Non-contact vital sign monitoring in the clinic
<p>Current monitoring systems available to track changes in the vital signs of patients (such as heart rate, respiratory rate or <em>SpO<sub>2</sub></em>) require adhesive electrodes or sensors to be in direct contact with the subjects. Most patients find the probes dif...
| 1. Verfasser: | |
|---|---|
| Weitere Verfasser: | |
| Format: | Abschlussarbeit |
| Sprache: | English |
| Veröffentlicht: |
2017
|
| _version_ | 1826316551298482176 |
|---|---|
| author | Montoya, M |
| author2 | Tarassenko, L |
| author_facet | Tarassenko, L Montoya, M |
| author_sort | Montoya, M |
| collection | OXFORD |
| description | <p>Current monitoring systems available to track changes in the vital signs of patients (such as heart rate, respiratory rate or <em>SpO<sub>2</sub></em>) require adhesive electrodes or sensors to be in direct contact with the subjects. Most patients find the probes difficult to attach or wear over prolonged periods of time. The ideal technology would involve sensors with no direct contact with the patient. Non-contact sensing provides several advantages, such as: no subject participation to set the equipment up, no skin preparation or irritation, decrease in the risk of infection, and the potential to be seamlessly integrated into the patient's lifestyle.</p> <p>It has been previously established that the colour and volume changes in superficial blood vessels during the cardiac cycle can be measured using a digital video camera over short time periods and under tightly-controlled conditions with relatively still and healthy subjects. Using an off-the-shelf camera and standard ambient light, this thesis proposes data fusion algorithms to compute estimates of heart rate and respiratory rate, and also to detect changes in <em>SpO<sub>2</sub></em> in a real hospital scenario, without interfering with regular patient care.</p> <p>A clinical study was designed to record a wide range of physiological values from patients undergoing haemodialysis treatment in the Renal Unit of the Churchill Hospital in Oxford. Video was recorded for a total of 104 dialysis sessions from 40 patients during the course of one year, comprising an overall video recording time of approximately 370 hours. Reference values were provided by two devices in regular clinical use.</p> <p>Algorithms are proposed to analyse the input video to identify time periods for which the location of the patient is known in the video frame, discarding periods for which there is high activity or which are affected by motion artefacts. The accuracy of heart rate estimation, for periods during which the subject is stable, is comparable to that of the two standard pulse oximeters used in the clinical study (positioned at the finger and earlobe respectively). The mean absolute error between the heart rate estimates from the camera and the average from two reference pulse oximeters was 2.8 beats/min for over 65% of the time. The mean absolute error between the respiratory rate estimates from the camera and the reference values (computed from the Electrocardiogram and a thoracic expansion sensor - chest belt) was 2.1 breaths/min for over 69% of the time for which the reference signals were valid. By calibrating the camera data with the reference pulse oximeters, changes in <em>SpO<sub>2</sub></em> could also be tracked during time periods with minimal patient motion.</p> <p>To increase the robustness of the algorithms, novel methods were devised for cancelling out aliased frequency components caused by the artificial light sources in the hospital, using auto-regressive modelling and pole cancellation. Maps of the spatial distribution of heart rate and respiratory rate information were developed from the coefficients of the auto-regressive models. Most of the periods for which the camera could not produce a reliable heart rate estimate were under 10 minutes, this opens the possibility to continuously monitor heart rate with a clinically relevant frequency.</p> |
| first_indexed | 2024-03-06T21:42:47Z |
| format | Thesis |
| id | oxford-uuid:488287d9-edf0-44de-9f83-ef25ef79a2e4 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-12-09T03:47:03Z |
| publishDate | 2017 |
| record_format | dspace |
| spelling | oxford-uuid:488287d9-edf0-44de-9f83-ef25ef79a2e42024-12-08T09:40:07ZNon-contact vital sign monitoring in the clinicThesishttp://purl.org/coar/resource_type/c_db06uuid:488287d9-edf0-44de-9f83-ef25ef79a2e4EnglishORA Deposit2017Montoya, MTarassenko, L<p>Current monitoring systems available to track changes in the vital signs of patients (such as heart rate, respiratory rate or <em>SpO<sub>2</sub></em>) require adhesive electrodes or sensors to be in direct contact with the subjects. Most patients find the probes difficult to attach or wear over prolonged periods of time. The ideal technology would involve sensors with no direct contact with the patient. Non-contact sensing provides several advantages, such as: no subject participation to set the equipment up, no skin preparation or irritation, decrease in the risk of infection, and the potential to be seamlessly integrated into the patient's lifestyle.</p> <p>It has been previously established that the colour and volume changes in superficial blood vessels during the cardiac cycle can be measured using a digital video camera over short time periods and under tightly-controlled conditions with relatively still and healthy subjects. Using an off-the-shelf camera and standard ambient light, this thesis proposes data fusion algorithms to compute estimates of heart rate and respiratory rate, and also to detect changes in <em>SpO<sub>2</sub></em> in a real hospital scenario, without interfering with regular patient care.</p> <p>A clinical study was designed to record a wide range of physiological values from patients undergoing haemodialysis treatment in the Renal Unit of the Churchill Hospital in Oxford. Video was recorded for a total of 104 dialysis sessions from 40 patients during the course of one year, comprising an overall video recording time of approximately 370 hours. Reference values were provided by two devices in regular clinical use.</p> <p>Algorithms are proposed to analyse the input video to identify time periods for which the location of the patient is known in the video frame, discarding periods for which there is high activity or which are affected by motion artefacts. The accuracy of heart rate estimation, for periods during which the subject is stable, is comparable to that of the two standard pulse oximeters used in the clinical study (positioned at the finger and earlobe respectively). The mean absolute error between the heart rate estimates from the camera and the average from two reference pulse oximeters was 2.8 beats/min for over 65% of the time. The mean absolute error between the respiratory rate estimates from the camera and the reference values (computed from the Electrocardiogram and a thoracic expansion sensor - chest belt) was 2.1 breaths/min for over 69% of the time for which the reference signals were valid. By calibrating the camera data with the reference pulse oximeters, changes in <em>SpO<sub>2</sub></em> could also be tracked during time periods with minimal patient motion.</p> <p>To increase the robustness of the algorithms, novel methods were devised for cancelling out aliased frequency components caused by the artificial light sources in the hospital, using auto-regressive modelling and pole cancellation. Maps of the spatial distribution of heart rate and respiratory rate information were developed from the coefficients of the auto-regressive models. Most of the periods for which the camera could not produce a reliable heart rate estimate were under 10 minutes, this opens the possibility to continuously monitor heart rate with a clinically relevant frequency.</p> |
| spellingShingle | Montoya, M Non-contact vital sign monitoring in the clinic |
| title | Non-contact vital sign monitoring in the clinic |
| title_full | Non-contact vital sign monitoring in the clinic |
| title_fullStr | Non-contact vital sign monitoring in the clinic |
| title_full_unstemmed | Non-contact vital sign monitoring in the clinic |
| title_short | Non-contact vital sign monitoring in the clinic |
| title_sort | non contact vital sign monitoring in the clinic |
| work_keys_str_mv | AT montoyam noncontactvitalsignmonitoringintheclinic |