Activity and Sleep Apnea Monitoring of Aged-Person using Image Processing / 대한의료정보학회지

OBJECTIVE: As the rapid progress of aged society, there must be solutions of preparation against unpredicted accidents for aged solitary people. The most important thing that we must consider is the unconstraint of daily life. So, we are to develop a system and algorithm which meet our objectives. METHODS: We have monitored the degree of activity of the aged solitary person. The CCD camera was used not to disturb the daily activity and we evaluated the degree of activity using image processing on personal computer. The activity monitoring during night was assumed by sleeping on the bed, so the major method was breath monitoring during sleeping. On the other hand, daily activity was monitored by wide viewing camera in the living room. To prevent the privacy trouble, the acquired image was converted to binary form and the degree of breath and moving factor was estimated. RESULTS: In this paper we propose a new processing algorithm to accurately measure breathing characteristics in sleep apnea sufferers. We improved the conventional center-of-mass method and further applied the projection-profile method. As a result, we have improved breath measurement accuracy. In a comparison with conventional polysomnography, our method was 92% effective in detecting apnea cases. CONCLUSION: As a result of this study, we can monitor sleep apnea more simply and with no sleep interference. In measuring the activity of daily life, these improved algorithms were applied. So, we established a monitoring method of no-constrained, quantitative measurement for the aged solitary people during the whole day.

A Projection-based Intensity Correction Method of Phased-Array Coil Images

PURPOSE: To develop a novel approach to calculate the sensitivity profiles of the phased array coil for use in non-uniform intensity correction. MATERIALS AND METHODS: The proposed intensity correction method estimates the sensitivity profile of the coil to extract intensity variations that represent the scanned image. The sensitivity profile is estimated by fitting a non-linear curve to various angles of projections through the imaged object in order to eliminate the high-frequency image content. Filtered back projection is then used to compute the estimates of the sensitivity profile of each coil. The method was applied both to phantom and brain images from 8-channel phased-array coil and 4-channel phased-array coil, respectively. RESULTS: Intensity-corrected images from the proposed method have more uniform intensity than those from the commonly used `sum-of-squares' approach. By using the proposed correction method, the intensity variation was reduced to 6.1% from 13.1%, acquired from the `sum-of-squares'. CONCLUSION: The proposed method is more effective at correcting the intensity non-uniformity of the phased-array surface-coil images than the conventional`sum-of-squares' method.

A Projection-based Intensity Correction Method of Phased-Array Coil Images

PURPOSE: To develop a novel approach to calculate the sensitivity profiles of the phased array coil for use in non-uniform intensity correction. MATERIALS AND METHODS: The proposed intensity correction method estimates the sensitivity profile of the coil to extract intensity variations that represent the scanned image. The sensitivity profile is estimated by fitting a non-linear curve to various angles of projections through the imaged object in order to eliminate the high-frequency image content. Filtered back projection is then used to compute the estimates of the sensitivity profile of each coil. The method was applied both to phantom and brain images from 8-channel phased-array coil and 4-channel phased-array coil, respectively. RESULTS: Intensity-corrected images from the proposed method have more uniform intensity than those from the commonly used `sum-of-squares' approach. By using the proposed correction method, the intensity variation was reduced to 6.1% from 13.1%, acquired from the `sum-of-squares'. CONCLUSION: The proposed method is more effective at correcting the intensity non-uniformity of the phased-array surface-coil images than the conventional`sum-of-squares' method.

A Study on the ROI Optimizing Technique for Accurate Breath Measurements in Sleep Apnea / 대한의료정보학회지

OBJECTIVE: A sleep apnea is the most frequent symptom among sleep disorders. As the number of aged people are increasing, research activities are also increasing to monitor sleeping disorders of the elderly who lives alone. We proposed a new processing algorithm to measure the quantity of breaths accurately. METHODS: We improved the conventional center-of-mass method and further apply the projection profile method. RESULTS: In this study a system that can monitors respiration non-invasively and automatically, by identifying the movement of the chest and abdomen using image processing technique during sleep, was implemented. We can see breathing status in real time based on the acquired breathing waves. To verify the designed system, the values from the polysomnography were compared and analyzed to validate the system accuracy. As a result, the mean accuracy is 96%. CONCLUSION: We can simply monitor the sleep apnea with no sleep interferences.