JP2006515675A - Method and apparatus for recovering dead pixels in a digital imaging system - Google Patents

Abstract

Gamma radiation events are individually received at the elements of the detector array (18), and at least one of the elements (P0) is defective. Each detector element converts incident radiation into a radiation event signal, which is digitized by an analog-to-digital converter (42) into the coordinate position (x, y) and energy (z) of the detector array. Is done. The event generator (48) generates a radiation event signal for each defective element based on radiation events received at contributing elements, such as the nearest neighboring elements (P1-P8). In the preferred embodiment, the contribution from each contributing element is randomized by passing a token (56) between the locations in the table (54) corresponding to each contributing element. Each time a radiation event is received at a contributing element for which a corresponding table position holds a token, the event also generates an event signal for the defective element and the token is passed (58). The energy of the generated event for the defective detector element is randomized (62), for example by replacing the least significant bit with a random number.

Description

  The present invention relates to digital imaging technology. The present invention finds particular application in the nuclear camera and will be described with particular reference thereto. It should be understood that the present invention is also applicable to other pixilated imaging devices, such as, for example, an astronomical detector for imaging blurred areas of outer space. Those skilled in the art will understand the applicability of the present invention in applications where the presence of dead or regressive pixels compromises the accuracy of the technique.

  Diagnostic nuclear medicine imaging is used to examine a subject's radionuclide distribution. Generally, one or more radiopharmaceuticals or radioisotopes are injected into a subject. Radiopharmaceuticals are generally injected into a subject's bloodstream to image the circulatory system or specific organs that absorb the injected radiopharmaceutical. A radiation detector is placed near the surface of the subject to monitor and record the emitted radiation. In many cases, the detector is rotated or indexed around the subject to monitor radiation emitted from multiple directions. These projection data sets are reconstructed into a three-dimensional image representing the radiopharmaceutical distribution within the subject.

  In general, each detector head has an array of photomultiplier tubes (PMTs) facing a large scintillation crystal. Each emission event produces a corresponding flash that is observed by the nearest photomultiplier tube. Each photomultiplier tube that observes the event outputs a corresponding analog pulse. Analog pulses from individual PMTs are digitized and combined to generate x and y space coordinates of the scintillation event location on the crystal plane.

  However, there are several problems with multipliers. Doublers tend to drift, take up a lot of space and are expensive. Many proposals have been made to replace multipliers with pixelated solid state arrays. Engineers and scientists have encountered additional problems as they began using solid state devices to examine a subject's nuclide distribution. A solid state array typically includes about 15,000 pixels or more of 1-2 mm. It is difficult to produce an array of this size without including any dead pixels or nonstandard pixels. In addition, over time, one or more pixels may become dead pixels or degrade, or otherwise fail to provide correct data in response to radiation events. There are many. Pixels that do not produce correct data will cause artifacts in the reconstructed image.

  The present invention contemplates a new and improved method and apparatus that solves the above and other problems.

  In accordance with one aspect of the present invention, a radiation detection device is provided. The array of elements converts the individual received radiation events into corresponding radiation event signals. One of the radiation conversion elements has a defect. A means digitizes radiation events. A means generates a radiation event signal for the defective radiation conversion element based on radiation event signals from other radiation conversion elements in the array.

  According to another aspect of the present invention, a method for detecting radiation is provided. A radiation event is received at the array of pixelated locations and a corresponding radiation event signal is generated. At least one of the positions has a defect. Radiation event signals from defect-free locations are digitized. A radiation event signal for the defective location is generated based on the radiation event signal from the location without the defect.

  Further advantages and benefits of the present invention will become apparent to those skilled in the art upon reading and understanding the following detailed description of the preferred embodiments.

  The invention can take form in various components and arrangements of components, and in various steps and arrangements of steps. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention.

  With reference to FIG. 1, a nuclear imaging apparatus generally includes a stationary gantry 10 that supports a rotating gantry 12. One or more detector heads 14 are carried by the rotating gantry to detect radiation events emanating from the region of interest 16. Each detector head has a two-dimensional array 18 of detector elements. The detector array is preferably a solid state detector that converts gamma radiation directly into charge. However, other arrays are also contemplated, such as an array of scintillators that are optically coupled to an array of photodiodes. Each head has a circuit 20 for converting each radiation response into a digital signal indicating its position (x, y) and its energy (z) in the detector plane.

  In general, an object 22 to be imaged is injected with one or more radiopharmaceuticals or radioisotopes and placed in the examination region 16. The presence of these medications within the subject 22 results in the emission of radiation from the subject. Radiation traveling along the trajectory defined by the collimator 24 is detected by the detector head 14. The detector head is angularly indexed or rotated around the examination area to collect emission data from multiple directions. The projection emission data (x, y, z) and angular position (θ) of the detector head around the examination area are stored in the data storage 26. The reconstruction processor 28 processes the event and detector orientation data from the data storage 26 into a volumetric image representation. The image representation is stored in the volume image memory 30 for manipulation by the video processor 32 and display on an image display 34 such as a video monitor, printer or the like.

