The Bergmann Foot Scanner and automated design system has two distinct functions. One function is the ability to

capture the foot and change it into a digitized form that can be evaluated, shown on the computer and transferred

to the orthotic lab over the phone line, reducing shipping time. The second function is the ability for the doctor or

lab to precisely manipulate and change the shape of this digitized data to form more accurate expansions,

platforms, or other corrections so that an orthotic fabricated using this digitized image will be both more

comfortable and effective in controlling the patient's biomechanical condition. This was the original purpose for

developing this technology. However, as the technology evolved many enhancements were added that have made

this system more than just a method to cast and manufacture orthotics. It has become a system that helps to

explain the basics of biomechanics to patients and others in a clear and concise way. It has also evolved into a

research tool that can be used to expand and verify our knowledge about orthotics and biomechanics.

    Although the technology is very sophisticated, we have tried to make this equipment easy to use with no

preconceived method of using it. You do not need to be computer literate to run it. We felt this philosophy would

give the equipment the widest use, which would in turn promote the most creativity in developing improved

orthotic designs. We want to encourage creativity, not stifle it. The following article will discuss this technology,

along with the additional benefits derived from it. Specifically, I will discuss the basic method for capturing the

three-dimensional shape of the foot, the methods used to check the accuracy of the image, evaluation of the

image, educational benefits, and the alteration of the image to produce orthotic corrections.

History of capture technology

    The technology to capture an image has been around for several years. When we first got involved with

automation it was divided into two methods, contact and non contact digitization.

Contact Digitization

    Contact digitization is the primary method used in industry today. It consists of a stylus or probe and a

computer to analyze the data. A data sampling is taken each time the stylus or probe touches the object to be

digitized. This method allows data samplings to be taken at whatever density required and can be highly accurate.

However, the closer the data samplings the longer it takes, since each point is taken separately. It is also

necessary to apply some sort of grid to the object if the data samplings are numerous, to map the position you will

be touching. So, to get data samplings of the plantar aspect of the foot that are 1/8 of an inch apart you would

need to touch the foot approximately 3000 times. This would take several hours at best. However, you could

spread out your data samplings and just get a basic contour of the plantar aspect of the foot. With a basic contour

of the foot you could then compare it to previously stored digitized foot shapes until you found the closest match.

This would change the method in which the lab technicians modify cast corrections. Instead of actually creating

cast expansions and designs such as heel cup height from the digitized image of the foot, you would select the

closest fitting predefined shape and designs. This would be like fitting a wide range of prefabricated orthotics to

the patients foot until you came up with the closest match. The creation of new types of cast corrections or

designs would be input as additional predefined foot shapes by the program developer not the orthotic specialists

or lab technician.

Non Contact Digitization

    Non contact digitization consists of a video camera, a light source such as a laser, and a computer to analyze

the data.  The light source is usually introduced as either a point of light or a band (or bands) or light. With the

video camera and light source positioned at known angles, calculations to determine the x, y and z coordinates of

any given point on the surface of the object can be made.1 The scan with the band or bands of light is faster than

the point of light and results in very quick scan times, usually taking 2 to 7 seconds with most types of imaging

equipment (fig 1). The data samplings can be as close as twenty thousandths (0.020) of an inch, which usually

results in very large data files. A point spacing of 40-80 thousandths (0.040-0.080) of an inch is usually sufficient

to give a very detailed representation of the contour of the foot. We call this method of image capture "optical

imaging". With this method cast corrections will fit exactly to the foot shape, because the cast expansions are

created, evaluated and modified directly from the digitized image of the foot. This method also gives the orthotic

specialist or lab technicians the same control over computer design of cast corrections that they have always had

with plaster corrections. This control is important since cast corrections are a creative process that needs to be

modified using the input of expertise on each cast as it is evaluated. We have been using this method for image

capture of positive and negative casts for automated orthotics since 1987.

    One test used to verify the accuracy of the optical image, is to image a negative plaster cast, then invert the data

points to a positive image in the computer and mill out a positive wax cast. The negative plaster cast is then put

over the positive wax cast to check for three-dimensional accuracy of fit (fig. 2). The negative casts is cut

longitudinally and transversely to check different local areas of fit.1 Another way in which we test for accuracy is

that we pour a positive plaster cast and then image it. The image is then milled to produce a wax duplicate of the

plaster cast. The orthotic is then thermoformed over the wax cast. When the orthotic is finished it is placed over

the original plaster cast to compare the fit.3 We have been using these tests for several years now and have found

the image produced from these non contact digitizers to be an accurate three-dimensional representation of the

original model, be it the negative, positive plaster cast, or the foot.

