The relatively long exposure times typically needed to obtain light readings with an acceptable noise to signal ratio mean it can take several minutes to image the whole sample 2.

Using techniques conventional in the field of optical scanners, the light readings from the light sensing elements 3A are processed by the process and control system 4 to construct image data comprising a set of pixel values which can be used to form an image, e.g. on an electronic display or via a printing method.

Unconventionally, time information is associated with each set of pixel values derived from light readings from the array of light sensing elements 3A during a single exposure. The time information is indicative of the time elapsed from the beginning of the scan at which the light reading(s) used to derive the pixel value were taken by the detector 3.

The process and control system 4 also associates position information with each set of pixel values. The position information indicates the position of the detector 3 relative to the sample 2 at the time the readings used to form the set of pixel values were taken. The position information is used by the controller 4A to accurately position the detector 3 relative to the sample 2 in order to take multiple luminance measurements at a single site as described below. The pixel values and associated time and position information are stored in a memory store of the process and control system 4.

The following describes example methods by which the process and control system 4 including controller 4A controls the function and movement of the detector 3 across the sample 2, and how the process and control system 4 manipulates the resulting image data to compensate for changes in the relative luminance of the sample 2 as the detector 3 passes across the sample 2.

In a first variant mode, the detector 3 is caused to pass across a first portion of the sample 2 in a forward direction Xi (solid line), left-to-right as viewed in Fig 2 at a speed (hereinafter scanning speed) that allows for light readings to be taken with the desired exposure time. Whilst the detector 3 passes over the sample 2, light readings from the first portion of the sample 2 are analysed by the process and control system 4 to determine a site Sl within the first portion that has sufficient luminance to be suitable for measuring a change of luminance of the sample 2. The position of the identified site Sl is recorded by the process and control system 4. A site might be selected by virtue of having an absolute luminance sufficient to induce a signal strength from a group of contiguous light sensor elements 3A above a threshold level (but does not saturate any sensor element 3A). Where multiple suitable sites are identified, the site corresponding with the strongest luminance may be selected.

After a period of time has passed since the light reading at S 1 was taken (the period of time can be arbitrary though long enough to allow more of the sample to be scanned between repeat measurements), the detector’s 3 position Rl over the sample, hereafter referred to as the first transition position, is recorded and the detector 3 caused to move backwards (right to left) Yi (dashed line) across the sample 2 to site S 1. A further light reading is taken of site S 1 in order to provide a second luminance reading of site Sl. Light readings are not taken (or if taken are discarded) whilst the detector 3 is moving backwards meaning that the detector 3 can travel back to S 1 at a greater speed (hereinafter referred to as a traversing speed) than the forward scanning speed. Following taking of second luminance reading, the detector 3 moves Zi in the first direction, at traversing speed, to the first transition position Rl, from which point the detector 3 is caused to resume its pass X ₂ over the sample 2 in the forward direction at scanning speed to take further readings of a yet-unscanned second portion of the sample 2.

After a period of time following taking of the second luminance reading at Sl has passed, the detector’s new position R2 over the sample 2 (hereinafter referred to as the second transition position) is recorded and the detector 3 moves backward Y ₂ across the sample (right to left) at traversing speed to take a third luminance reading at site Sl before returning Z ₂ to the second transition position R2 and thence continuing forward at scanning speed to obtain light readings of a further portion of the sample 2. The detector 3 may make further reciprocal backward and forward movements in order to take additional luminance readings at S 1 as desired until the detector 3 has passed across the whole sample to take all light readings required to complete the image data of the sample.

To ensure comparable reading at Sl are obtained, it is preferable that the same exposure time and resolution is used when taking each reading. Nevertheless, in variant methods, different exposure times could be used so long as an adjustment is made to compensate.

Figure 3 illustrates, in chart form, the relative luminance of a fictitious sample during the time taken to scan it in its entirety. Each data point is derived from a luminance reading (in this example five) taken at site Sl at a different times t during the scan. The e first reading is assigned a relative luminance value of unity; subsequent relative luminance values are determined by dividing the measured luminance value of a new reading by the measured luminance value of the first reading.

In order to derive a relative luminance for a time during the scan between luminance readings at site Sl, interpolation may be carried out using known methods, e.g. linear interpolation or through modelling using part of all of the luminance readings. Similarly, extrapolation methods can be used to provide an indication of relative luminance at times occurring before the first reading at Sl and/or after the last reading of S 1.

To normalise the resultant image data, a normalisation factor is applied to each pixel value of the image data based on the relative luminance of the sample 2 at the time during imaging that the light reading(s) corresponding to that pixel value was taken. For example, if the relative luminance RF of the sample at a time ti from the start of the scan was determined, e.g. through interpolation, to be 1.4, i.e. the sample was 1.4 times more luminous at time ti than a reference luminance, a factor of 1/1.4 will be applied to all pixel values taken substantially at time ti. Similarly, if at another time, t2 (not shown), the relative luminance of the sample was 0.5, then a factor of 1/0.5 will be applied to pixel values taken at t2.

