8-bit Depth Research Articles

Teledermatologic views at 16-bit color depth yield accurate diagnoses

Teledermatologic views at 16-bit color depth yield accurate diagnoses

Changes realized from extended bit‐depth and metal artifact reduction in CT

High-Z material in computed tomography (CT) yields metal artifacts that degrade image quality and may cause substantial errors in dose calculation. This study couples a metal artifact reduction (MAR) algorithm with enhanced 16-bit depth (vs standard 12-bit) to quantify potential gains in image quality and dosimetry. Extended CT to electron density (CT-ED) curves were derived from a tissue characterization phantom with titanium and stainless steel inserts scanned at 90-140 kVp for 12- and 16-bit reconstructions. MAR was applied to sinogram data (Brilliance BigBore CT scanner, Philips Healthcare, v.3.5). Monte Carlo simulation (MC-SIM) was performed on a simulated double hip prostheses case (Cerrobend rods embedded in a pelvic phantom) using BEAMnrc∕Dosxyz (400,000,0000 histories, 6X, 10 × 10 cm(2) beam traversing Cerrobend rod). A phantom study was also conducted using a stainless steel rod embedded in solid water, and dosimetric verification was performed with Gafchromic film analysis (absolute difference and gamma analysis, 2% dose and 2 mm distance to agreement) for plans calculated with Anisotropic Analytic Algorithm (AAA, Eclipse v11.0) to elucidate changes between 12- and 16-bit data. Three patients (bony metastases to the femur and humerus, and a prostate cancer case) with metal implants were reconstructed using both bit depths, with dose calculated using AAA and derived CT-ED curves. Planar dose distributions were assessed via matrix analyses and using gamma criteria of 2%∕2 mm. For 12-bit images, CT numbers for titanium and stainless steel saturated at 3071 Hounsfield units (HU), whereas for 16-bit depth, mean CT numbers were much larger (e.g., titanium and stainless steel yielded HU of 8066.5 ± 56.6 and 13,588.5 ± 198.8 for 16-bit uncorrected scans at 120 kVp, respectively). MC-SIM was well-matched between 12- and 16-bit images except downstream of the Cerrobend rod, where 16-bit dose was ∼6.4% greater than 12-bit. Absolute film dosimetry in a region downstream of a stainless steel rod revealed that 16-bit calculated dose, with and without MAR, agreed more closely with film results (1%-2% less than film) as compared to 12-bit reconstructions (5.6%-6.5% less than film measurements). Gamma analysis revealed that 16-bit dose calculations were better matched to film results than 12-bit (∼10% higher pass rates for 16-bit). Similar results were observed in two patient cases; the largest discrepancy was observed for a femur case where 12-bit doses, both with and without MAR correction, were 6-7 Gy lower (∼17%-20% of the prescription dose) as compared to 16-bit dose calculations. However, when beams are not directly traversing metal, such as a prostate cancer case with bilateral hip prostheses; the impact of 16-bit reconstruction was diminished. These results suggest that it may be desirable to implement 16-bit MAR-corrected images for treatment planning purposes, which can provide a more accurate dosimetric approach coupled with improved visualization by suppression of CT artifacts.

Reversible and High-Capacity Data Hiding in High Quality Medical Images

Via the Internet, the information infrastructure of modern health care has already established medical information systems to share electronic health records among patients and health care providers. Data hiding plays an important role to protect medical images. Because modern medical devices have improved, high resolutions of medical images are provided to detect early diseases. The high quality medical images are used to recognize complicated anatomical structures such as soft tissues, muscles, and internal organs to support diagnosis of diseases. For instance, 16-bit depth medical images will provide 65,536 discrete levels to show more details of anatomical structures. In general, the feature of low utilization rate of intensity in 16-bit depth will be utilized to handle overflow/underflow problem. Nowadays, most of data hiding algorithms are still experimenting on 8-bit depth medical images. We proposed a novel reversible data hiding scheme testing on 16-bit depth CT and MRI medical image. And the peak point and zero point of a histogram are applied to embed secret message k bits without salt-and-pepper.

