Spotting of disc haemorrhages on fundus pictures; on Kodachrome slides and on digitally enhanced images.

PURPOSE: To find out whether digital processing can facilitate interpretation of fundus photographs. METHODS: Digitised fundus pictures were manipulated and enhanced by a process called histogram equalisation. Essentially, the gamma values for red (R), green (G), and blue (B) were separately modified. The principle is explained and exemplified. MATERIAL AND RESULTS: Small disc haemorrhages, which may normally be hard to detect, became more visible after histogram equalisation, which is demonstrated in test series. CONCLUSION: Digital histogram equalisation offers a new way of looking at fundus pictures, which, alone or combined with other capabilities of digital processing, has great potential for diagnosis, as well as for follow up studies. The possibility to decide the true nature of suspect spots after applying the enhancement procedure is likewise of great value.

A computerised adaptometer.

PURPOSE: The design of an automatic instrument for the estimation of subjective dark adaptation (minimum perceptible for light) by an uninterrupted chain of trials. METHODS: A light emitting diode is used as test light. The patient answers by pressing a button when light is perceived. The process is controlled by means of a lap top computer and traces the adaptation course step by step. RESULTS: The instrument was tested by recording the course of dark adaptation (DA) in a clinical series of normals and of patients. The threshold sensitivity (after 25 min in darkness) showed very satisfying reproducibility. CONCLUSION: The instrument described was found to be reliable and well suited for clinical routine as well as experimental work.

An instrument for recording the ocular pulse wave.

An instrument for recording the ocular pressure pulse is described. Examples of recordings showing individual differences in the shape of the pulse curve are demonstrated. The properties of the instrument as to the frequency response in the range 0.5-10 Hz are established with the aim of correcting the recordings for instrumental errors.

A model for pulsatile and steady ocular blood flow.

BACKGROUND: The aim was to elucidate two controversial questions. (1) is the pulsatile inflow dominant or is there a sizeable steady inflow component? (2) Is the outflow of blood steady or pulsatile in induced and in glaucomatous hypertension? METHODS: The OBF system (Langham) was used for pulse recording. A simple electric flow model was conceived and analysed in terms of Fourier series, using linear circuit theory. RESULTS AND CONCLUSIONS: Possible explanations of the seemingly inconsistent observations are indicated by the model. (1) A steady inflow of similar magnitude to the pulsatile one is predicted. The total flow can be determined at present theoretically from recordings of the pulsatile inflow and of the systolic and diastolic intraocular blood pressure. Some methodological improvements are desirable in order to obtain reliable estimates in practice. (2) The outflow is normally steady, probably even in ocular hypertension and glaucoma. In hypertension induced with the suction cup this may be doubted. The eye has similarities with the device called the Starling resistor which describes the behaviour of collapsible vessels. Acute pressure increase (plus the pulsations of the IOP) may put the collapsible vessels into a state which disturbs the steady outflow. The effects of changes in IOP and in other components on the size and shape of the flow curve are demonstrated.

Disk hemorrhages and retinal vein occlusions in glaucoma.

There are several reasons to consider disk hemorrhages, branch retinal vein occlusions and central retinal vein occlusion as manifestations of the same vascular disease, the only difference among them being the size of the vessel affected. There is a close association of these vascular events with open angle glaucoma, and all of them increase with increasing follow-up time. The morphological changes described in the retinal veins in glaucoma and in central vein occlusions are endothelial proliferations causing progressive increase of flow resistance. Similar changes most probably also cause branch vein occlusions and disk hemorrhages. This vascular condition which obstructs the vessels with hindrance of the blood flow and impaired nutrition of neuronal tissue might be the primary cause of glaucoma.

Arguments for a vascular glaucoma etiology.

