Technique of Receiving Data from Medical Devices to Create Electronic Medical Records Database

INTRODUCTION

In the electronic medical record processing model, the process of digitising medical data is an important key to creating an electronic data source to deploy intelligent IT applications as well as to evaluate the optimisation of the model. Meanwhile, patients’ medical data is diverse and complex, with many different types of data, from different types of medical devices, from various manufacturers. Thus, the process of digitising these medical data is an enormous task. In addition, the strict requirements of the integrity, accuracy and confidence level when the patient data is electronically processed is also a challenge [I].

Imaging data from diagnostic devices comply with Digital Imaging and Communications in Medicine (DICOM), while modern laboratories use data management systems such as Radiology Information System (RIS), Picture Archive and Communication System (PACS) or Laboratory Information System (LIS). Therefore, the electronic process of medical data is quite simple and accurate. However, other forms of medical data such as videos, graphics, numerals and others still must be entered manually from a computer keyboard and specialised scanners are used to digitise data from film, and then printed on paper or stored separately on CD and DVD drives [2]. The digitising of medical data using the above methods has shown many shortcomings. According to a study, on average, in Australia about 13.3% of the doctors’ conclusions had at least one error caused by typing from a computer keyboard. In New Zealand the average is 50%, the US is 59%, Canada is 60%, Ireland is 65.5% and Sweden is 66% [3]. In developing countries, a major feature is the numerous models of general hospitals with a variety of types and origins of medical devices that generate medical data. Therefore, the process of digitising this data source is even more complicated. This chapter presents several techniques to collect data from different types of medical devices and evaluate the quality of data after receiving it to form a database of I-EMR in an electronic hospital model [4-5].

AUTOMATIC TECHNIQUE OF ACQUIRING ANALOG ELECTRONIC MEDICAL IMAGES

Introduction and Classification

Medical data in the form of images, video and audio are mainly generated from the current diagnostic imaging devices such as diagnostic ultrasound equipment, endoscopic equipment or electron microscopes. These are devices that are considered routine in the health sector during medical examination and treatment. Therefore, the number of medical records of this type created daily in health facilities is very large. However, the manufacturers of this device group only provide data processing functions directly on the device such as display, printing, analysis and temporary storage in the device’s memory, CD/DVD storage or film printing. To provide the function of communicating with external peripherals for data processing purposes, the devices are mainly designed in the analog video output port. Therefore, in order to be able to receive and store these types of medical data in the form of electronic medical records on computers, a method of collecting and converting these data into digital forms is required [6-8].

The hardware module automatically captures the analog electronic medical image from the imaging device with standard communication interface to convert to digital format on the computer

FIGURE 10.1 The hardware module automatically captures the analog electronic medical image from the imaging device with standard communication interface to convert to digital format on the computer.

Automatic Acquisition of Analog Electronic Medical Image from Imaging Equipment with Standard Communication Interface

The principal diagram for the process of automatically collecting analog electronic medical data from imaging equipment with standard communication is shown in Figure lO.l. Accordingly, analog electronic medical data including images, video and audio from imaging devices with communication standards will be collected and converted into digital data by the hardware circuit module. This digital data is then sent to computers and processing software to perform the process of formatting, reproducing, displaying and storing medical data into the database in a digitised form [9].

The technique to implement the proposed process includes two main contents: (1) Designing hardware circuit to capture and convert electronic medical data in the form of analog images and videos into digital form and (2) building computer processing software to format, reproduce and store electronic medical data in the form of digital images and videos.

Hardware circuit design captures and converts electronic medical data in the form of analog images and videos into digital form.

