Graphic file compression on a Macintosh

AutoDoubler, Compact Pro, DiskDoubler, Stufflt Deluxe, and Stufflt SpaceSaver compression applications for the Macintosh were tested for size of compressed files, and time to compress and expand. EPS, TIFF, PICT and Photoshop native format file types were used, in 1-bit, 8-bit greyscale, 8-bit colour (indexed), 24-bit colour (red, blue, green (RGB)), and 32-bit colour (cyan, magenta, yellow, black (CMYK)) modes. JPEG compression is illustrated by comparing an uncompressed image with three images compressed at different rates.     

基于JPEG2000的医学图像压缩

本文结合头部CT和MRI图像压缩，研究了JPEG2000标准和感兴趣区域（ROI）编码技术在医学图像压缩中的应用。通过计算峰值信噪比（PSNR），对编码后的图像进行率失真评价，进行了无损压缩和适当的有损压缩实验，并对JPEG2000标准和JPEG标准的应用效果进行对比。实验结果和分析表明，JPEG2000标准和ROI技术应用于医学图像压缩，对保证医学图像重要信息不失真，又能够提高压缩比，是一个优越的方案。     

Evaluation of a PC-based teleconferencing system for reading chest radiographs

Chest radiographs from 50 patients, including 25 with lung cancer, were digitized at a resolution of 100 dots/inch (40 dots/cm) and saved in JPEG format at a low compression ratio. The average size of the stored images was 55 kByte (range 6-86 kByte). Four respiratory specialists observed these images on the cathode ray tube (CRT) display of a PC-based teleconferencing system with a resolution of 800 x 600 pixels. One month later they observed the original chest radiographs. A receiver operating characteristic (ROC) analysis was performed on their diagnoses, based on a five-point confidence scale about the presence of lung cancer. The observer-specific Az values for the teleconferencing system ranged from 0.801 to 0.944, and the corresponding Az values for the conventional radiographs from 0.926 to 0.957. No significant differences were found between the teleconferencing images and the original images, which suggests that a PC-based teleconferencing system could be useful in the diagnosis of lung cancer on chest radiographs.     

基于JPEG2000算法的传统X光片数字化处理与质量评价

目的用JPEG2000算法对传统X光胶片进行数字化处理及有效压缩，并对有损压缩造成的图像信息缺失对临床诊断影响程度进行质最评价。方法选取80幅保存较好的传统X光胶片，通过专用胶片扫描仪对其进行数字化转换，用JPEG2000压缩软件对数字化图像分别进行无损压缩与有损压缩处理，采用主观图像质量测量方法对压缩图像进行质量评估。结果统计结果表明标准有损压缩与无损压缩的图像质量无明显差异，高压缩比有损压缩（20：1）与无损压缩存在图像质量差异，但不影响诊断。结论对传统X光胶片的数字化处理可以进行适度的有损压缩（低于20：1）。     

The effect of resolution, compression, colour depth and display modality on the accuracy of accident and emergency telemedicine

There are no current recommendations regarding the minimum technical specification for realtime telemedicine consultation in accident and emergency (A and E) practice. We assessed the effect of image resolution, compression, colour depth and display modality on perceived image quality and telediagnosis. Test sets of digitized radiographs and clinical images were subjected to a series of standardized manipulations and the resulting output files were evaluated by an expert panel using image scoring and receiver operating characteristic (ROC) analysis. For telemedicine in A and E work, the minimum technical specification should be regarded as images containing at least 250,000 pixels, compressed at up to JPEG 50 (or GIF for colour images) and displayed on a high-resolution computer monitor. These specifications resulted in average file sizes of 17 kByte for digital images and 9 kByte for radiographs.     

影像压缩对于乳腺小结节诊断的影响

使用JPEG压缩算法分别对于数字化后的乳腺影像进行无损压缩及有损压缩,并由3名放射科医师对于原始影像、无损压缩及有损压缩影像进行盲评,三次读片结果进行ROC分析,三次读片结果间未发现显著性差异。可以使用JPEG有损压缩算法来传输与存储乳腺影像。     

JPEG2000压缩技术对乳腺微钙化灶诊断性能影响的研究

目的:探讨采用JPEG 2000有损压缩技术压缩乳腺X摄影影像对乳腺簇状微钙化灶诊断性能的影响。方法:选择25例乳腺簇状微钙化灶X线摄影影像及30例正常对照病例影像,采用JPEG 2000压缩算法分别对其进行无损压缩及有损压缩,由2名放射科医生分两次进行阅片,并对是否发现乳腺簇状微钙化灶进行评价。采用受试者操作特性曲线对阅片结果进行分析。结果:2名放射科医生对采用2种不同压缩算法处理后的影像阅片结果相似,两者之间没有显著性差异。结论:采用有损压缩算法也能够满足乳腺簇状微钙化灶病例的阅片需求。     