A material of Open Angle Glaucoma (G) collected by an eye practitioner (B S) from 1980 to 1991 is analysed with special respect to the occurrence of disc haemorrhages (Hs) and retinal vein occlusions (VOs). There is a close association between Hs and VOs and G, which in all categories increases with increasing follow-up time. Furthermore, the occurrence of mixed cases--H in one eye and VO in the same or the other--support the opinion that Hs, branch vein occlusions and central vein occlusions are manifestations of the same vascular disease, the difference between them being the size of the vessel affected. The well known morphological changes in the retinal veins of glaucomas and in central vein occlusions are endothelial proliferations causing progressive increase of flow resistance. Similar changes most probably also cause branch occlusions and disc haemorrhages. The vascular affection which obstructs the vessels with hindrance of the blood flow and impaired nutrition of neuronal tissue, might be the primary cause of glaucoma.

Measurement of progressive disc change in glaucoma.

We describe a videographic procedure for detecting minute changes in disc topography of glaucomatous eyes that involves measuring the displacement of selected circumlinear vessels. The problem of recognizing the difference between parallactic errors and true changes in topography is emphasized.

Calculation of the pulsatile ocular blood flow.

Based on reasonable assumptions, it is theoretically possible to determine the pulsatile ocular blood flow. The pertinent parameter is not the pulse amplitude but the steepest decay value of record (ie, the minimum of dV(t)/dt). Some advantages of developing the record in Fourier series are discussed. The importance of careful analysis of the recording system is stressed.

Correction of optic disc measurements on fundus photographs.

If defined as diameter of image/diameter of object, the magnification M of the eye-camera system is kB, where k is a camera constant and B represents the vergence of the internal axis of the eye. The magnification may be assessed (1) from the length l of the eye, using the formula 1.336 k/(l-0.0016); (2) from the glass refraction G, using the formula kD/(1-G/D), in which D is the "normal" refractive power of the eye; or (3) from the principal point refraction A and the deviation delta D1 from the "normal" refractive power of the cornea, using the formula kD.(1 + delta B/D), where delta B = A + 0.84 delta D1. If k is 0.042 m, as in the Zeiss camera, and D is about 60 dpt, the formulas may be written as (1) M = 0.056/(l-0.0016); (2) M = 2.5/(1-0.017 G); and (3) M = 2.5.(1 + 0.017 delta B).

Correlations in series of perimetric performance values.

The correlation function for consecutive performance (P) values in a material of simple glaucoma is studied. The outcome of perimetry sessions recorded with about 2 min intervals are significantly correlated, whereas those obtained 2 h apart or more are not so. Partial correlations of the right and left P values as well as IOP and P value are estimated and discussed in connection with a statistical model for the sensitivity fluctuations. A slow trend, a random error component and an irregular undulation of sensitivity are discerned. The cause of the fluctuations as well as procedures for reducing the dispersion of performance values are discussed.

Computerized patient record system for the glaucoma ward.

A system for computerized glaucoma patient record, developed from a data base called 'Advanced Revelation', is described. The record is made up from free text as well as from choices among listed alternatives. All data are stored on magnetic media and also printed out on paper in a well disposed and easily surveyable form. Apart from saving personnel and time in the clinical routine, the main advantages of the system is a powerful tool for clinical research. Data retrieval occurs on 'list', 'count', or 'sort' orders, facilitating retrospective studies. The system has been used for over a year at a glaucoma ward (Lund) to increasing satisfaction.

High frequency of asymptomatic visual field defects in subjects with transient ischaemic attacks or minor strokes.

The hypothesis that asymptomatic visual field defects can be found in patients with carotid transient ischaemic attacks (TIA) or minor strokes was tested. Twenty-two consecutive male patients with TIA and 18 patients with minor strokes from the carotid artery territory were examined by perimetry, cerebral computerised tomography and regional cerebral blood flow. Asymptomatic visual field defects were found in many TIA and minor stroke patients, 29% (5/17) and 57% (8/14), respectively (NS). Eighty-five per cent (11/13) of the scotomas were solely or predominantly located in the upper part of the visual field (P = 0.008 for absolute defects and P = 0.03 for relative defects). We conclude that both carotid territory TIA and minor stroke patients have a high frequency of asymptomatic visual field defects, predominantly located in the upper part of the visual field.