To be able to automatically receive electronic medical data in the form of analog images and videos, first it is necessary to convert these data from analog to digital format with acceptable errors in the processing of medical information. Based on the characteristics of the analog electronic image data and video data in communication standards on medical imaging devices (including horizontal scan frequency, vertical scan frequency, video signal bandwidth, frame rate), this function is proposed to be carried out on electronic circuits with detailed block diagram described in Figure 10.2. In particular, the analog-to-digital convertor (ADC) continuously performs the conversion of analog video signals to digital signals. To ensure the ability to accurately reproduce video signals after digitisation, the sampling frequency selected in the high range is 15 Msps. This sampling rate is greater than the data transfer rate via universal serial bus (USB), so a high-speed 256-kB medium memory has been used to store data temporarily. Accordingly, the lines of continuous output from the ADC are stored in the intermediate memory one by one. The storage process is paused when a complete frame is encoded to begin the data transfer process. The pause is

Circuit diagram proposed to receive and convert electronic video medical data format into digital format at a rate of 3 to 12 frames per second

FIGURE 10.2 Circuit diagram proposed to receive and convert electronic video medical data format into digital format at a rate of 3 to 12 frames per second.

maintained until all frame data in the memory is transferred to the computer software via USB [10].

On the other hand, an analog video signal is sent to the synchronous pulse detector. The output of the synchronous pulse detector is a line and frame synchronisation signal that is fed into the logic block and the time synchronisation to ensure that the storage process on the intermediate memory is performed correctly and without overlapping. This block ensures that the archiving is carried out line by pixel until the end of the frame. In addition, control signals received from the USB port sent by the software on the computer are fed into the console block to affect other blocks in the system. This block only affects memory block, logic block and time synchronisation, and 8-bit buffer. Thus, the operation of the ADC block, the synchronous pulse separator and the clock generator are continuous and relatively independent of each other. The detailed design diagram for the acquisition and conversion of analog electronic video medical data into digital format is proposed as shown in Figure 10.3. The parameters to achieve by following design circuit are as follow's: inputs are electronic medical data in the form of analog images and videos taken from the composite video plug standard; the output is digitised data and communicates according to USB standard to the computer; conversion speed is from 3 to 12 frames per second; the maximum size of each frame is 730x288 pixels; the encoder resolution is 8 bits per pixel; colour decoding supports Phase Alternating Line (PAL) and National Television System Committee (NTSC) colour systems for processing software.

Detailed design diagram of the receiving and converting electronic medical data from analog video to digital format

FIGURE 10.3 Detailed design diagram of the receiving and converting electronic medical data from analog video to digital format.

Schematic of acquisition and reproduction of video medical data from digital stream of hardware circuit

FIGURE 10.4 Schematic of acquisition and reproduction of video medical data from digital stream of hardware circuit.

Algorithm flowchart archives medical data in the form of images and videos into a database on a computer

FIGURE 10.5 Algorithm flowchart archives medical data in the form of images and videos into a database on a computer.

3. Develop an algorithm for reading and displaying images from a database: The algorithm for reading and displaying image medical data stored in a database as a binary byte is described in Figure Ю.6. The binary byte data stored in the database is read and checked in turn to be the last byte of the image. If not, the binary byte data will be converted to bitmap data and transferred to the Picture Box image data display tool. When the last byte is reached, the bitmap data in the Picture Box tool will be displayed as a digital image.

Algorithm flowchart reads and displays medical data in the form of images and videos from the database

FIGURE 10.6 Algorithm flowchart reads and displays medical data in the form of images and videos from the database.

4. Following are the main processing functions:

AxVideoCapl.Device()//function receives data from hardware circuit posts/

AxVideoCapl. VideoFormat, for example RGB24 (800x600)///function selects data format for image and video posts//

AxVideoCapl.VideoStandard, for example PAL standard or NTSC standard;//function determines the video standard post// AxVideoCapl.Start ()///function starts to receive and recreate video data//

AxVideoCapl.Pause()///function pauses the process of receiving and recreating//

AxVideoCapl.Stop ()///function finishes getting the process// AxVideoCapl.SnapShot2Picture<)///function receives and saves each photo frame//

Dim fs AsNewFileStream (Trim (pathf ile, FileMode .Open) // function of converting image data to bit stream//

DimDataf) AsByte = New [Byte] (fs. Length) „•//function of converting bit stream to byte array//

The result of an automated process of obtaining analog electronic medical data in image and video type from diagnostic ultrasound equipment,

FIGURE 10.7 The result of an automated process of obtaining analog electronic medical data in image and video type from diagnostic ultrasound equipment, (a) Connect and receive data from Siemens ACUSON X300 and (b) the software interface displays the post-acquisition and processing results.

