DOC PREVIEW
Personal Healthcare

This preview shows page 1 out of 4 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 4 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 4 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Diagnostic Quality Driven Physiological Data Collection for Personal Healthcare David Jea, Rahul Balani, Ju-Lan Hsu, Dae-Ki Cho, Mario Gerla, and Mani B. Srivastava Abstract—We believe that each individual is unique, and that it is necessary for diagnosis purpose to have a distinctive combination of signals and data features that fits the personal health status. It is essential to develop mechanisms for reducing the amount of data that needs to be transferred (to mitigate the troublesome periodically recharging of a device) while maintaining diagnostic accuracy. Thus, the system should not uniformly compress the collected physiological data, but compress data in a personalized fashion that preserves the “important” signal features for each individual such that it is enough to make the diagnosis with a required high confidence level. We present a diagnostic quality driven mechanism for remote ECG monitoring, which enables a notation of priorities encoded into the wave segments. The priority is specified by the diagnosis engine or medical experts and is dynamic and individual dependent. The system pre-processes the collected physiological information according to the assigned priority before delivering to the backend server. We demonstrate that the proposed approach provides accurate inference results while effectively compressing the data. I. INTRODUCTION Remote health monitoring is probably one of the most fundamental functionalities that a personalized healthcare system should provide. Smartphones are fast emerging as the platform for wireless healthcare via affordable medical sensors either embedded in phone or connected through Bluetooth [14][17]. These technologies can not only trans-form the way healthcare is managed and delivered to indi-viduals, but also enable more accurate aggregation of indi-vidual data into health models. It has been known that real-time collection and transmis-sion of high rate physiological data impose a huge burden on the limited energy resources available to smartphones and sensors. Our tests show that at low data rates (approximately 60bytes/s) and frequent transmissions (once every 0.5-1sec), a cell phone can only last 10-12 hours at most while continually transmitting data over a strong WiFi connection. This means a recharge cycle of every half day for users. It has also been shown that WiFi radios are more energy efficient than cellular radios for infrequent bulk data transfers [12]. Thus, from usability perspective, it is more favorable to reduce the amount of data for transmission to prolong recharge cycle. We believe that each individual is unique with diverse physical conditions. Automated or semi-automated persona-lized healthcare should reflect this uniqueness in physiologi-cal signals and associated parameters of an individual. A remote monitoring system should preserve “important” data features in the related physiological signal for accurate di-agnosis during the transmission from user end to server. General data reduction mechanisms either uniformly com-press the data or preserve only certain features in the data. The system thus needs a higher-level personalized descrip-tion to instruct the proper data processing mechanism for each individual. Medical diagnosis is usually based on inferences drawn, not from raw sensor sample streams, but from the data fea-tures extracted from them. The importance of a data feature is best defined in terms of its diagnostic quality, or the in-cremental inference accuracy derived from it. With a nota-tion of priorities encoded into wave segments, we present a diagnostic quality driven framework. The priority is speci-fied by the diagnosis engine or medical experts and is dy-namic and individual dependent. The assigned priorities allow a data reduction while maintaining diagnostic accura-cy by preserving specific signal features. To demonstrate and assess the concept of diagnostic quality driven band-width-scalable physiological data collection, we design and implement a supporting processing framework. II. RELATED WORK Lossy ECG compression achieves a higher compression ratio by allowing distortion. Various techniques have been well developed [1, 8 - 10]. Quality controlled compression recently has grabbed researchers’ attention [19 - 21]. These systems uniformly compress data to a given threshold. [22] developed a compression that is specialized for signal com-ponents. We propose to encode priorities into waveforms such that a system can process ECG accordingly in a more personal fashion. A concept similar to the one discussed in this paper has been explored in the field of image/video compression in the context of “Region of Interest”. The JPEG 2000 Standard [5] provides mechanisms to label and compress different parts of a picture into a different degree of fidelity. The MP3 encodes audio based on psychoacoustic models and discards less audible (to human hearing) components [7]. III. DIAGNOSTIC INFORMATION IN ECG We choose ECG signal, as the subject throughout rest of this study, to investigate the idea of preserving diagnosis-related features while transmitting compressed data. ECG has the benefits that it has been well studied by the medical community and has proven its clinical importance. Besides, everyone has a unique ECG [16] and thus motivates the im-portance of personalized healthcare. Other physiological signals also apply to the same idea in preserving features. A typical ECG waveform of one normal heartbeat con-sists of three characteristic waves: a P wave, a QRS complex and a T wave. There is a small U wave that is not always visible. ECG reflects the electrical activity of the heart mus-cle and is recorded by electrodes attached to the body sur-face. P wave relates to the depolarization of the atria. QRS complex captures ventricular depolarization and T wave indicates ventricular repolarization of the heart. With differ-ent origins in heart, it is natural that each wave segment in ECG has its own clinical significance in diagnosis of vari-ous cardiac related diseases. The concept of preserving diagnostic information of phy-siological signals


Personal Healthcare

Download Personal Healthcare
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Personal Healthcare and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Personal Healthcare 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?