  Referring to FIG. 2, circuit 20 optionally includes an array 40 of signal processing circuits for amplifying or otherwise processing the output of each detector element of array 18. An array of analog-to-digital converters 42 converts each event into a digital signal indicating the (x, y) position of the detector plane and digitizes the pulse amplitude to provide an energy signal (z).

  During the pre-calibration process, the detector surface is illuminated by a uniform flood field. Due to a uniform flood of radiation, all detectors in the array should have the same count and the event must be of consistent amplitude. The control circuit or processor 44 monitors the output of each detector element, either directly or by reading the memory 26, to see if each of the detector elements has approximately the same count number and approximately the same energy distribution. To do. If any of the detectors is greater than the preselected deviation and different from the others, the control processor 44 causes the switching means 46 to delete signals from elements that do not work correctly, or the preamplifier 40 An element that does not work correctly (for example, pixel P0) is disconnected from the analog-to-digital converter 42. The control processor 44 causes the output of a plurality of nearest neighbor pixels or other contributing pixels (eg, pixels P1-P8) to be sent to both the memory 26 and the event generation circuit 48. The output of the event generation circuit 48 is connected to the data memory 26 to provide (x, y, z) radiation events for elements that do not work correctly according to the events received by the contributing pixels.

  Referring to FIG. 3, the event generator 48 of the preferred embodiment has an input 50 in which events from each contributing pixel, such as the nearest neighbor pixels (P1-P8), for example. Is received. Each time an event is received at one of the contribution pixel inputs, the look-up and comparison circuit 52 looks at the corresponding pixel table 54 and the contribution pixel where the event occurred corresponds to the table location having the token 56. Decide if you have. Illustratively, if the token 56 is at the table location P1 corresponding to the nearest neighboring pixel P1, and an input event is received from the pixel P1, the lookup and comparison circuit 52 outputs an output indicating the event received at the location P0. Is generated. The backup and comparison circuit 52 can further cause the token passing control circuit 58 to cause the table position P1 to pass (pass) the token to one of the table positions P2-P8 corresponding to the other contributing pixels. . The tokens may be passed in sequence between table positions, but are preferably passed randomly. Thus, if there are 8 contributing pixels, approximately 1/8 of the events that occur in each of the contributing pixels are applied not only to the contributing pixel location, but also to the location of the defective pixel.

  In one embodiment, the energy of the contributing pixel event is passed to the data memory as the energy of the event at P0. However, it is preferred that the energy circuit 60 replace the actual energy of the shared event with the expected average energy of the injected radioisotope. More specifically, the energy of the event is generally distributed on a bell curve. Dither circuit 62 preferably vibrates the energy along the bell curve to produce a more general energy distribution. In one embodiment, the dither circuit 62 deletes the least significant bits of the energy value from the contributing pixels, eg, the three least significant bits (70). The dither circuit 60 has a random number generator 72 for generating a random value for the detected least significant bit, which replaces the deleted bit (74).

  Referring to FIG. 4, it should be understood that the contributing pixels need not be the eight nearest neighbor pixels. In the example of FIG. 4, the contributing pixels include the 24 nearest neighboring pixels. More preferably, the token passing circuit 58 passes the token based on the proximity of the contributing pixels. The randomness of the token passing algorithm is selected so that the nearest neighbor pixels P1-P8 receive tokens up to four times more than the next nearest neighbor pixels P9-P24. In another alternative embodiment, matrix corner pixels P9, P13, P17 and P21 are not used as contributing pixels or contribute less frequently. In another alternative embodiment, only the positive-sign array pixels, P11, P2, P6, P19 and P22, P8, P4, P15, contribute to the numbering method of FIG. In another alternative embodiment, only pixel columns, rows or diagonal rows contribute. Various other configurations of contributing pixels can be selected as appropriate to select pixels that produce approximately the same amount of output as would be expected from a defective pixel.

  In one example, two defective pixels P0 may appear side by side as shown in FIG. 5, or may appear from corner to corner. In the example of FIG. 5, the nearest neighboring pixels P1-P10 that enclose the pair are selected as contributing pixels. Data is processed in much the same way as in FIG. 3 except that there are two tokens. One token corresponds to each of the two defective pixels. Whenever an input is received from a contributing pixel corresponding to a table location that holds a token for one of the pixels, an event is generated for that pixel and the token is passed.

  The invention has been described with reference to the preferred embodiments. Variations and changes will be made by reading and understanding the foregoing detailed description. The present invention is intended to be construed as including all such modifications and variations as fall within the scope of the appended claims or their equivalents.