A reversible data hiding method by histogram shifting in high quality medical images

Enormous demands for recognizing complicated anatomical structures in medical images have been demanded on high quality of medical image such as each pixel expressed by 16-bit depth. Now, most of data hiding algorithms are still applied in 8-bit depth medical images. We proposed a histogram shifting method for image reversible data hiding testing on high bit depth medical images. Among image local block pixels, we exploit the high correlation for smooth surface of anatomical structure in medical images. Thus, we apply a different value for each block of pixels to produce a difference histogram to embed secret bits. During data embedding stage, the image blocks are divided into two categories due to two corresponding embedding strategies. Via an inverse histogram shifting mechanism, the original image will be accurately recovered after the hidden data extraction. Due to requirements of medical images for data hiding, we proposed six criteria: (1) well-suited for high quality medical images, (2) without salt-and-pepper, (3) applicable to medical image with smooth surface, (4) well-suited sparse histogram of intensity levels, (5) free location map, (6) ability of adjusting data embedding capacity, PSNR and Inter-Slice PSNR. We proposed a data hiding methods satisfying above 6 criteria.

The Study of Data Hiding in Medical Images

Reversible data hiding plays an important role in medical image systems. Many hospitals have already applied the electronic medical information in healthcare systems. Reversible data hiding is one of the feasible methodologies to protect the individual privacy and confidential information. With application in several high quality medical devices, the detection rate of diseases and treating are improved at the early stage. Its demands have been rising for recognizing complicated anatomical structures in high quality images. However, most data hiding methods are still applied in 8-bit depth medical images with 255 intensity levels. This paper summarizes the existing reversible data hiding algorithms and introduces basic knowledge in medical image.

A Joint Encryption/Watermarking System for Verifying the Reliability of Medical Images

In this paper, we propose a joint encryption/water-marking system for the purpose of protecting medical images. This system is based on an approach which combines a substitutive watermarking algorithm, the quantization index modulation, with an encryption algorithm: a stream cipher algorithm (e.g., the RC4) or a block cipher algorithm (e.g., the AES in cipher block chaining (CBC) mode of operation). Our objective is to give access to the outcomes of the image integrity and of its origin even though the image is stored encrypted. If watermarking and encryption are conducted jointly at the protection stage, watermark extraction and decryption can be applied independently. The security analysis of our scheme and experimental results achieved on 8-bit depth ultrasound images as well as on 16-bit encoded positron emission tomography images demonstrate the capability of our system to securely make available security attributes in both spatial and encrypted domains while minimizing image distortion. Furthermore, by making use of the AES block cipher in CBC mode, the proposed system is compliant with or transparent to the DICOM standard.

Shadow Correction in High Dynamic Range Images by Aerial Digital camera

With the widespread use of aerial digital cameras such as the DMC, ADS40, RMK-D, and UltraCamD, high dynamic range imaging is generally expected for generating minuteness orthophotos in digital aerial photogrammetry. However, high dynamic range images (12-bit, 4,096 gray levels) are generally compressed into an 8-bit depth digital image (256 gray levels) owing to huge amount of data and interface with peripherals such as monitors and printers. This means that a great deal of image data is eliminated from the original image, and this introduces a new shadow problem. In particular, the influence of shadows in urban areas causes serious problems for generating minuteness orthophotos and performing house detection. In order to resolve the shadow problem in digital aerial photogrammetry, shadow areas should be recognized and corrected automatically without the loss of luminance information. With this motive, a practical shadow correction method for 8-bit depth digital image which is derived from 12-bit real data of DMC is investigated in this paper.

MULTIFRACTAL ANALYSES OF GRAYSCALE AND BINARY SOIL THIN SECTION IMAGES

Multifractal analyses of binary images of soil thin sections (STS) are widely used to characterize pore structure. However, no geometrical model is known to exist for a binary multifractal. Thus, the multifractality of binary images, and the accuracy of multifractal parameters estimated from them, need to be carefully evaluated. We captured 8-bit depth resolution digital grayscale images of three STS images with dimensions of 1024 × 1024 pixels and a pixel length of 1.9 μm. Random grayscale geometrical multifractal fields (GMF) with similar dimensions and known multifractal parameters were constructed using generators extracted from the STS images. The STS and GMF grayscale images were objectively thresholded to give six binary images. The method of moments was used to compute the log-transformed partition function, log (χ(q, δ)) versus log(δ) where δ is box size, for each grayscale image and its binary counterpart. Consistent linearity was observed in the resulting functions for the grayscale images, indicating, by definition, multifractal behavior. In contrast, the log (χ(q, δ)) versus log(δ) plots for the binary images exhibited a two-region response, with a flat plateau at small scales and linearity at larger scales, indicating they were not true multifractals. Generalized dimensions (Dq) computed from the linear portions of the binary log-transformed partition functions were significantly over estimated for q ≪ 0 and underestimated for q ≫ 0 relative to corresponding Dq values for the grayscale images. Based on these results we contend that binary images are not mathematical multifractals, and that generalized dimensions estimated from them cannot be used to quantify pore space geometry. Instead we encourage further exploration of the use of grayscale images for multifractal characterization of soil structure. This direct approach is theoretically sound and does not require any intermediate thresholding step, which is known to influence the results of multifractal analyses.