The automated process of obtaining electronic medical data in the form of analog images and videos from diagnostic imaging devices has been designed and built by the author using the technique described above as illustrated in Figure 10.7.

Evaluation of the Quality of Electronic Medical Image after Acquisition

The author has collected image and video data from several diagnostic ultrasound devices that are being commonly used in hospitals, along with the original data stored temporarily on devices for obtaining evaluation and analysis data. Due to the characteristics of medical data in the form of images and videos including grayscale type (2D black-and-white ultrasound) and colour type (3D, 4D, endoscopy, electron microscopes), it is important to assess the quality of data collected for both types of images.

For grayscale image data, the acquisition and evaluation of three 2D black and white diagnostic ultrasound equipment, each device randomly captures 20 images of 20 different patients. The equipment performing the assessment includes: LogiglOO-GE device, SSDIOOO-Aloka device and AcusonX300-Siemens device. Parameters received for grayscale images from these devices are 640x480 pixels image size, 8 bits encoding. Performing the calculation of peak signal-to-noise ratio (PSNR) values for the corresponding image pairs, we are graphed showing the trend of variation of PSNR values described in Graph I0.l [Ю, 11].

For colour image and video, acquisition and evaluation on two popular colour diagnostic ultrasound devices, Accuvix XQ-Medison and EUB-6000-Hitachi. These are devices capable of producing colour ultrasound images as well as 3D colour videos when performing ultrasound. On each device, the author randomly collected five colour videos of five different patients. Each video, author evaluates to 20 frames, respectively. The

GRAPH 10.1 The chart shows the trend of changing the PSNR value for 20 pairs of multilevel grey ultrasound image data received on the computer and the original image data on the temporary memory of three survey devices: LogiglOO, SSD1000 and AcusonX300. In the 20 pairs of images assessed, the PSNR value varied from 35.49 dB to 41.63 dB.

received colour video data parameters from these devices include: frame size 640x480 pixels, 24 bits encoding for three colours RGB. The PSNR values for each pair of frames on the two original colour videos from the device’s temporary memory and the colour video captured on the computer by the proposed method are shown in Graph 10.2.

GRAPH 10.2 The graph shows the trend of fluctuating PSNR values for 20 pairs of five colour video frames received on the computer and original video data received from the temporary memory of EUB6000 3D colour ultrasound equipment. In 20 pairs of rated frames, PSNR values varied from 37.62 dB to 41.40 dB.

From the charts showing the trend of fluctuating PSNR values for data acquisition and quality assessment of medical data in the form of images and videos we see: For the LogiglOO device, the value of PSNR varies from 35.49 dB to 37.96 dB. the average is 36.84 dB. For SSD1000 devices, the value of PSNR varies from 36.97 dB to 39.21 dB, the average is 38.03 dB. For the AcusonX300, the value of PSNR varies from 37.96 dB to 41.63 dB, the average is 39.72 dB. For Accuvix XQ devices, the value of PSNR varies from 38.19 dB to 41.82 dB, the average is 39.93 dB. For EUB6000, the value of PSNR varies from 37.62 dB to 41.40 dB, the average is 39.87 dB. Thus, compared to the assessment criteria of image and video data quality in processing health information at a good level, PSNR ranges from 35 to 40 dB [11, 12], the method of obtaining analog electronic medical data in the form of images and videos presented above is completely responsive. It can be used to automate the process of digitising of analog image and video medical data from medical imaging devices.

 
Source
< Prev   CONTENTS   Source   Next >