Occurrence of disc haemorrhages in open-angle glaucoma treated with pilocarpine or timolol.

1573 examinations were performed on 397 eyes. hh was found in 55 eyes--thromboses included--at 91 examinations. Part of this material qualified for use in a comparative study with pilocarpine in 81 eyes and timolol in 82 eyes. The relative frequency of hh in the treated groups and in the pretreatment group was not significantly different from that of the whole material. No indication of pressure reduction with effect on the occurrence of hh could be detected. The average probability of hh at an examination chosen at random was 6 per cent in this material of ordinary simple and capsular glaucoma cases. The occurrence of haemorrhages in relation to age, sex, general hypertension and diabetes are discussed.

Visual field testing with reduced sets of test points. A computerized analysis.

A computerized analysis of visual fields is described. In a second step these results are condensed into a 'reduced' set of test points so constructed that if testing is limited to 6-8 points nearly all defective visual fields would have been spotted with at least one defective point. The 'reduced' set involves the testing of few points several times in a session rather than spending very few trials on a great number of points. This is a new approach which merits further investigation.

Observations concerning the course of glaucoma.

Four cases with glaucomatous damage to the optic nerve head preceding a rise in intraocular pressure from normal or borderline to high values, are reported. The question of whether such cases are exceptional or typical and, therefore, best explained by a coincidence of diseases or by one disease going through successive stages is discussed on the basis of 10 further cases from a population study. Many clinicians claim that there are two groups of patients: a large one with 'ordinary' glaucomas supposedly caused by high intraocular pressures and a small one with low tension glaucomas presumed to have low perfusion pressures for other reasons. This classification circumvents the discussion on etiology and gives theoretical support to conventional treatment. Accumulated clinical observations indicate, however, that it should not be accepted without discussion. We describe here a few selected cases with glaucomatous damage to the optic nerve head preceding a rise in the intraocular pressure from normal or borderline to high values.

Temporal summation and perimetry.

The threshold sensitivity at perimetry was determined experimentally as a function of the exposure time of the test objects. The summation capability of the retina is expressed as a 'memory' function with approximately exponential decay. The efficient stimulus is found from the convolution integral of the functions for memory and light stimulus. The effacing of relative defects by a prolonged exposure time is explained by means of a hypothetical model, based on the rarefaction of optic-nerve fibres.

Treatment versus no treatment in chronic open angle glaucoma.

In a controlled randomized study 15 patients (20 eyes) with chronic open angle glaucoma and visual field defects were followed by greater than 1 year, 12 of them were followed for 3 years. Half of the group were untreated controls, the other half treated with pressure reducing medical therapy. At least 5 consecutive computerized visual fields were recorded (COMPETER) on each eye, and the linear regression coefficient was calculated. With the reservation for uncontrolled compliance no significant difference in the line of favourable effect of pressure reduction could be spotted, in spite of an average pressure reduction in the treated group of 4 mmHg. More important than this result, which is open to criticism for the smallness of the material, uncontrolled compliance etc, is the lesson that a randomized experiment with treated and non-treated glaucoma cases carried out in accordance with the Helsingfors convention is hardly feasible.

Disc haemorrhages and glaucoma in a general ophthalmic practice.

The composition of the clientele from an ophthalmological practitioner's office is described with special reference to the occurrence of glaucoma and disc haemorrhages (h in singular; hh in plural). This study could not be planned as an epidemiological survey and gives no clue to sensitivity or specificity of hh in glaucoma. During a period of about 10 years ending with 1986 there were 731 patients with h and/or glaucoma. When detected, 185 patients had h but no glaucoma, 33 had both h and glaucoma and 513 had glaucoma but no h. During the follow-up period hh were detected in 83 cases of glaucoma, and glaucoma developed in 27 cases with hh. The detection rate of hh among glaucoma was low but steady, indicating that hh may occur at any stage of the glaucoma process. This study shows no predilection for hh in cases with general hypertension or diabetes, nor is the frequency of hh among pseudoexfoliation cases significantly lower than among cases without this stigma.