It will be appreciated that instead of multiplying by the reciprocal of relative luminance, each pixel value could be divided by the relative luminance. Alternatively some other function dependent on the relative luminance of the sample at time t _x could be applied to pixel values derived from readings taken at time t _x .

In this way pixel values of the image data can be adjusted to compensate for changes in the relative brightness of the sample 2 during the period of time taken to image the whole sample 2.

Figure 4 illustrates a variant mode of scanning a luminescent sample in which the changes in the relative luminance of the sample 2 are determined by taking multiple luminance readings at each of several different sites of the sample.

The detector 3 is caused to pass across a portion of the sample 2 in a forward direction X’i, left-to-right as before. During the pass light readings from the sample 2 are analysed by the process and control system 4 to determine a site Sl suitable to be used for measuring change of luminance. The position of the site Si is recorded by the process and control system 4. After a period of time has passed since the detector 3 took the luminance reading at first site Sl the detector’s position Rl (first transition position) is recorded and the detector 3 is moved backwards Y’i across the sample 2 (right to left at traversing speed) to the position of site Sl. A second luminance reading is taken of site Sl following which the detector 3 is moved forwards Z’i, at traversing speed, to the first transition position Rl, and thence caused to continue its forward pass X’ ₂ over an as yet unscanned portion of the sample 2 taking light readings to add to the image data.

After a period of time following a first reading at a second suitable site S2 identified during movement X’ ₂ beyond the first transition position Rl, and which may have a different absolute luminance to Sl, the current position R2 (herein after the second transition position) of the detector 3 is recorded and the detector 3 moved backwards Y’ ₂ (right to left at traverse speed) to take a reading for a second luminance measurement at site S2. The detector 3 is then moved forward Z’ _I at a traversing speed to the second transition position R2 and thence continues X’ ₃ in the forward direction at scanning speed to image a further portion of the sample. This is repeated in order to obtain multiple luminance readings at a further site S3 and again for any number of further sites depending on, for example, the desired speed of the scan, sample size and the desired accuracy of normalisation.

In any one of movements X’^ X’ ₂ X’ ₃ etc, multiple sites spaced apart along the axis of movement of the detector 3 may be identified The detector 3 may then return to the first of the identified sites to take a second luminance measurement and later to the second identified site to take a second luminance measurement there. In this way the travel distance during a reciprocal movement of the detector 3 can be reduced even if no suitable further sites are identified in a region of the sample scanned between taking a pair of luminance measurements.

Each pair of readings taken from a site Sl .. S _x provides an indication of the relative change in luminance of the sample during the time interval between the pair. These derived relative changes in luminance can be used to provide an indication of the relative change in luminance of the sample over the scan time.

So that the relative intensity measurements taken at Sl can be used with relative intensity measurements taken at site S2, it is preferable that the first reading at site S2 is taken before the second (or last) reading at site (e.g. Sl). Similarly it is preferred that the first reading at the third site S ₃ is taken before the second readings at the second site S2 and so on.

Through interpolation, the first and second readings (e.g. at times tl and t2 respectively) at site Sl can be used to identify the relative luminance of the sample when the first reading at site S2 at time t3 is taken (t3 occurring between tl and t2). The ratio between luminance levels from the first reading taken at time t3 at the second site S2 and a second reading at time t4 at the second site S2 (t4 occurring after t2) provides a change in relative luminance of the sample 2 between t3 and t4. The value for the change in relative luminance can be multiplied by the relative luminance of the sample at t3 in order to provide the relative luminance of the sample at time t4.

It is possible to combine readings from sites where all the readings from one site are taken before (or after) the readings for the other site by extrapolating from the readings from one site to determine the relative luminance of the sample at the time one of the readings from the other site was taken. For example, where readings at S2 taken at times t3 and t4 both occur after the second reading at S 1 taken at time t2, an extrapolation process using determined relative luminances at times tl and t2 can be used to determine the relative luminance of the sample at t3. The ratio between the luminance readings taken at t3 and t4 provide the relative change in luminance of the sample between t3 and t4 which can be multiplied by the relative luminance determined at t3 in order to arrive at the relative luminance of the sample at time t4.

The sampling interval, i.e. the period of time between which luminance readings from a site (e.g. Sl) are taken (or where multiple sites, the period of time between readings at the same site), may be fixed before the scan, based e.g. on the size of the sample and the exposure time.

In a variant scanning method, the detector 3 may be moved in the first direction over substantially the whole sample 2 at scanning speed to take readings to construct the image data, and identify and take a first luminosity measurement at one or more sites of the sample, and then move back over substantially the whole sample at a relatively fast speed compared with the scanning speed to take a second measurement at the one or more sites.

