MFER FAQs

Japanese

[Q-1] Why MFER standard is required?: Around 2000, MFER committee found the following problems about standardization of medical waveform throughout the evaluation of existing standards such as HL7, SCP-ECG, and DICOM.; (1) Specialized in data communication (HL7 V2 waveform): [Data exchange method]; - Sending side: Convert original data format into HL7 V2 waveform.; - Receiving side: Convert HL7 V2 waveform into original data format. (Receiving side can store and use HL7 data without conversion if needed); HL7 will be able to describe data of any form. However, in some cases, it requires data cutdown and it cannot describe manufacturers' characteristic specifications. Sending side and Receiving side have same problems. Both sides require specialized technical knowledge of waveform and signal processing to solve the problems.; (2) Only 12 lead ECG description (SCP-ECG): Essential feature of SCP-ECG was organized in 1993. In those days, data compression was a key issue for data storage and communication, because memory was expensive and communication speed was low. So, the main challenge was how to storage more data on a floppy disk and how to communicate fast by RS232C. SCP-ECG realized high-compression by converting precipitous waveforms such as QRS and gentle parts such as base line using following techniques.; - Extract typical waveform data.; - Filter waveform data.; - Out put the difference of waveform data.; - Change sampling frequency.; SCP-ECG requires implementation of CRC and format check. In addition, manufacturer code is standardized so as to evaluate diagnosis performance. As a result of standardization, validation became complex. Since SCP-ECG specializes in 12 lead ECG, some functions are excluded from the specifications. It causes a difficulty on confirmation of linearity. (e.g. only 1 minute ECG is acquired); (3) Wide application and excessively spelled out standard (DICOM): DICOM had been considered a waveform standard around 2000. It focused on a waveform of heart catheterization in those days. Holter ECG was added to the standard afterward reflecting market demand. There are several waveforms exist in the medical front. So, it is desirable that all waveforms have DICOM standards. Since DICOM requires detailed and strict standardization, it disturbs legislative process. See from a cost-effectiveness standpoint, DICOM waveform standards are inefficient. Therefore, as for 12 lead ECG, only limited functions (e.g. 13 channel waveform and 13 sec description) are standardized.; Following items are required for medical waveform handling.; (1) There is a wide range of variations in waveform description: Most often-used waveform is electrocardiogram. Particularly, 12 lead resting ECG is used on a routine basis. Usually, 12 lead resting ECG acquires 10 or 15 sec of limb lead (I, II , III, aVR, aVL, avF) and chest leads (V1, V2, V3, V4, V5, V6) data. Extra chest leads such as V3R, V4R, V7, and V8 are added as needed basis. In some cases, ECG is acquired for a few minutes. There are many additional needs such as the following.; - Record reverse polarity such as Cabrera lead (-aVR); - Change the leads order; - Capture typical waveform such as P-QRS-T; - Exercise stress test; - Used as a kind of monitoring waveform at ICU and OR; Since those needs are necessary for clinical practice, we cannot ignore them.; (2)Storage method and combination of waveform data: Waveform data is used as single agent or in combination. For example, blood pressure waveform, pulse wave, phonocardiogram, and electroencephalogram are measured with electrocardiogram. Also, there are several inspection time from short (spot check) to long period (such as 24h holter test). From the usage point of view, 24h holter test uses complete data. By contrast, monitoring uses ever-changing waveform. Many of the latest ECGs only storage "I" and "II" lead of limb lead and other leads are calculated by operation. Accurate synchronization, linearity, and standardized storage condition are required for this function. Needless to say, waveform data stored according to the unified standard is easy to handle.; (3) Merit and demerit of computerization: Display and storage of electronically-stored data depend on performance and feature of display/recording devices. So, there is a possibility that devices are inadequate for clinical qualities and requirements. As for display devices including computer, resolution, display colors and performance affect waveform accuracy. It makes an impact on diagnosis. Take electrocardiogram as an example. If display performance does not meet required resolution, P wave or ST cannot be displayed accurately. In addition, if display/recording devices do not meet required measurement accuracy such as "1mV/1cm" or "1sec/1cm", it will affect diagnoses. Measurement functions such as enlarged display and scale are required to cover the performance deficiency. On the other hand, if original data are stored electronically, it is easy to perform filtering and processing because they are recoverable. Not only put computerization in the place of paper, but it is also necessary to add merit of computerization for standard convention.; (4) It is widely utilized for clinical practice, research, and education: Computerization of medical waveform is required not only for clinical practice, but also for research and education. It is ideal that medical waveform data acquired by clinical practice are used in research and education consistently. This will require specifications and structures that support accurate waveform description and anonymity.; (5) Require technical knowledge: Specialized technical knowledge (medical knowledge, signal processing, and waveform processing) is required to deal medical waveform. Recently, clinicians and researchers are familiar with medical knowledge. And signal processing is well supported by general engineering. So, it is easy to acquire those specialists and learn a skill. In contrast, waveform processing is acquired as know-how of researchers and manufacturers' engineers. Each manufacturers process the waveform in the unique way. So, it is difficult to standardize it. If technical knowledge is required for usage of JPEG and MPEG, digital camera could not widely spread what it is today. Source of success is that specialists studied and provided the standard of image processing. As for medical information, specialists of medical waveform have to standardize it in the same manner.