Live, Video-Rate Super-Resolution Microscopy Using Structured Illumination and Rapid GPU-Based Parallel Processing

Structured illumination fluorescence microscopy is a powerful super-resolution method that is capable of achieving a resolution below 100 nm. Each super-resolution image is computationally constructed from a set of differentially illuminated images. However, real-time application of structured illumination microscopy (SIM) has generally been limited due to the computational overhead needed to generate super-resolution images. Here, we have developed a real-time SIM system that incorporates graphic processing unit (GPU) based in-line parallel processing of raw/differentially illuminated images. By using GPU processing, the system has achieved a 90-fold increase in processing speed compared to performing equivalent operations on a multiprocessor computer--the total throughput of the system is limited by data acquisition speed, but not by image processing. Overall, more than 350 raw images (16-bit depth, 512 × 512 pixels) can be processed per second, resulting in a maximum frame rate of 39 super-resolution images per second. This ultrafast processing capability is used to provide immediate feedback of super-resolution images for real-time display. These developments are increasing the potential for sophisticated super-resolution imaging applications.

Performance comparison between 8- and 14-bit-depth imaging in polarization-sensitive swept-source optical coherence tomography

Recently the effects of reduced bit-depth acquisition on swept-source optical coherence tomography (SS-OCT) image quality have been evaluated by using simulations and empirical studies, showing that image acquisition at 8-bit depth allows high system sensitivity with only a minimal drop in the signal-to-noise ratio compared to higher bit-depth systems. However, in these studies the 8-bit data is actually 12- or 14-bit ADC data numerically truncated to 8 bits. In practice, a native 8-bit ADC could actually possess a true bit resolution lower than this due to the electronic jitter in the converter etc. We compare true 8- and 14-bit-depth imaging of SS-OCT and polarization-sensitive SS-OCT (PS-SS-OCT) by using two hardware-synchronized high-speed data acquisition (DAQ) boards. The two DAQ boards read exactly the same imaging data for comparison. The measured system sensitivity at 8-bit depth is comparable to that for 14-bit acquisition when using the more sensitive of the available full analog input voltage ranges of the ADC. Ex-vivo structural and birefringence images of equine tendon indicate no significant differences between images acquired by the two DAQ boards suggesting that 8-bit DAQ boards can be employed to increase imaging speeds and reduce storage in clinical SS-OCT/PS-SS-OCT systems. One possible disadvantage is a reduced imaging dynamic range which can manifest itself as an increase in image artifacts due to strong Fresnel reflection.

