Apparatus and method for solid state storage of episodic signals

• US 4,920,489 A
• Filed: 08/14/1987
• Issued: 04/24/1990
• Est. Priority Date: 08/14/1987
• Status: Expired due to Fees

• Current Assignee: Cruickshank Gary W Trustee
• Original Assignee: Cardiodata, Inc.
• Inventors: Shadmon, David, Hubelbank, Mark
• Primary Examiner(s): Smith, Jerry
• Assistant Examiner(s): Kibby, Steven G.
• Application Number:
  US07/085,879
• Time in Patent Office: 984 Days
• Field of Search: 364/417, 364/416, 364/415, 364/413.03, 364/413.04, 364/413.05, 364/413.06, 128/710, 128/702, 128/709
• US Class Current: 600/519
• CPC Class Codes: G16H 10/60 for patient-specific data, e.g. for electronic patient records
• CPC Class Codes: G16H 40/63 for local operation

Abstract

A method and apparatus for digitally storing episodic signals, such as, electrocardiogram signals is disclosed. In a preferred embodiment, the apparatus includes, for each signal channel, an input amplifier, an analog-to-digital converter, and a double data buffer. The incoming signal on each channel is converted to a multibit digital number a fixed sampling rate. Each multibit digital number is stored in a location in the double data buffer. When a fixed amount of data is stored in the buffer, a processor examines the data to determine the periods during which the signal is undergoing the most rapid rate of change. For these periods, both an indicator of the period, and a first number of data samples per unit time is stored in a bulk solid state memory. Preferably, the data stored is a number relating to the difference between the value of the sample and a number related to the value of the last sample. For other periods, having rates of data change high, but not as high as those of the first period, a smaller number of samples per unit time is stored. These samples are chosen to faithfully reproduce the minima and maxima of the waveform. For all other waveform periods, every Rth sample is stored. The storage media is preferrably dynamic random access memory, the refresh rate of which is determined by the temperature of the circuitry.

Claims

1. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from at least one channel of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) storing a number related to each digital sample from the ones of said predetermined number of groups,

(f) storing a number related to every Rth digital sample from the digital samples outside of said predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

2. The method of claim 1 wherein R is an integer between 2 and 6.

3. The method of claim 1 including the further step of storing a window-start code byte prior to storage of the first number related to each digital sample from each one of said predetermined number of groups.

4. The method of claim 1, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

5. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from at least one channel of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying groups of a predetermined number of consecutive digital samples wherein the sums of the absolute values of said differences exceed a predetermined threshold,

(e) storing each digital sample from said groups,

(f) storing every Rth digital sample from the digital samples outside of said groups, where R is an integer greater than one and is a function of the sampling rate.

6. The method of claim 5 wherein R is an integer between 2 and 6.

7. The method of claim 5 including the further step of storing a window-start code byte immediately prior to storage of the first digital sample from each one of said groups,

8. The method of claim 5, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

9. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from at least one channel of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying groups of a predetermined number of consecutive digital samples wherein the sums of the absolute values of said differences exceed a predetermined threshold,

(e) storing a number related to each digital sample from said groups,

(f) storing a number related to every Rth digital sample from the digital samples outside of said groups, where R is an integer greater than one and is a function of the sampling rate.

10. The method of claim 9 wherein R is an integer between 2 to 6.

11. The method of claim 9 including the further steps of:

storing a window-start code byte immediately prior to storage of the first digital sample from each one of said groups.

12. The method of claim 9, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

13. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from at least one channel of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying groups of a predetermined number of consecutive digital samples wherein the sums of the absolute values of said differences exceed a predetermined threshold,

(e) storing from said groups a digital value related to the difference in amplitude between each digital sample and the digital sample preceding it,

(f) storing a digital value related to the difference in amplitude between every Rth digital sample from the digital samples outside of said groups, where R is an integer greater than one and is a function of the sampling rate.

14. The method of claim 13 wherein R is an integer between 2 and 6.

15. The method of claim 13 including the further steps of:

storing a window-start code byte immediately prior to storage of the first digital sample from each one of said groups.