Alternatively, the sampling interval may be varied during the scan based on the rate of change of relative luminance of the sample determined during the scan. For example, if after taking two luminance readings at site Sl it is determined that the rate of change of relative luminance of the sample 2 is above a threshold, the process and control system 4 may reduce the sampling interval before the next luminance reading taken at Sl (or of a different site depending on mode). If further luminance readings indicate a further increase of rate of change in luminance, the sampling interval for the next measurement can be reduced further; conversely if the rate of change has substantially reduced the sampling period may be increased.

This enables the apparatus to respond to changing rates of change in luminance in order to increase the accuracy of normalisation when rates of change are high, whilst reducing the number of repeat measurements taken when rates of change of luminance are low, in order to accelerate the scanning process.

Rates of change in luminance of the sample determined whilst the detector 3 is passing over the sample 2 can also be used by the process and controller 4 to alter one or more of the exposure time of the light readings, relative speed of movement between detector 3 and sample 2, and resolution of image derived from the image data. For example, where determined that the sample’s 2 absolute luminance and rate of decrease of luminance are such that the sample’s luminance may drop below a recordable level before the scan is expected to complete, the process and control system 4 may instruct the detector 3 to reduce the exposure time and/or reduce the resolution of light readings, e.g. through a conventional binning process of signals from contiguous light sensor elements 3A of the array, and increase the forward scanning speed of the detector 3 in order to accelerate the scanning rate to capture light readings of the whole sample 2 before the luminance of the sample drops to a level in which the noise will be unacceptable.

Figure 5 illustrates variant apparatus that differs primarily in the presence of a first detector 3 A and a second detector 3B. In order that each detector 3 A 3B can move independently of the other they are driven by separate motors 5A 5B (and optionally separate actuators 6) under control of control 4A. The first detector is arranged to pass across substantially the whole of the sample 2 in the forward direction taking light readings to construct the image data.

The light readings from the first detector 3A are analysed by the process and control system 4 to determine one or more sites Sl of the sample 2 spaced apart along the axis of movement of the detector 3 across the sample 2.

The second detector 3B is arranged to move to the position of the first site Sl determined from the light readings from the first detector 3A, and take light readings to record the change in luminance at the first site Sl whilst the first detector 3 A continues to scan the sample 3. This arrangement allows for more frequent (optionally substantially continuous) luminance readings (and hence relative luminance measurements) to be taken without the need to slow the scanning process through back-tracking of the first detector 3A.

In a first variant the first and second detectors 3A 3B are arranged to each pass over the sample 2 spaced by a fixed distance apart along the axis of movement across the sample. In this way two readings of the whole sample will be taken with a fixed time interval again enabling substantially continuous luminance readings (and hence relative luminance measurements) to be taken without the need to slow the scanning process through back-tracking of the first detector 3A. In this variant both detectors can be driven by a single motor.

In a further variant, the two detectors 3A 3B are arranged to move across different portions (e.g. different halves) of the sample 2, each adapted to function in the manner described in relation to Figs 2 or 4 in order to image their respective portions.

Typically each sensor element will exhibit a black signal level that is different from at least some of the other sensor elements (providing what is known as fixed pattern noise) and which may change independently from the other sensor elements. During a scan, the black signal level of the sensor elements may change as a result, for example of a change in the sensor’s temperature.

To compensate for fixed pattern noise, the or each detector 3 may be caused by the process and control 4 to take a light reading at a black position; a black position being a position where no light from the sample 2 or an external source is expected to be received by the detector 3. This allows for the determination of a black signal level for each sensor element 3A of the array which can be used to correct for variation in readings taken from different light sensor elements attributable to different black signal levels. To account for a change in fixed pattern noise during the course of imaging, the or both detectors 3 may be caused by the process and control system 4 to move either during or after the scan to a black position in order to update the black signal levels. The updated black signal levels can then be used in the construction of image data derived from readings taken thereafter.

The black position may, for example, be a stowed position that the detector 3 moves to when not in use, e.g. before and after a scan. Updating the black level during scanning may be of particular use where the detector comprises a CMOS image sensor as such sensors have a higher level of fixed pattern noise.

For samples with a relatively low absolute luminance and/or where a relatively fast scan rate is required, the processor and control means 4 may be arranged to combine pixel values of groups of contiguous pixels of the image (analogous to a binning operation) in order to form a macropixel having a value equal to the total or average value of the combined pixels. This provides an image with improved signal to noise ratio at the expense of resolution.

A macro pixel may be comprised of two or more pixel values derived from light sensor readings from the same sensor element following movement of the detector relative to the sample during the scan. This allows the formation of macro pixels from pixels spaced in the direction of movement of the detector 3 across the sample, even where the array of light sensors 3A is only one element wide. Macro pixels may instead or in addition be comprised using light readings from contiguous light elements 3 A of the detector’s array.

In an alternative though less preferred variant, the apparatus may be arranged such that the sample 2 is moved relative to the detector 3 in which case the bed 1 may be dispensed with in favour, for example, of a sample moving mechanism such as, for example a drum or other feeder.

Although preferred, the apparatus need not adopt a traversing speed when moving to and from a site to take a second or further reading.