4K-UHD 비디오 시청환경 특성분석을 위한 주관적 화질평가 분석

본 논문에서는 초고해상도(UHD: Ultra High Definition) 영상의 방송 서비스를 위한 UHD 영상의 주관적 화질 평가를 수행하고 이에 따른 화질평가 결과를 분석하였다. TV, 인터넷, 개인 미디어 기기들의 보급으로 인해 영상 콘텐츠의 소비가 늘어나고 있으며 이와 더불어 고화질 영상에 대한 수요 또한 증가하고 있다. 현재 HD급(<TEX>$1920{\times}1080$</TEX>) 영상이 DTV, DVD, 디지털 캠코더, 감시 비디오 및 기타 멀티미디어 기기 등을 통하여 다양한 형태로 소비되고 있으며, 최근에는 이러한 HD 해상도를 넘어 4K-UHD(<TEX>$3840{\times}2160$</TEX>) 해상도의 디지털 시네마 콘텐츠가 활발히 제작되고 있고, 멀티미디어 시장에서도 이를 지원하는 카메라, 빔 프로젝터, 디스플레이 등의 기기가 출시되기 시작하였으며, 곧 방송<TEX>${\times}$</TEX>통신 환경에서도 4K-UHD 비디오 서비스가 출현할 것으로 예상된다. 이에 본 논문에서는 UHD TV 서비스 도입 시에 UHD 비디오의 시청환경 및 신호규격을 결정하는데 도움을 주기 위해 4K-UHD 영상에 대해 공간 해상도, 컬러포맷, 프레임률, 압축률을 포함한 4 가지 항목에 대한 주관적 화질 평가 실험을 수행하였다. 평가 결과, HD 해상도 대비 UHD 해상도의 영상에 대해 더 높은 주관적 화질 평가 결과를 보였으며, 컬러포맷이 YUV444인 4K-UHD 영상이 YUV422, YUV420인 4K-UHD 영상들 보다 화질이 더 높은 주관적 평가를 받았고, YUV422와 YUV420 4K-UHD 영상 간의 화질 비교에서는 주관적 평가 차이는 미미한 것으로 나타났다. 프레임률에 있어서는 60fps의 4K-UHD 영상이 30fps의 4K-UHD 영상보다 움직임이 많은 영상일수록 더 높은 주관적 화질 평가 결과를 보였으며, HD 영상에 대한 비트심도 비교에서는 10bit인 영상과 8bit인 영상 간의 주관적 화질 평가차이는 매우 작은 것으로 나타났다. 마지막으로 복<TEX>${\times}$</TEX>부호화된 4K-UHD 영상의 PSNR이 높을수록 높은 주관적 화질평가를 받았으며 시청거리가 멀어질수록 부호화에 의한 화질 열화를 인지하는 능력이 떨어지면서 PSNR이 낮은 영상도 화질이 양호한 것으로 평가되는 경향이 있었다. In this paper, we perform subjective visual quality assessments on UHD video for UHD TV services and analyze the assessment results. Demands for video services have been increased with availabilities of DTV, Internet and personal media equipments. With this trend, the demands for high definition video have also been increasing. Currently, 2K-HD (<TEX>$1920{\times}1080$</TEX>) video have been widely consumed over DTV, DVD, digital camcoders, security cameras and other multimedia terminals in various types, and recently digital cinema contents of 4K-UHD(<TEX>$3840{\times}2160$</TEX>) have been popularly produced and the cameras, beam projects, display panels that support for 4K-UHD video start to come out into multimedia markets. Also it is expected that 4K-UHD service will appear soon in broadcasting and telecommunications environments. Therefore, in this paper, subjective assessments of visual quality on resolutions, color formats, frame rates and compression rates have been carried to provide basis information for standardization of signal specification of UHD video and viewing environments for future UHDTV. As the analysis on the assessments, UHD video exhibits better subjective visual quality than HD by the evaluators. Also, the 4K-UHD test sequences in YUV444 shows better subjective visual quality than the 4K-UHD test sequences in YUV422 and YUV420, but there is little perceptual difference on 4K-UHD test sequences between YUV422 and YUV420 formats. For the comparison between different frame rates, 4K-UHD test sequences of 60fps gives better subjective visual quality than those of 30fps. For bit-depth comparison, HD test sequences in 10-bit depth were little differentiated from those in 8-bit depth in subject visual quality assessment. Lastly, the larger the PSNR values of the reconstructed 4K-UHD test sequences are, the higher the subjective visual quality is. Against the viewing distances, the differences among encoded 4K-UHD test sequences were less distinguished in longer distances from the display.

Performance of reduced bit-depth acquisition for optical frequency domain imaging

High-speed optical frequency domain imaging (OFDI) has enabled practical wide-field microscopic imaging in the biological laboratory and clinical medicine. The imaging speed of OFDI, and therefore the field of view, of current systems is limited by the rate at which data can be digitized and archived rather than the system sensitivity or laser performance. One solution to this bottleneck is to natively digitize OFDI signals at reduced bit depths, e.g., at 8-bit depth rather than the conventional 12-14 bit depth, thereby reducing overall bandwidth. However, the implications of reduced bit-depth acquisition on image quality have not been studied. In this paper, we use simulations and empirical studies to evaluate the effects of reduced depth acquisition on OFDI image quality. We show that image acquisition at 8-bit depth allows high system sensitivity with only a minimal drop in the signal-to-noise ratio compared to higher bit-depth systems. Images of a human coronary artery acquired in vivo at 8-bit depth are presented and compared with images at higher bit-depth acquisition.

