Measuring Breathing Airflow Using a Heat Sensitive Camera (ThermFlow)

High Resolution Thermal Imaging Developments for Non-Contact Airflow Measurement: A Novel Child-friendly Method to Monitor Paediatric Airflow.

ThermPaed is a Sheffield Children's Hospital research collaborative which investigated the clinical diagnostic capabilities of high resolution thermal imaging (HRTI) in paediatrics. This research utilises thermal imaging techniques to detect fractures, arthritic inflammation and infection. ThermPaed proved HRTI has potential in paediatrics and resulted in development of techniques that has led to the current research study. The purpose of this study is to develop HRTI as a non-contact child-friendly means of measuring and monitoring paediatric airflow.

Sheffield Children's Hospital Sleep Unit performs 500 sleep studies a year to investigate paediatric sleep-related breathing disorders. 1-6% of children have problems with their breathing during sleep and have breathing pauses called apnoeas. The condition needs accurate diagnosis and treatment to ensure correct blood oxygen levels. During a sleep study a number of physiological parameters are measured. These include heart rate, chest movements, airflow, oxygen levels, body position, snoring, movement and video. A total of 13 sensors are attached to the child whilst awake, and then the child settles to sleep. Airflow parameters are measured by two contact sensors; one measures airflow by temperature and the other measures airflow by pressure. Both sensors are placed in or very close to the nostrils to detect and classify apnoea. Due to their contact nature, these sensors are poorly tolerated. In a recent survey of 100 sleep studies on the sleep unit, 50% children did not allow the sensors to be attached or removed them immediately causing difficulties in interpreting their condition.

Breathing generates changes in temperature around nose and mouth which can be detected with HRTI. The investigators will recruit 30 children who are undergoing a sleep study. The investigators will compare airflow measurements from the existing clinical nasal sensors and HRTI to evaluate whether this new technique will provide an accurate non-contact alternative to measure respiration airflow.

The study will form part of a wider project currently ongoing which is looking into development of HRTI for clinical diagnosis and monitoring. It builds on the significant progress the investigators made through the ThermPaed study. Sheffield Hallam University supports this study by allocating an MPhil student (UR). The student will be jointly supervised at the University by Professor Saatchi who is well-established in digital signal and image processing and Professor Elphick and Dr Kingshott.

This study design is explained below. Measurement of thermal imaging alongside standard clinical parameters Following ethics and HRA approval, the investigators will identify 30 suitable children attending Sheffield Children's Hospital Sleep Unit who are undergoing cardiorespiratory PSG investigation for sleep-related breathing disorders.

Each patient would undergo the standard diagnostic cardiorespiratory PSG and wear all the sensors as usual. In addition, following consent, HRTI will be performed in the Sleep Unit overnight. The HRTI device will be positioned alongside the standard video camera used in PSG. The standard recording time of a PSG study is between 8-12 hours and is determined by the child and their age-related sleep and wake times. Even if the child does not tolerate the nasal sensors, the investigators will record HRTI to identify potentially missed apnoea events. The study therefore does not interfere in any way with the patient's normal procedures.

Patient, carer and staff opinions of the HRTI versus conventional nasal sensors.

As part of the research, the participating children, their attending carer and overnight sleep nursing staff will complete an evaluation questionnaire [Appendix 1] in the morning following the PSG.

Development of software to integrate the HRTI data with the PSG sleep system

The HRTI data recording and processing will be carried out by the MPhil student, supervised by his university tutor. The data will be stored on a password secure laptop during the recording and then transferred and stored in a safe, secure and password protected drive at Sheffield Hallam University. Each participating patient will be assigned a unique identifying number (see sections 5.3 and 6.0). There will not be any identifiable patient information with these recordings. The PSG data are downloaded onto a sleep software system at the hospital and then the anonymised version of the airflow signals will be stored at Sheffield Hallam University research secure drive. This sleep software has the ability to export its data anonymously and so currently can be integrated alongside the HRTI data. Workflow on integrating these two sets of data will include:

• Discussion with the sleep software company to directly integrate the thermal imaging data into the sleep software.
• Time synchronising the two data sets. Blinded scoring analysis of the signals to compare apnoea event scoring and classification Once each child's data are collected, Dr Kingshott will interpret the PSG data to identify apnoea events. Following standard AASM scoring guidelines apnoeas will be marked using the nasal sensor channels versus the HRTI channel. In order to minimise scoring bias, the processing order for HRTI versus the data from the nasal sensors will be randomised.

Inclusion Criteria:

• Children age 0-15.9 years Are attending the sleep unit for a cardiorespiratory sleep study to investigate or monitor sleep-disordered breathing

Exclusion Criteria:

• Families who do not understand written or verbal English where no interpreter is present
• Children who do not attend with their parent / legal guardian and so are therefore unable to give informed consent
• Children who are undergoing tracheostomy or ventilation sleep studies and so will not be expected to wear nasal airflow sensors during their sleep study