Automatic medical sign monitor

• US 5,671,734 A
• Filed: 11/03/1995
• Issued: 09/30/1997
• Est. Priority Date: 11/03/1995
• Status: Expired due to Term

• Current Assignee: The United States of America As Represented By The Secretary of Agriculture
• Original Assignee: the united states of america as represented by the secretary of the navy
• Inventors: Pugh, Jamie K.
• Primary Examiner(s): Bahr, Jennifer
• Assistant Examiner(s): HUANG, STEPHEN D
• Application Number:
  US08/552,818
• Time in Patent Office: 697 Days
• Field of Search: 128/671, 128/630, 128/637, 128/668, 128/687, 128/670, 128/672, 128/691, 128/695 R, 128/716, 128/725, 364/413.03, 364/413.02, 364/415, 364/416
• US Class Current: 600/301
• CPC Class Codes: G16H 15/00 ICT specially adapted for medical reports, e.g. generation or transmission thereof
• CPC Class Codes: G16H 20/13 delivered from dispensers
• CPC Class Codes: G16H 40/63 for local operation

Abstract

An automatic medical monitor of the present invention comprises medical s sensors for collecting a time-ordered set of values representative of medical signs such as pulse, respiration, and blood pressure. The medical sign sensors are coupled to a medical sign data processor that forms statistics from the medical sign data and forms a modified Fast Initial Response (FIR) Shewhart cumulative sum and a variance cumulative sum to detect changes in health state. When a change in health state is detected, the medical sign data processor displays the statistics on a display and logs them on a printer.

Claims

1. An automatic medical sign monitor comprising:

a medical sign sensor for collecting a time ordered set of medical sign data;

a medical sign data processor coupled to said medical sign sensor for detecting a change in health state by forming statistics from said medical sign data comprising;

a fast initial response (FIR) Shewhart cumulative sum modified to include;

a Shewhart statistic having a first upper threshold and a first lower threshold selected to result in a probability of false alarm substantially equal to a reciprocal of an average run length,

a fast initial response cumulative sum having a second upper threshold and a second lower threshold selected to result in a probability of false alarm substantially equal to said reciprocal of said average run length for detecting a shift in mean value,

and a second cumulative sum for estimating a current mean value and an elapsed time of said shift in mean value;

and a variance cumulative sum modified to detect a shift in variance and to estimate a current variance value and an elapsed time of said shift in variance, and to include a third upper threshold and a third lower threshold selected to result in a probability of false alarm substantially equal to said reciprocal of said average run length;

and an output device coupled to said data processor for outputting at least one of said medical sign data and said statistics.

2. The automatic medical sign monitor of claim 1 wherein said output device comprises a display.

3. The automatic medical sign monitor of claim 1 wherein said output device comprises a printer for logging said statistics.

4. The automatic medical sign monitor of claim 1, wherein said statistics further include an event table of said medical sign data.

5. The automatic medical sign monitor of claim 1 wherein said statistics further include at least one of weighted rate of change, damped weighted rate of change, quadratic regression, and prediction equations developed from non-linear multivariate regression techniques for forecasting future values of said medical signs.

6. The automatic medical sign monitor of claim 1 wherein said medical sign data processor compares dispensed quantities of medication with prescription directions to estimate consumption and displays medication shortfalls.

7. The automatic medical sign monitor of claim 1 wherein said medical sign processor forecasts estimates of Hansen'"'"'s disease Erythema Nodosum Leprosum (ENL) reaction, reversal reaction, and bacilli index from prediction equations developed from non-linear multivariate regression techniques.

8. A method for automatically monitoring medical signs comprising the steps of:

sensing medical signs to form a time ordered set of medical sign data;

detecting a change in health state by forming statistics from the medical sign data comprising;

a fast initial response (FIR) Shewhart cumulative sum modified to include;

a Shewhart statistic having a first upper threshold and a first lower threshold selected to result in a probability of false alarm substantially equal to a reciprocal of an average rim length,

a fast initial response cumulative sum having a second upper threshold and a second lower threshold selected to result in a probability of false alarm substantially equal to the reciprocal of the average run length for detecting a shift in mean value,

and a second cumulative sum for estimating a current mean value and an elapsed time of said shift in mean value;

and a variance cumulative sum modified to detect a shift in variance and to estimate a current variance value and an elapsed time of the shift in variance, and to include a third upper threshold and a third lower threshold selected to result in a probability of false alarm substantially equal to the reciprocal of the average run length;

and outputting at least one of the medical sign data and the statistics.

9. The method of claim 8 wherein the step of outputting forecasts includes displaying the medical sign data and statistics.

10. The method of claim 8 wherein the step of outputting forecasts includes logging the statistics on a printer.

11. The method of claim 8 wherein the step of outputting forecasts includes forming an event table of the medical sign data.

12. The method of claim 8 wherein the step of forming statistics further includes at least one of weighted rate of change, damped weighted rate of change, quadratic regression, and prediction equations developed from non-linear multivariate regression techniques for forecasting future values of the medical signs.

13. The method of claim 8 further including the step of comparing dispensed quantities of medication and prescription directions to estimate consumption and displaying medication shortfalls.

14. The method of claim 8 further including the step of estimating Hansen'"'"'s disease Erythema Nodosum Leprosum (ENL) reaction, reversal reaction, and bacilli index forecasts from prediction equations developed from non-linear multivariate regression techniques.