What Do We See Best

In diagnostic images such as mammograms, incredibly small and low-contrast structures such as microcalcifications have to be appreciated.

f A structure is perceived best if it is viewed at a distance i at which it subtends an angle of 5° (Fig. 3.7a).

We implement this physiological truth in our daily work by getting closer to an image if we look for very small de-tails—until we hit the limits of accommodation. Accommodation is, of course, phenomenal in children (if they show you a piece of paper to read, they will hold it close to your nose). Radiologists of adequate biological age, however, need to carry a magnifying glass around to compensate for their accommodation deficit. It is interesting to note that the phenomenon also works the other way around: large, low-contrast lesions are better perceived if you view them with a minifying glass. The number of radiologists carrying these around is much smaller, however. The normal crowd just takes a step back and then another look.

How well we view small structures also depends on the brightness, or optical density as the physicists call it. As typical animals of the steppe we see contrasts best at the brightness of a summer afternoon. At this optical density our optical cones work optimally while the intraocular scatter is minimal. If it becomes darker, we turn to rod vision. The detectability decreases significantly, which is why you a have a reading lamp at your bedside. The windows of a typical ward room are only usable as "light boxes" if main rounds take place on a summer afternoon and if clear skies dominate in the region—thus in hospitals for the upper few. Late in the evening, however, a broader public can also check the healing progress while standing in the parking lot.

As our eye adapts to the brightness of the total visual field, an image on the light box should be very well masked if you do not want to miss low-contrast lesions. Room (ambient) light must also be adapted to prevent reflections on the image, and the dilation of the pupils, which would increase intraocular scatter and activation of rod vision. The proper examination of an image that has been generated with great care is thus far from trivial (Fig. 3.7b). Special computer-assisted light boxes or viewing monitors that also control room light optimize our perception (Fig. 3.7c).

I Perception

Room lighting

Intraocular scatter

Light box

Intraocular scatter

Light box

Glare due —^ to room light

Reflection of room light


Clare due to light box

Light box

a Glare and reflections in the visual field, intraocular scatter, and insufficient masking of the image on the light box are reasons for impaired perception (upper). If these mistakes are corrected and if structures are viewed so that they subtend an angle of 5° perception is optimal (lower). b Chief's round on the surgical ward. The films were fixed to the windows. Obviously rounds have to take place before sundown on this ward. Nice to know also that the therapeutic progress can be checked another time late at night while walking over to the parking lot. c This is a computer-assisted lightbox that masks the film and adapts ambient light automatically, thus optimizing perception (by Smartlight, Inc.). d This is a modern flat panel display used in digital mammography. In this special field the quality requirements are extremely high. The display also needs to be placed in dedicated, light-adjusted surroundings (by Fuji Medical Systems). ►

What Else Influences Our Perception?

Even if a structure is optically well discernible it must also be perceived and, eventually, be evaluated and classified. Is the structure pathological, normal, or just a variant of the normal? The image is scanned optically and compared to an internalized standard, the "gestalt" of, for example, a chest radiograph. The more complex the normal image, i.e., the radiological anatomy, the more difficult is the detection of pathology. A given nodule is readily perceived in the periphery of the lung but is easily missed when it is close to the lung hilum, because the large vessels can look like or camouflage the nodule. The negative influence of complex anatomy on the detectability of pathology is also called "anatomical noise" (see p. 21).

After you have detected a definitely pathological structure, your attention may fade, especially ifyou are still inexperienced. This "satisfaction of search" effect (see p. 25) hinders the careful examination of the rest of the image. Make it a point to take the time for a second thorough pass over the image!

The independent review of an examination by another colleague ("double reading") can increase its diagnostic value significantly (see p. 32)—four eyes do see more than two!

f Images are generated with extremely expensive and soi phisticated equipment, sometimes with substantial risk to the patient. They must be studied with great care and under optimal conditions.

Blood Vessels The Eye Zebrafish
0 0

Post a comment