International transmission of tele-ophthalmology images

Tele-ophthalmic consultations were conducted between a hospital in East Java, Indonesia, and an ophthalmology centre in Perth, Western Australia. Twenty-two eyes of 14 subjects were screened for glaucoma using a hand-held fundus camera. Optic disc images comprised 267 x 234 pixels at 24 bit/pixel (187 kByte). The images were compressed and stored together with patient information on a laptop computer. The images were then transmitted to Perth using either a mobile phone or a satellite phone, taking 170 s or 240 s, respectively. Images were also compressed to five different compression levels before transmission. At a compression ratio of 1:5, the images were 36 kByte in size and took 29 s to transmit by mobile phone and 60 s by satellite phone. To measure the loss of quality, the root mean square error was calculated for each colour component, comparing the transmitted and original images. The coefficients of variation were 10% (green), 15% (blue) and 22% (red). The images received in Perth were considered to be of excellent quality and readily interpreted by ophthalmologists in terms of the likely presence of glaucoma.     

医学成像设备图像分析系统的功能及质量控制

通过对医学图像分析系统的主要功能及医学图像分析的一般工作步骤的阐述,说明静态图像分析、动态图像分析及实时图像分析的方法。OLYMPUS荧光图像分析系统的质量控制方法在于医学图像分析系统使用过程中样品制备时对组织切片、载玻片、标本的质量要求。同时,在获取图像时应注意图像的存储格式,用TIFF格式保存图像;分析图像时,为了保证灰度值的可比性,应尽量使用同一批染色的切片,并在同一亮度光源下摄取图像。将分析测量信息存入电子资料表中,进行资料分析、统计及绘图。采用统计、直方图、曲线等输出方式,输出图像分析结果。     

医学图像压缩方式的探讨

目的探讨医学图像压缩的方式。方法根据CT、MRI图像的特点,采取不同于JPEG和JPEG2000的压缩方式。结果新的压缩方案可以很好地满足要求。结论应当根据不同医学图像的特点采取适当的压缩方法。     

Computer-assisted planimetry associated with Sturge-Weber syndrome

Photographs of the optic nerve head (optic disc) in the eye are used for the clinical assessment of the disease glaucoma. These images are usually subjectively assessed by a clinician. A case of Sturge-Weber Syndrome which includes glaucoma as a symptom, is presented here. Narrowing of the rim of the optic disc was measured using custom-made measurement software confirming glaucomatous progression. To maximize the information obtained from optic disc images, low cost software can assist with quantifying disc parameters aiding clinical interpretation.     

Realtime transfer of live video images in parallel with three-dimensional modelling of the surgical field in computer-assisted telesurgery

We have carried out three-dimensional, computer-assisted, functional endoscopic sinus telesurgery. Surgeons at different locations up to 300 km apart could not only see and transfer video images but also transfer three-dimensional computer models and manipulate them in realtime during surgery. Two different approaches were used. In the first telesurgery procedure we used M-JPEG compression and transmitted the data using fibre optic connections (ATM OC-3) at a bandwidth of 155 Mbit/s. In the second telesurgery procedure video images were transmitted over four E1 digital lines, amounting to about 8 Mbit/s of bandwidth, with better compression standards, such as MPEG1 and 2. We found that MPEG2 video compression produced the best picture quality for the operating field and endoscopic cameras. For conferencing and consultation between two or more connected sites during the surgery, we used JPEG and MPEG1 video compression with audio. The main feature of our three-dimensional telesurgery was the use of three-dimensional modelling of the operative field. This is important for emergency surgical interventions. We do not advocate that inexperienced surgeons operate on patients, not even with the guidance of a remote surgeon. However, three-dimensional telesurgery may become very valuable for experienced surgeons in the future.     

A preliminary study of the feasibility of wound telecare for the elderly.

We have developed a method for remote wound assessment in the elderly. Wound images were taken with a digital camera at a resolution of 640x480 pixels. JPEG compression was then used to produce images of about 100 kByte. Selected clinical data were transmitted by email, together with standardized digital images of wounds. The remote physician then read the clinical data and viewed the digital images on a 38 cm colour display monitor, at a resolution of 800x600 pixels, in 16-bit colour, using standard software. Three elderly inpatients with pressure sores or leg ulcers had both bedside and remote examinations, by different physicians. The diagnosis and therapeutic recommendations proposed after each of the two examinations were compared qualitatively. There was reasonable agreement between the two physicians in the assessment of wound size, anatomical classification, wound bed and status of infection. However, the lack of palpation represented a major limitation to remote wound assessment, despite the use of probes to delineate the depth of any opening in the wound bed.     