16. The method of claim 13, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

17. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from at least one channel of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) storing from said predetermined number of groups, a digital value related to the difference in amplitude between each digital sample and the digital sample preceding it,

(f) storing a digital value related to the difference in amplitude between every Rth digital sample from the digital samples outside of said predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

18. The method of claim 17 wherein R is an integer between 2 and 6.

19. The method of claim 17 including the further step of storing a window-start code byte prior to storage of the first digital value related to each digital sample from each one of said predetermined number of groups.

20. The method of claim 17 including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

21. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from at least one channel of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a first predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) storing a number related to each digital sample from said predetermined number of groups,

(f) identifying a second predetermined number of groups of a predetermined number of consecutive samples having sums of the absolute values of said differences which exceed a predetermined threshold,

(g) storing a number related to selected digital samples from said second predetermined number of groups,

(h) storing a number related to every Rth digital sample from the digital samples outside of said first predetermined number of groups and said second predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

22. The method of claim 21 wherein R is an integer between 2 and 6.

23. The method of claim 21 including the further steps of:

storing a window-start code byte immediately prior to storage of the first number related to each digital sample from each one of said first predetermined number of groups, and,

storing a pseudo-window start code byte immediately prior to storage of the first number related to selected digital samples from said second predetermined number of groups.

24. The method of claim 21, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

25. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from at least one channel of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a first predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) storing each digital sample from said first predetermined number of groups,

(f) identifying a second predetermined number of groups of a predetermined number of consecutive samples having sums of the absolute values of said differences which exceed a predetermined threshold,

(g) storing selected digital samples from said second predetermined number of groups,

(h) storing every Rth digital sample from the digital samples outside of said first predetermined number of groups and said second predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

26. The method of claim 25 wherein R is an integer between 2 and 6.

27. The method of claim 25 including the further steps of:

storing a window-start code byte immediately prior to storage of the first digital sample from each one of said first predetermined number of groups, and,

storing a pseudo-window start code byte immediately prior to storage of the first of the selected digital samples from said second predetermined number of groups.

28. The method of claim 27, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

29. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from at least one channel of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a first predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) storing from said first predetermined number of groups, a digital value related to the difference in amplitude between each digital sample and the digital sample preceding it,

(f) identifying a second predetermined number of groups of a predetermined number of consecutive samples having sums of the absolute values of said differences which exceed a predetermined threshold,

(g) storing a digital value relates to the difference in amplitude between selected digital samples from the second predetermined number of groups,

(h) storing a digital value related to the difference in amplitude between every Rth digital sample from the digital samples outside of said first predetermined number of groups and said second predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

30. The method of claim 29 wherein R is an integer between 2 and 6.

31. The method of claim 29 including the further steps of:

storing a window-start code byte immediately prior to storage of the first digital sample from each one of said first predetermined number of groups, and,

storing a pseudo-window start code byte immediately prior to storage of the first of the digital values related to the difference in amplitude between selected digital samples from said second predetermined number of groups.

32. The method of claim 29, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

33. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from two channels of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a selected channel having the fewest artifacts,

(e) identifying from said selected channel a predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(f) storing a number related to each digital sample from the ones of said predetermined number of groups,

(g) storing a number related to every Rth digital sample from the digital samples outside of said predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

34. The method of claim 33 wherein R is an integer between 2 and 6.

35. The method of claim 33 including the further step of storing a window-start code byte prior to storage of the first number related to each digital sample from each one of said first predetermined number of groups.

36. The method of claim 33, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

37. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from two channels of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a selected channel having the fewest artifacts,

(e) identifying from said selected channel a predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(f) storing each digital sample from the ones of said predetermined number of groups,

(g) storing every Rth digital sample from the digital samples outside of said predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

38. The method of claim 37 wherein R is an integer between 2 and 6.

39. The method of claim 37 including the further step of storing a window-start code byte prior to storage of the first digital sample from each one of said first predetermined number of groups.