Independent two-photon measurements of albumin GSC give low values

Independent two-photon measurements of albumin GSC give low values

The scope for a rapid digitiser of mammograms and other large-file, high-resolution images

A new high-speed, high-resolution digitiser has provided a facility with which new means of communication and interaction become possible. 1. The DicomNow Digitiser camera system will scan mammography film at a 16-bit depth in less than 1 second. 2. The thresholds for compression of mammogram images have previously been confirmed at a ratio between 65:1 and 80:1. 3. New opportunities for transmission between centres; difficult analogue films can be digitised quickly, then compressed and sent for consensus/arbitration views from an expert panel or submitted to computer-aided diagnosis. 4. Applications to multidisciplinary team meetings in which all types of images, patient information and pathology are shared via any Internet link in a virtual meeting room. 5. The speed of DicomNow and the compression system removes time limitations for telemedicine communication speeds in mammography. 6. Archive potential to reduce storage cost and support disaster-recovery initiatives. 7. Modular training programmes will benefit students with access to high-resolution images via the scanning and compression technology. 8. Higher level studies of subtle features of image interpretation for validated case studies. Each of these elements will be expanded to illustrate the effects of high-speed digitising on breast cancer management.

Definition of Compression Ratio: Difference Between Two Commercial JPEG2000 Program Libraries

The objective was to demonstrate the difference in the definition of compression ratio between two popular commercial JPEG 2000 program libraries. An institutional review board approved this study and waived informed consent. Using each of two JPEG 2000 libraries (libraries A and B), 20 abdomen computed tomography images with 12-bit depth (from scanner 1) and 20 images with 16-bit depth (from scanner 2) were compressed to three different nominal compression ratios: 10:1, 15:1, and 20:1. Achieved compression ratios (the original image file size to the compressed size) were compared with the nominal compression ratios using one-sample t-test tests. At each nominal compression level, the achieved compression ratios for scanner 1 images compressed using library A were approximately 1.33-fold greater than the nominal compression ratio (p < 0.0001), while the achieved compression ratios for the remaining three scanner-library combinations (scanner 1-library B, scanner 2-library A, and scanner 2-library B) were approximately the same as the nominal compression ratio (p-value range, 0.22-0.93). The definition of compression ratio is different between commercial JPEG 2000 program libraries. The definition should be standardized to facilitate the adoption and communication of an acceptable compression level.

Real-time implementation of multiview image synthesis system

In this paper, we generate a virtual right image corresponding to an input left image by using given RGB texture data and 8-bit gray-scale depth data from a depth camera. First we transform the depth data to disparity data. Then, using this disparity data, we produce a virtual right image. Further, we propose a real-time multiple-view synthesis algorithm using geometric transformations, such as rotation and translation after the stereo image synthesis, and also a fast LUT (look-up table) method for real-time processing through analyzing the correlation between the depth data and the disparity data. As a result, we are able to achieve real-time synthesis of 11-viewpoint stereo images using only RGB texture data and the 8-bit depth information.

The Glucose Transporter 4-regulating Protein TUG Is Essential for Highly Insulin-responsive Glucose Uptake in 3T3-L1 Adipocytes

Insulin stimulates glucose uptake in fat and muscle by redistributing GLUT4 glucose transporters from intracellular membranes to the cell surface. We previously proposed that, in 3T3-L1 adipocytes, TUG retains GLUT4 within unstimulated cells and insulin mobilizes this retained GLUT4 by stimulating its dissociation from TUG. Yet the relative importance of this action in the overall control of glucose uptake remains uncertain. Here we report that transient, small interfering RNA-mediated depletion of TUG causes GLUT4 translocation and enhances glucose uptake in unstimulated 3T3-L1 adipocytes, similar to insulin. Stable TUG depletion or expression of a dominant negative fragment likewise stimulates GLUT4 redistribution and glucose uptake, and insulin causes a 2-fold further increase. Microscopy shows that TUG governs the accumulation of GLUT4 in perinuclear membranes distinct from endosomes and indicates that it is this pool of GLUT4 that is mobilized by TUG disruption. Interestingly, in addition to translocating GLUT4 and enhancing glucose uptake, TUG disruption appears to accelerate the degradation of GLUT4 in lysosomes. Finally, we find that TUG binds directly and specifically to a large intracellular loop in GLUT4. Together, these findings demonstrate that TUG is required to retain GLUT4 intracellularly in 3T3-L1 adipocytes in the absence of insulin and further implicate the insulin-stimulated dissociation of TUG and GLUT4 as an important action by which insulin stimulates glucose uptake.