[Q-2] Why HL7 is inadequate?: Regarding feature of HL7 standard, see Q-1. HL7 excels at V2 message conversion and CDA document description. For this reason, it is actual and reasonable to use MFER (waveform data reference) in combination with V2 (ordering and patient information management) and CDA (examination report).
[Q-3] Why DICOM is inadequate?: Regarding feature of DICOM standard, see Q-1. Hospitals are using DICOM server for radiological image management. If medical waveform data described by MFER can be saved on the same server, it is user friendly. Therefore, it is one of a subject of future investigation.
[Q-4] Why SCP-ECG is inadequate?: Regarding feature of SCP-ECG standard, see Q-1. MFER committee provides a guidebook for converting SCP-ECG into MFER. It is necessary to pay attention to the missing data caused by nonlinearity and filter.
[Q-5] Does MFER committee provide analysis software?: MFER committee does not have a plan to release analysis software at the moment (August, 2010). We will study based on the market situation. There is a possibility that we start at small task such as heart rate, maximum RR time detection, and noise evaluation. Since analog holter recorders have compatibility, analyzers spread around medium scale hospitals. Holter recorder manufacturers did not need to care about it. Digital holter recorders seem to be spreading rapidly because they have advantages in size and accuracy than 12 leads ECG. However, they are incompatible with other manufacturers' one. In other words, data acquired by manufacturer A's recorder cannot be analyzed by manufacturer B's analyzer. Besides, it is necessary to study specifications and price of holter recorders and analyzers. If new methodology such as HRV is developed, all manufacturers have to mount an analysis module on their products. To salvage the situation, MFER-compliant recorders are required. MFER committee has been studying compatibility of holter electrocardiogram with Japanese well-known manufacturers. Recently, some analyzers can convert manufacturers' original data into MFER and/or directly receive MFER data from a recorder.
[Q-6] Is electrocardiogram as image information inadequate for medical care information system?: Typically, Electronic Medical Record system such as electronic health record system implements industry standard of image processing. JPEG or PDF waveform data are handled as part of the category. In case image information displayed on the screen or printed out, it has a risk to cause distortions (JIS strictly regulates the standard of electrocardiogram) because of resolution and printer setting. It poses a potential risk of incorrect diagnosis. Image information is sufficient and convenient to give an explanation to a patient or to conduct brief overview. However, image information cannot be used for secondary use such as re-measurement or reprocessing. It is necessary to standardize universal waveform standard so as to ensure Human readability, Integrity, and Persistence for medical waveform data. While image information cannot convert into waveform data such as MFER, waveform data described by MFER can convert into image information.
[Q-7] When MFER-compliant ECG will be launched?: Japanese manufacturers have already launched MFER-compliant ECGs. They have also released data receiving and viewer systems. MFER committee will work hard to increase MFER-compliant products.
[Q-8] Is it better to add preamble?: Identifier is defined as follows according to the MFER's basic encoding rule. @MFR (40, 20, 4D, 46, 52, �c.) [MWF_PRE(64)]. Since description of preamble is discretionary, it depends on a project's conclusion.
[Q-9] Is it preferable to use UID like DICOM?: If unique identification of waveform is necessary, UID (Unique Identifier) is available [MWF_UID(87)]. For example, UUDI (Universally Unique Identifier) and OID (Object Identifier) is selectable. OID is unique in general. However, ID will be able to track back by issuing institution. For example, OID of MFER (1.2.392.200119.6) has uniqueness as noted in the attached. UUDI is standardized by OSF (Open Software Foundation). GUID ( ? ) developed by Microsoft is now used commonly. Since UUDI uses random number, it is unable to track back the object. Please make an appropriate decision about usage of UID including utilization purpose, privacy issue, and availability.