GB1564169A - Scan conversion - Google Patents

Description

(54) IMPROVEMENTS IN OR RELATING TO SCAN CONVERSION (71) We, E M I LIMITED, a British company of Blyth Road, Hayes, Middlesex, do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to performed, to be particularly described in and by the following statement: The present invention relates to scan conversion arrangements, in particular such arrangements for converting the display scan known as plan position indicator (PPI) or sector scan, from for example an ultrasonic investigative system, to a raster scan such as that known for television.

It is known to store PPI data as rows of data ina suitable store, for which the word 'row' does not necessarily imply a physical distribution. In order to produce a point of a raster scan, it is then necessary to read data for the appropriate area from the store, and then to interpolate between those data.

However, this requires that twodimensional interpolation be applied at the raster data rate which, for television standard, is typically ten samples per microsecond. If, instead of interpolation, the nearest point is used, there is still a high computation rate, since an algorithm is required to find the nearest point.

Alternatively, the interpolation can be applied at the data store input. so that the data is stored in the form of the raster scan.

This, however, still leads to problems in interpolating a PPI scan, of which the number of lines may be variable.

It is an object of this invention to provide an alternative scan conversion arrangement.

According to the invention there is provided a scan conversion arrangement for converting data from radially disposed lines of a sector scan to data for parallel lines of a raster scan, including means for sampling the data for each radial line at a rate proportional to the cosine of the angle 0 of inclination of the respective line to a perpendicular of the raster lines, storage means for storing the data and means for reading the data from the storage means in sequences, each sequence corresponding to a line of the raster scan for subsequent display.

In order that the invention may be clearly understood and readily carried into effect, an example thereof will now be described with reference to the accompanying drawings, of which: Figure 1 illustrates a representative PPI scan Figure 2 illustrates a mode of organisation of the data of that scan, Figure 3 shows a block diagram for one example of the invention and Figures 4 and 5 show modifications of the arrangement shown in Figure 3.

Turning now to Figure 1, there is shown a PPI scan centred on an origin 0, only one half of the sector being shown in full.

Typically, a complete PPI scan would cover a 90" sector and would comprise several hundred lines. Each line in the sector is considered to have the same length OA = OB. The projection OB' of line OB on line OA, which is vertical in the drawing, has length OB Cosy.

It is desired to provide raster lines which are, in space. perpendicular to OA. This invention reads the data into the store along the PPI lines as desired, but samples them at intervals which allow the data to be read out directly as raster lines. To achieve this, the input data are sampled at a rate which gives constant spacing of sample points in the direction of line OA. This is shown in Figure 2. Such spacing is provided if the sample frequency along each line is proportional to Cos. or sine (90-0) where 0 is the inclination of that line to AO (Figure 1). If the input data are already in the form of samples, then the resampling required involves a one-dimensional interpolation.

This can be performed by a suitable low pass filter before the resampling. or else by digital interpolation.

The sampled data are read into rows of the data store, with no geometry being implied, and are, in one embodiment of the invention, read out in lines of data, such as that indicated in Figure 2, which have the same projection on OA with no interpolation required.

Clearly, closer to the apex of the PPI scan the data read-out rate is higher. This affects the line readout rate in two ways:- first, lines closer to the apex must be read out more quickly than more distant lines; second, the rate along each line varies as the angle 0 increases. The second variation is as much as 2:1 over a 90" sector scan. The two are in fact two aspects of the same variation and the rate changes for the two aspects can be effected together or separately. Since many suitable data stores have limited ranges of reading rates, it is convenient to read the data at a constant rate into a line buffer.

such as a digital (shift register) or analogue (charge coupled) device, from which it can be derived at the desired variable rate. If, close to the apex at O. the data rate becomes too high. some of the points can be omitted to reduce the clock frequencies.

This can be achieved by reading only every nth point.

If the number of lines and the line spacing of the PPI signal are variable. the output clock rate is controlled according to the scan geometry.

A block diagram of a suitable system, for this aspect of the invention. is shown in Figure 3. in which the data flow is shown in full line and clock and sync signals. from a frame store and timing generator 1 and input signal timing generator 2. are shown in broken line.

The PPI signal from a source 3 is read into line buffers 4 at the rate proportional to CosO as discussed hereinbefore. The source 3 may typically be an ultrasonic sector scanner such as a rotating probe system or a phased arrav of transducer elements. The data are read into a data store 5 at a suitable clock rate and then transferred to a TV display 6 via further line buffers 7. which ensure the correct rate required for TV. It should be noted that in this example, since the two timing generators are independent, the input signal and video output are probably asynchronous. Furthermore, the data store can conveniently have separate periods for reading and writing.

It will be realised that the different TV lines are filled to different proportions bv the PPI data. Lines closer to the apex have only data for a progressively shorter central part. It is, therefore, important to provide zeroes to line buffer 7, before and after the PPI derived data, to ensure that each line is complete.

The example discussed hereinbefore does not require output interpolation but does require a data reading rate inversely proportional to the distance from the apex. This results from packing approximately the same number of samples into a variable proportion of the line.

If, however, interpolation is employed, between the read out from store 5 and the T.V. display, the data points can be evenly spaced by a suitable choice of sampling rate.