1 is a schematic view of a nuclear digital imaging apparatus according to the present invention. Schematic of a radiation event signal generation circuit. FIG. 4 shows a technique for generating a radiation event signal for a defective pixel. FIG. 6 illustrates an alternative embodiment of a technique for generating random events. FIG. 4 illustrates a technique for generating a radiation event signal for a defective pixel that is in line.

Claims (20)

An array of radiation conversion elements for converting individual received radiation events into corresponding radiation event signals, wherein one of said radiation conversion elements is defective;
Means for digitizing radiation event signals from at least defect-free radiation conversion elements;
Means for generating a radiation event signal for the defective radiation conversion element based on radiation event signals from other radiation conversion elements of the array;
A radiation detection apparatus. Each of the radiation conversion elements is
A solid-state detector element and a pair of scintillation crystal and photodiode,
The apparatus of claim 1, having one of the following:   Means for assigning at least two of the individual radiation conversion elements of the array as contributing pixels, the output signal of the contributing pixels being the means for generating the output signal of the defective pixel; The apparatus of claim 1, wherein the apparatus is supplied. The means for generating an output signal of the defective pixel;
A table with a position for each of the contributing pixels;
Means for passing a token between the table positions;
In response to receiving a radiation event signal from the contributing radiation conversion element and accessing the table to determine whether the corresponding table position holds the token, wherein the corresponding table position holds the token Means for generating a radiation event signal for the defective radiation conversion element and passing the token to the means for passing the token;
The apparatus of claim 3, comprising:   Two adjacent radiation conversion elements are defective, and the apparatus further comprises two tokens, one token corresponding to each of the defective radiation event conversion elements and passing the token The apparatus of claim 4, wherein the means passes the token between the table positions.   4. The means for generating an output signal for the defective pixel generates an output signal for the defective radiation event transducer as a random part of an event in the contributing radiation event transducer. The device described.   The apparatus according to claim 1, further comprising energy output means for assigning a radiation energy value to the generated radiation event signal for the defective pixel.   8. The apparatus of claim 7, further comprising energy changing means for changing an energy output of the energy output means in a preselected limited range. The energy changing means is
Means for removing a preselected number of least significant bits of the energy value;
A random number generator that randomly generates the least significant bit;
Means for replacing the removed least significant bits with the randomly generated least significant bits;
9. The apparatus of claim 8, comprising: Means for reconstructing radiation event information into an image representation;
Means for storing the image representation;
Means for converting at least a portion of the image representation into a human-readable display;
The apparatus of claim 1, further comprising: A two-dimensional array of radiation detector elements that receive incident gamma radiation events and generate corresponding output signals, wherein one of the radiation detector elements is defective;
At least one analog-to-digital converter that converts the output signal of the element into a digital value indicative of the spatial position of the array and the energy of the incident gamma radiation event;
A virtual event generator that generates a digital output signal for the defective radiation detection element based on output signals from other contributing radiation detection elements of the array;
With gamma camera. Receiving a radiation event at an array of pixelated locations and generating a corresponding radiation event signal, wherein at least one of the locations is defective;
Digitizing radiation event signals from a defect-free location;
Generating a radiation event signal for the defective detection position based on the radiation event signal from the defect-free position;
A method for detecting radiation comprising: Illuminating the pixelated position with a flood gamma radiation field;
Monitoring at least one of the radiation event signals to determine the location of the defect;
The method of claim 12, further comprising:   The method further comprises assigning a position adjacent to each of the defective locations as a contributing pixel location, wherein the output of the contributing pixel location forms the basis of a radiation event signal for the defective location. 14. The method according to 13. Providing a token to at least one of the contributing pixel locations;
In response to receiving a radiation event signal corresponding to the contributing pixel location with the token, generating a radiation event signal for the defective pixel location and passing the token to another contributing pixel location When,
15. The method of claim 14, comprising:   The two adjacent pixel locations are defective and the step of providing the token further comprises providing two tokens, one token corresponding to each of the defective pixel locations, The method of claim 15, wherein the method is passed separately.   The contributing pixels are the nearest neighbor pixel location and the next nearest neighbor pixel location, and the step of passing the token is more frequently than passing a token between the next nearest neighbor pixel locations; The method of claim 15, comprising passing the token between the nearest neighboring pixel locations.   13. The radiation event signal of claim 12, wherein the radiation event signal indicates a location and energy of a received radiation event, and the method further comprises randomly changing a digital energy value corresponding to the defective pixel location. Method. Removing the least significant bit of the digital energy value of the radiation event at the contributing pixel location;
Assigning a randomly generated value as the least significant bit;
The method of claim 18, further comprising: The digitized event signal includes an array position value indicating the position of the array where the radiation event was received;
Reconstructing the digital position value into a three-dimensional image representation;
Converting a portion of the image representation into a human-readable display;
The method of claim 12, further comprising:

Images