40. The method of claim 37, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

41. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from two channels of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a selected channel having the fewest artifacts,

(e) identifying from said selected channel a predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(f) storing from said predetermined number of groups, a digital value related to the difference in amplitude between each digital sample and the digital sample preceding it,

(g) storing a digital value related to the difference in amplitude between every Rth digital sample from the digital samples outside of said predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

42. The method of claim 41 wherein R is an integer between 2 and 6.

43. The method of claim 42 including the further step of storing a window-start code byte prior to storage of the first digital value from each one of said first predetermined number of groups.

44. The method of claim 42, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

45. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from two channels of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a selected channel having the fewest artifacts,

(e) identifying from said selected channel a first predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(f) storing a number related to each digital sample from said first predetermined number of groups,

(g) identifying a second predetermined number of groups of a predetermined number of consecutive samples having sums of the absolute values of said differences which exceed a predetermined threshold,

(h) storing a number related to selected digital samples from said second predetermined number of groups,

(i) storing a number related to every Rth digital sample from the digital samples outside of said first predetermined number of groups and said second predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

46. The method of claim 45 wherein R is an integer between 2 and 6.

47. The method of claim 45 including the further steps of:

storing a window-start code byte immediately prior to storage of the first number related to each digital sample from each one of said first predetermined number of groups, and,

storing a pseudo-window start code byte immediately prior to storage of the first of the numbers related to said selected digital samples from said second predetermined number of groups.

48. The method of claim 45, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

49. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from two channels of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a selected channel having the fewest artifacts,

(e) identifying from said selected channel a first predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(f) storing each digital sample from said first predetermined number of groups,

(g) identifying a second predetermined number of groups of a predetermined number of consecutive samples having sums of the absolute values of said differences which exceed a predetermined threshold,

(h) storing selected digital samples from said second predetermined number of groups,

(i) storing every Rth digital sample from the digital samples outside of said first predetermined number of groups and said second predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

50. The method of claim 49 wherein R is an integer between 2 and 6.

51. The method of claim 49 including the further steps of:

storing a window-start code byte immediately prior to storage of the first digital sample from each one of said first predetermined number of groups, and,

storing a pseudo-window start code byte immediately prior to storage of the first of the selected digital samples from said second predetermined number of groups.

52. The method of claim 49, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

53. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate from at least one channel of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a selected channel having the fewest artifacts,

(e) identifying from said selected channel a first predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(f) storing from said first predetermined number of groups a digital value related to the difference in amplitude between each digital sample and the digital sample preceding it,

(g) identifying a second predetermined number of groups of a predetermined number of consecutive samples having sums of the absolute values of said differences which exceed a predetermined threshold,

(h) storing a digital value related to the difference in amplitude between selected digital samples from said second predetermined number of groups,

(i) storing a digital value related to the difference in amplitude between every Rth digital sample from the digital samples outside of said first predetermined number of groups and said second predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

54. The method of claim 53 wherein R is an integer between 2 and 6.

55. The method of claim 53 including the further steps of:

storing a window-start code byte immediately prior to storage of the first digital value from each one of said first predetermined number of groups, and,

storing a pseudo-window start code byte immediately prior to storage of the first of the digital values related to the differences in amplitude between selected digital samples from said second predetermined number of groups.

56. The method of claim 53, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

57. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate of at least 100 samples per second from at least one channel of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) referencing a first table for said predetermined number of groups, said table having a series of ranges of said differences in amplitude between consecutive digital samples, wherein each range corresponds to a mapped value and an unpacked-to value,

(f) storing, for each sample from the ones of said predetermined number of groups, the mapped value from said first table corresponding to the one in said series of ranges within which the difference between said sample and the unpacked-to value for the immediately previously stored sample falls,

(g) referencing a second table for said digital samples outside of said predetermined number of groups, said table having a series of ranges of said differences in amplitude between consecutive digital samples, wherein each range corresponds to a mapped value and an unpacked-to value,

(h) storing, for each Rth sample from said samples outside of said predetermined number of groups, the mapped value from said second table corresponding to the one in said series of ranges within which the difference between said sample and the unpacked-to value for the immediately previously stored sample falls, wherein R is an integer greater than one and is a function of the sampling rate.