A simple telemedicine system using a digital camera

Radiographs on a viewing box were photographed at a remote hospital in South Africa using a digital camera with a resolution of 1024 x 768 pixels at 24-bit colour depth. The resultant images were stored in JPEG format and transmitted as email attachments to be read on a PC monitor by radiologists in Durban and Cape Town. Twenty-seven images were received, of which 23 were of diagnostic quality (85%). The mean file size was 120 kByte. For quality control purposes, 100 chest radiographs were photographed at a base hospital and read by a radiologist blinded to the diagnosis. In this study 96 images were of diagnostic quality (96%) and the correct diagnosis was made in 90 cases (94%). Incorrect readings were made in six cases (6%): small pulmonary nodules (less than 1 cm in diameter) were missed in five cases and in one case early apical tuberculosis was missed. Digital camera technology permits simple, inexpensive telemedicine. Limited spatial resolution is a concern when reading chest images with small pulmonary nodules and infiltrates.     

Tele-ophthalmology screening for retinal and anterior segment diseases

A digital indirect ophthalmoscope (DIO) was developed and tested for use in tele-ophthalmology screening for posterior and anterior segment diseases. Using custom software, images from the DIO were digitized, compressed, stored and transmitted to a centralized eye clinic for interpretation. A total of 43 subjects were primarily screened for glaucoma using the DIO, a hand-held fundus camera and a stereo fundus camera. The photographic slides from the stereo fundus camera used as gold standards. Images (390 x 300 pixels x 3 bits) were stored using a laptop computer together with patient information. Image quality received at the central eye clinic was good and showed adequate diagnostic information. An ophthalmologist estimated cup-disc ratios and graded the quality of the images. The sensitivity and specificity of each instrument was calculated. A high sensitivity and specificity was found when using the DIO, indicating that it could be used in tele-ophthalmic screenings. Further modifications are needed to make the instrument more user-friendly and to enable it to be used with undilated pupils, so that it can be easily operated by health-care personnel in remote areas. The camera can also be used to image gross external eye pathology.     

医学图像的无损压缩

JPEG、JPEG-LS和JPEG2000是ISO/IEC组织制定的三个静止图像压缩的国际标准。这三个标准虽采用的方法不同,但均能实现无损压缩。本文分别介绍了这三个静止图像压缩标准和编码过程,并对三种压缩标准在处理临床静止图像的无损压缩性能方面做了比较。     

基于小波变换的JPEG2000算法及其在医学图像压缩中的应用

医学图像数据量巨大，使用医学图像的不同人群对图像的分辨率有不同的需求。JPEG2000算法可以同时兼容无损压缩和有损压缩。本文描述了JEPG2000算法及其在医学图像压缩中的应用，并将压缩后的图像在图像的质量及大小两方面与原始图像作了比较，便于医务人员选用不同的压缩比得到适合的图像。     

CT图像经济性存储初步尝试

目的：通过对Dicom格式图像的刻,改变传统胶片存储CT图像的方法，实现CT图像存储的无胶片化。方法：利用SieNet Magic View300软件的功能，实现CT图像Dicom,Jpeg或Tiff格式存储。结果；此方法成功地实现了CT图像以Dicom,Jpeg或Tiff格式的刻录，图像质量合格，并能将图像返回CT主机进行再处理。结论：此方法投入少，效果好，图像质量符合Dicom标准，并节省了大笔的相关费用。     

JPEG2000及其在医学图像中的应用

JPEG2000是ISO／IEC组织制定最新的静止图像压缩国际标准。它不仅解决了已有的2个压缩标准JPEG和JPEG—LS存在的不足。而且还增加了很多新的功能，其中感兴趣区域编码、图像渐进式传输、运动（序列）图像压缩（Motion JPEG2000）、三维图像压缩（JP3D）以及图像安全性（JPSEC）都非常适合应用于医学图像。首先概要地介绍了JPEG2000所具有的新特性，然后详细地说明编码过程的预处理、离散小波变换、量化、第一层和第二层熵编码各个处理步骤，最后列举了医学图像中的几个典型应用。     

Development of a web-based picture archiving and communication system using satellite data communication

Using the JAVA language we have developed a Web-based picture archiving and communication system (PACS) which allows a remote hospital to access medical images. An asymmetric satellite data communication system (ASDCS) provided a receive-only link for data delivery and a conventional terrestrial link (which could be the conventional telephone network) allowed data transmission. The satellite communication link was 10-30 times faster than the conventional terrestrial link. To increase image transmission speeds over the Internet connection, JPEG and wavelet compression methods were used. The resulting images were evaluated quantitatively by measuring the peak signal:noise ratio and qualitatively by radiologists. Compression ratios of 10:1 or less were deemed acceptable for diagnostic purposes. The system appears to be suitable for teleradiology and telemedicine.