Calibration of a Soft X-Ray Digital Imaging System for Biological Materials

Soft X-ray imaging of agricultural products for quality determination is gaining worldwide interest. Recent advances in technology have made it possible to capture, store, and instantaneously process digital X-ray images at higher resolution for lower cost. An X-ray imaging system was developed and calibrated to capture images with a resolution of 1024 × 1024 pixels over a 50 × 50 mm area. Maximum X-ray tube voltage was 50 kVp, maximum current was 1 mA, and signal integration time ranged from 460 to 6700 ms. The system consists of an X-ray source tube, a camera composed of a CMOS photodiode array, a frame grabber, and a data acquisition and control card. Because the X-ray beam is polychromatic, and the detector responses are variable, calibration is necessary to relate image intensity to X-ray attenuation. Imprecision of the system was 0.64% of the response range (1.63 gray level for an 8-bit pixel depth). Response, measured as mean pixel intensity, varied linearly with X-ray tube current and integration time and had a quadratic relationship with peak tube voltage. A regression model was developed to estimate blank image intensity at higher voltages and currents. Prediction and validation errors for the model were 0.46% and 1.06% (1.18 and 2.72 gray level), respectively. Beam hardening effect was demonstrated using a polystyrene target. The procedure explained in this article can be used to calibrate a soft X-ray imaging system for radiometric measurements.

On algorithmic rate-coded AER generation

This paper addresses the problem of converting a conventional video stream based on sequences of frames into the spike event-based representation known as the address-event-representation (AER). In this paper we concentrate on rate-coded AER. The problem is addressed as an algorithmic problem, in which different methods are proposed, implemented and tested through software algorithms. The proposed algorithms are comparatively evaluated according to different criteria. Emphasis is put on the potential of such algorithms for a) doing the frame-based to event-based representation in real time, and b) that the resulting event streams ressemble as much as possible those generated naturally by rate-coded address-event VLSI chips, such as silicon AER retinae. It is found that simple and straightforward algorithms tend to have high potential for real time but produce event distributions that differ considerably from those obtained in AER VLSI chips. On the other hand, sophisticated algorithms that yield better event distributions are not efficient for real time operations. The methods based on linear-feedback-shift-register (LFSR) pseudorandom number generation is a good compromise, which is feasible for real time and yield reasonably well distributed events in time. Our software experiments, on a 1.6-GHz Pentium IV, show that at 50% AER bus load the proposed algorithms require between 0.011 and 1.14 ms per 8 bit-pixel per frame. One of the proposed LFSR methods is implemented in real time hardware using a prototyping board that includes a VirtexE 300 FPGA. The demonstration hardware is capable of transforming frames of 64 x 64 pixels of 8-bit depth at a frame rate of 25 frames per second, producing spike events at a peak rate of 10(7) events per second.

Quantitative analysis of retinopathy in type 2 diabetes: identification of prognostic parameters for developing visual loss secondary to diabetic maculopathy

To describe whether quantitative assessment of early changes in the morphology of retinopathy lesions can predict development of vision-threatening diabetic maculopathy. We used a nested case-control study, and we studied 11 type 2 diabetes patients who had developed visual loss secondary to diabetic maculopathy. For each diabetes patient, we also studied three matched control patients who had been followed for a comparable period of time without developing visual loss. Fundus photographs describing the early development of retinopathy were digitized and subjected to a full manual quantitative grading on a computer monitor. Differences in the early development of retinal morphology were compared between the two groups. The outcome parameters were changes in the number and area of haemorrhages and exudates in different regions of the fundus, and the weighted distance of these lesions from the fovea and the optic disc. In patients who developed visual loss secondary to diabetic maculopathy there was significant early progression in the total area and number of haemorrhages and exudates. The haemorrhages had progressed in all retinal areas except the area around the optic disc and the temporal vascular arcades. The exudates had progressed temporally from the fovea and in the retinal periphery. The results suggest that a quantitative description of the regional development of early diabetic retinopathy may help in identifying patients who will later develop vision-threatening maculopathy.