58. The method of claim 57 wherein R is an integer between 2 and 6.

59. The method of claim 57 including the further steps of:

storing a window-start code byte immediately prior to storage of the first mapped value from each one of said predetermined number of groups.

60. The method of claim 57, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

61. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate of at least 100 samples per second from at least one channel of an electrocardiogram apparatus,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) referencing a first table for said first predetermined number of groups, said table having a series of ranges of said differences in amplitude between consecutive digital samples, wherein each range corresponds to a mapped value and an unpacked-to value,

(f) storing, for each sample from the ones of said predetermined number of groups, the mapped value from said first table corresponding to the one in said series of ranges within which the difference between said sample and the unpacked-to value for the immediately previously stored sample falls,

(g) identifying a second predetermined number of groups of a predetermined number of consecutive samples having sums of the absolute values of said differences which exceed a predetermined threshold,

(h) referencing a second table for said second predetermined number of groups, said table having a series of ranges of said differences in amplitude between consecutive digital samples, wherein each range corresponds to a mapped value and an unpacked-to value,

(i) storing, for selected samples from said second predetermined number of groups, the mapped value from said second table corresponding to the one in said series of ranges within which the difference between said sample and the unpacked-to value for the immediately previously stored sample falls,

(j) referencing a third table for said digital samples outside of said first predetermined number of groups and said second predetermined number of groups, said table having a series of ranges of said differences in amplitude between consecutive digital samples, wherein each range corresponds to a mapped value and an unpacked-to value,

(k) storing, for each Rth sample from said samples outside of said first predetermined number of groups and said second predetermined number of groups, the mapped value from said third table corresponding to the one in said series of ranges within which the difference between said sample and the unpacked-to value for the immediately previously stored sample falls, wherein R is an integer greater than one and is a function of the sampling rate.

62. The method of claim 61 wherein R is an integer between 2 and 6.

63. The method of claim 61 including the further steps of:

storing a window-start code byte immediately prior to storage of the first mapped value from each one of said first predetermined number of groups, and,

storing a pseudo-window start code byte immediately prior to storage of the first mapped value from said second predetermined number of groups.

64. The method of claim 61, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte for each channel representing the present value of the amplitude of the sampled signal on that channel.

65. A method of digitally storing episodic signals in solid state memory, including the steps of:

(a) sampling analog episodic signals at a fixed sampling rate,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) storing a number related to each digital sample from the ones of said predetermined number of groups,

(f) storing a number related to every Rth digital sample from the digital samples outside of said predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

66. The method of claim 65 including the further step of storing a window-start code byte prior to storage of the first number related to each digital sample from each one of said first predetermined number of groups.

67. The method of claim 65, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte representing the present value of the amplitude of the sampled signal.

68. A method of digitally storing episodic signals in solid state memory, including the steps of:

(a) sampling analog episodic signals at a fixed sampling rate,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a first predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) storing a number related to each digital sample from said first predetermined number of groups,

(f) identifying a second predetermined number of groups of a predetermined number of consecutive samples having sums of the absolute values of said differences which exceed a predetermined threshold,

(g) storing a number related to selected digital samples from said second predetermined number of groups,

(h) storing a number related to every Rth digital sample from the digital samples outside of said first predetermined number of groups and said second predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

69. The method of claim 68 including the further steps of:

storing a window-start code byte immediately prior to storage of the first number related to each digital sample from each one of said first predetermined number of groups, and,

storing a pseudo-window start code byte immediately prior to storage of the first number related to selected digital samples from said second predetermined number of groups.

70. The method of claim 68, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte representing the present value of the amplitude of the sampled signal.