By doing this a variable rate read-out from store 5 may be avoided. It is, of course, replaced by a variable sampling rate but that variation will be directly proportional to the distance from the apex, since lines closer to the apex will have less samples. In certain circumstances the substitution of a direct for an indirect variation can be an advantage.

Figures 4 and 5 show two variations of the Figure 3 circuit to implement this embodiment including interpolation. The control paths from timing generator 1 have been omitted but it will be understood that control is effected thereby as in Figure 3.

The circuit of Figure 4 includes two analogue line stores 8 and 9 to which the data from store 5 are applied via a digital to analogue (D/A) converter 10 and low pass filter 11. The data are derived from store 5 at a constant rate f2 for all lines and then converted into analogue form. The characteristics of filter 11 are chosen to provide a suitable interpolation function between the samples at rate f2. A resampling is now effected by sampling at a variable rate f3.

The resampling comprises clocking the data into one of the analogue line stores in response to a variable sample clock at f3.

The rate f3 incorporates both aspects of the rate variation discussed hereinbefore, being now a directly proportional relation as mentiond. The required number of leading and trailing zeroes are also included at this stage and the data read out for display at T.V. rate f4.

If the line stores can have simultaneous inputs and outputs or different rates then one line store could be used. However in this example two are used so that as a line is stored in one store the preceding line is read from the other for display. In an alternative use of the same Figure 4 circuit, the line readout clock can be caused to vary as Cos-0 along each line. This rate change is then the same for each line but does not include the grosser variation from one line to the next which is still effected bv variations in f3. Thus f3 is constant for each line.

but varies between lines, and may be more easily derived from a system clock. In such an embodiment the low pass filter 11 should be tracked in cut-off frequency with f2.

However a suitable compromise fixed cutoff can be used.

Figure 5 shows a digital equivalent, of Figure 4, in which the stores 8 and 9 become digital line stores 8D and 9D respectively.

The DIA converter is now required at the output of the line stores and is replaced, together with filter 11, at the input by a suitable digital interpolator 12. Interpolator 12 runs at a clock rate, f2', high enough to be greater than the highest f2 expected. In this example f2' is an integral multiple of f2, being n f2 where n is an integer.

This system also may be operated so that f2 is varied as Cos 20. In that case, since f2' = n f2, the interpolation function automatically tracks the variation in f2 across each line.

In a further variation, not illustrated, the digital interpolator of Figure 5 may be followed by a D/A converter and analogue line stores as in Figure 4. In this case if f2 varies as Cos20, the interpolation output is still constant rate and a fixed cut-off filter may be used without disadvantage. However the filter will perform some interpolation so that f2' can be a lower frequency. For some examples f2' could be 2f2.

It will be apparent that other variations of this embodiment of the invention may be devised. For example the signal in the analogue variants could redigitised to use digital line stores. The choice of implementation will depend on relevant design constraints. The circuit elements may be of any known type. For example the line stores can be charge coupled devices or commutated storage capacitors.

WHAT WE CLAIM IS: 1. A scan conversion arrangement for converting data from radially disposed lines of a sector scan to data for parallel lines of a raster scan, including means for sampling the data for each radial line at a rate proportional to the cosine of the angle t) of inclination of the respective line to a perpendicular of the raster lines, storage means for storing the data and means for reading the data from the storage means in sequences, each sequence corresponding to a line of the raster scan for subsequent display.

2. An arrangement according to Claim 1 in which the means for reading is arranged to read the data at a rate inversely proportional to the distance of the respective sampled data from the apex of the sector scan.

3. An arrangement according to Claim 1 in which the means for reading is arranged to read the data at a constant rate and including means for interpolating between the data to provide further data substantially equally spaced on the raster line.

4. An arrangement according to Claim 3 in which the means for interpolating is arranged to resample the data at a rate directly proportional to the distance of the respective sampled data from the apex of the sector scan.

5. An arrangement according to Claim 4 in which the rate directly proportional to said distance is inversely proportional to cosine20.

6. An arrangement according to Claim 1 in which the means for reading is arranged to read the data at a rate inversely proportional to cosine20 and including means for interpolating between the data to provide further data substantially equally spaced on the raster line.

7. An arrangement according to any of Claims 3 to 6 in which the means for interpolating includes a digital interpolator.

8. An arrangement according to any of Claims 3 - 7 including means for converting the first mentioned data, or data derived therefrom. to analogue form and in which the means for interpolating includes an analogue interpolator.

9. An arrangement according to Claim 8 in which the analogue interpolator includes a low pass filter.

10. An arrangement according to any preceding claim including means for converting the first mentioned or further data to a television line rate for television standard display.

11. An arrangement according to any preceding claim in which the means for sampling at a rate proportional to cosine (3 includes a digital interpolator.

12. A scan conversion arrangement substantially as herein described with reference to Figures 1 - 3 of the accompanying drawings.

13. A scan conversion arrangement substantially as herein described with reference to Figures 1, 2, 4 and 5 of the accompanying drawings.

14. A display system, for displaying data derived from a sector scan ultrasonic transducer arrangement, including a scan conversion arrangement according to any preced

Claims (1)

1. ing claim.

   15. An ultrasonic examining system including a sector scanning transducer arrangement and a display system according to Claim 14.