71. A method of digitally storing episodic signals in solid state memory, including the steps of:

(a) sampling analog episodic signals at a fixed sampling rate,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a first predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) storing each digital sample from said first predetermined number of groups,

(f) identifying a second predetermined number of groups of a predetermined number of consecutive samples having sums of the absolute values of said differences which exceed a predetermined threshold,

(g) storing selected digital samples from said second predetermined number of groups,

(h) storing every Rth digital sample from the digital samples outside of said first predetermined number of groups and said second predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

72. The method of claim 71 including the further steps of:

storing a window-start code byte immediately prior to storage of the first digital sample from each one of said first predetermined number of groups, and,

storing a pseudo-window code start byte immediately prior to storage of the first of the selected digital samples from each of said second predetermined number of groups.

73. The method of claim 71, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte representing the present value of the amplitude of the sampled signal.

74. A method of digitally storing episodic signals in solid state memory, including the steps of:

(a) sampling analog episodic signals at a fixed sampling rate,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) storing from said predetermined number of groups, a digital value related to the difference in amplitude between each digital sample and the digital sample preceding it,

(f) storing a digital value related to the difference in amplitude between every Rth digital sample from the digital samples outside of said predetermined number of groups, where R is an integer greater than one and is a function of the sampling rate.

75. The method of claim 74 including the further step of storing a window-start code byte prior to storage of the first digital value from each one of said first predetermined number of groups.

76. The method of claim 74, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte representing the present value of the amplitude of the sampled signal.

77. A method of digitally storing episodic signals in solid state memory, including the steps of:

(a) sampling analog episodic signals at a fixed sampling rate,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying from said selected channel a first predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) storing from said first predetermined number of groups a digital value related to the difference in amplitude between each digital sample and the digital sample preceding it.

(f) identifying a second predetermined number of groups of a predetermined number of consecutive samples having sums of the absolute values of said differences which exceed a predetermined threshold,

(g) storing a digital value related to the difference in amplitude between selected digital samples from said second predetermined number of groups,

(h) storing a digital value related to the difference in amplitude between every Rth digital sample from the digital samples outside of said first predetermined number of groups and said second predetermined number of groups, where R is an integer between 2 and 6.

78. The method of claim 77 including the further steps of:

storing a window-start code byte immediately prior to storage of the first digital value from each one of said first predetermined number of groups, and,

storing a pseudo-window start code byte immediately prior to storage of the first of the digital values related to the differences in amplitude between selected digital samples from each of said second predetermined number of groups.

79. The method of claim 77, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte representing the present value of the amplitude of the sampled signal.

80. A method of digitally storing episodic signals in solid state memory, including the steps of:

(a) sampling analog episodic signals at a fixed sampling rate,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) referencing a first table for said predetermined number of groups, said table having a series of ranges of said differences in amplitude between consecutive digital samples, wherein each range corresponds to a mapped value and an unpacked-to value,

(f) storing, for each sample from the ones of said predetermined number of groups, the mapped value from said first table corresponding to the one in said series of ranges within which the difference between said sample and the unpacked-to value for the immediately previously stored sample falls,

(g) referencing a second table for said digital samples outside of said predetermined number of groups, said table having a series of ranges of said differences in amplitude between consecutive digital samples, wherein each range corresponds to a mapped value and an unpacked-to value,

(h) storing, for each Rth sample from said samples outside of said predetermined number of groups, the mapped value from said second table corresponding to the one in said series of ranges within which the difference between said sample and the unpacked-to value for the immediately previously stored sample falls, wherein R is an integer greater than one and is a function of the sampling rate.

81. The method of claim 80 including the further steps of:

storing a window-start code byte immediately prior to storage of the first mapped value from each one of said predetermined number of groups.

82. The method of claim 80, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte representing the present value of the amplitude of the sampled signal.

83. A method of digitally storing electrocardiogram signals in solid state memory, including the steps of:

(a) sampling analog electrocardiogram signals at a fixed sampling rate,

(b) converting each sampled signal to a multibit digital sample representing its amplitude,

(c) calculating the differences in amplitude between each consecutive digital sample,

(d) identifying a first predetermined number of groups of a predetermined number of consecutive digital samples having the highest sums of the absolute values of said differences,

(e) referencing a first table for said first predetermined number of groups, said table having a series of ranges of said differences in amplitude between consecutive digital samples, wherein each range corresponds to a mapped value and an unpacked-to value,

(f) storing, for each sample from the ones of said predetermined number of groups, the mapped value from said first table corresponding to the one in said series of ranges within which the difference between said sample and the unpacked-to value for the immediately previously stored sample falls,

(g) identifying a second predetermined number of groups of consecutive samples having sums of the absolute values of said differences which exceed a predetermined threshold,

(h) referencing a second table for said second predetermined number of groups, said table having a series of ranges of said differences in amplitude between consecutive digital samples, wherein each range corresponds to a mapped value and an unpacked-to value,

(i) storing, for selected samples from said second predetermined number of groups, the mapped value from said second table corresponding to the one in said series of ranges within which the difference between said sample and the unpacked-to value for the immediately previously stored sample falls,

(j) referencing a third table for said digital samples outside of said first predetermined number of groups and said second predetermined number of groups, said table having a series of ranges of said differences in amplitude between consecutive digital samples, wherein each range corresponds to a mapped value and an unpacked-to value,

(k) storing, for each Rth sample from said samples outside of said predetermined number of groups and said second predetermined number of groups, the mapped value from said third table corresponding to the one in said series of ranges within which the difference between said sample and the unpacked-to value for the immediately previously stored sample falls, wherein R is an integer greater than one and is a function of the sampling rate.

84. The method of claim 83 including the further steps of:

storing a window-start code byte immediately prior to storage of the first mapped value from each one of said first predetermined number of groups, and,

storing a pseudo-window start code byte immediately prior to storage of the first mapped value from each of said second predetermined number of groups.

85. The method of claim 83, including the further step of providing a reset byte every fixed number of seconds, immediately followed by a byte representing the present value of the amplitude of the sampled signal.

86. Portable apparatus for digitally storing electrocardiogram signals in solid state memory, including:

input amplifier means for conditioning and amplifying analog electrocardiogram signals,

digitizing means, connecting to said input amplifier means, for converting said analog electrocardiogram signals to digital representations thereof,

buffer means, connected to said digitizing means, for temporarily storing a plurality of said digital representations,

processor means, communicating with said buffer means, for determining, from said plurality of digital representations, the existence of window and pseudo-window intervals therein and the data to be stored in accordance therewith, and for directing the storage of said date,

a dynamic random access memory, communicating with said buffer means and said processor means, having a plurality of storage locations, for storing said data,

circuit means for determining the temperature of said apparatus,

refresh control means, responsive to said circuit means, for controlling the refresh rate of DRAMs of said dynamic random access memory in response to the temperature of said apparatus.

87. The portable apparatus of claim 86 wherein said refresh control means controls the refresh rate of dynamic random access memory to be 150 mSec at 25°

C. and 80 mSec at 45°

C.

88. The portable apparatus of claim 87 wherein said refresh control means causes said refresh rate to vary approximately exponentially with temperature.

89. Portable apparatus for digitally storing electrocardiogram signals in solid state memory, including:

first and second input amplifier means for conditioning an amplifying analog electrocardiogram signals on a first and a second channel, respectively,

first and second digitizing means, connected to said first and second input amplifier means respectively, for converting said analog electrocardiogram signals to digital representations thereof,

first and second buffer means, connected to said first and second digitizing means respectively, for temporarily storing a plurality of said digital representations,

processor means, communicating with said first and second buffer means for determining which of the signals on said first and second channels has the fewest artifacts, for determining, from said plurality of digital representations, data to be stored, and for directing the storage of said data,

a dynamic random access memory, communicating with said buffer means and said processor means, having a plurality of storage locations, for storing said data,

circuit means for determining the temperature of said apparatus,

refresh control means, responsive to said circuit means, for controlling the refresh rate of DRAMs of said dynamic random access memory in response to the temperature of said apparatus.

90. The portable apparatus of claim 89 wherein said refresh control means controls the refresh rate of dynamic random access memory to be 150 mSec at 25°

C. and 80 mSec at 45°

C.

91. The portable apparatus of claim 89 wherein said refresh control means causes said refresh rate to vary approximately exponentially with temperature.