    It is very important to note that an automated orthotic is only as accurate as its least accurate measuring point.

The least accurate point is the image capture. If you have taken data samplings of 1/8 of an inch apart, then the

orthotic accuracy is 1/8 of an inch. Even though a milling machine can mill out a part to a tenth of a thousandths

(0.00010) of an inch, your orthotic will only be accurate up to 1/8 of an inch because that was your image capture

resolution. 

In office optical scanners

        The first scanner we sold was a non contact scanner for digitizing the foot in the office. It used bands of

white light that shift to capture the image. Image capture requires a couple of seconds and image processing takes

about thirty seconds. The scanner was used to take a non-weight bearing neutral cast. The patient must be in the

prone position. The scanner weighs about 150 pounds and has a video monitor which the doctor uses to view the

foot while placing it into the neutral position before taking the scan. The image produced is a three-dimensional

digitized image with data samplings that are about seventy-five thousandths of an inch (0.075) apart.

The Bergmann Optical Scanner

    We decided to develop our own scanner in 1990 and we asked doctors at that time what they considered

would be important features to include. They stated that they would like a scanner that they could transport to

different locations for seminars, so they wanted one that was compact and lightweight. Doctors also felt it was

very important to have a scanner that didn't change drastically the casting procedure they presently used, one

which would produce an accurate duplication of the foot without being technically difficult to use.2 The scanner

we developed takes into account all the above. It weighs only 40 pounds and is very compact. It is designed to

have the patient sitting and the foot in the classical non weight bearing neutral position, but it can also be used with

the patient prone or weight bearing. We didn't want to limit ourselves to one method that could change as we

learn more about biomechanics. So the orthotic specialist can put the foot in any position that is felt to be

appropriate, or if research is being done on different casting techniques, this scanner has the flexibility to

accomplish it.

Scanner Parameters

    The scanner takes data samplings that are four hundredths (0.040) of an inch apart. There are about 9000 data

samplings taken of the plantar, medial and lateral aspect of an average foot. The time it takes to take an optical

image (scan) is about six seconds. The scanner head goes up and down a vertical track. The scan is taken on the

down stroke. The scanning area is 6 inches wide by 3.5 inches deep by 14 inches high (fig 3). In comparing

plaster casting to optical casting, both methods depend upon the doctor holding the foot in the correct neutral

position in order to get an accurate cast. Both methods give an accurate three-dimensional contour of the foot.

Plaster casting is total contact, and optical casting, as stated above is a data sampling every 400 hundredths of an

inch. Even though both methods are accurate we feel that optical casting has been found to be, over all, more

accurate. This is because the doctor can more precisely check the optical scan on the computer to see if it reflects

his biomechanical examination, such as the forefoot varus or valgus. If it doesn't reflect his examination, he can

take another scan in 7 seconds. With plaster casting, since it takes twenty to thirty minutes you are less likely to

recast. So, although both are accurate you are more likely to scan until its correct than you are to recast until

correct, making optical scanning the more accurate method overall.

    Most doctors who have the scanner use the classic neutral suspension casting technique with the patient supine.

However, the foot can be held in any position desired, such as: a pronated position, supinated position, tripod

technique or corrected position. Since no plaster is used with optical imaging, the doctor has the advantage of

looking directly at the foot, making it easier to judge if the neutral position of the foot is being maintained while

scanning. Markings that have been made on the foot can be seen since they are not covered up by plaster. Also,

since you are only holding the foot for about 6 seconds, it is easier to maintain your casting position. Optical

scanning is excellent for children who tend to move because of the time it ordinarily takes for a plaster cast. If the

cast is incorrect just recast, no materials are, wasted just a little electricity. In a busy practice, especially if the

doctor is doing all of the casting, the reduction of casting time can really be helpful.

    The scanner can also be used to take weight bearing casts. We have designed a weight bearing platform that

has a glass top. To take a weight bearing scan, just lay the scanner on its back and place the platform over the

scanner head. The patient stands on this platform and the scanner scans the foot through the glass. The patient can

stand at angle and base of gait with the subtaler joint held in neutral. Several new uses have been found for this

position besides just casting. One use is to scan the foot in different positions in sports. For example, in golf you

could take a person with a rigid forefoot valgus and show the lateral instability on the back swing by scanning the

patient's foot while they are in that position. Another use is to overlay on the computer a non weight bearing and

weight bearing scan of a patient's foot, then compare the difference between the longitudinal arch heights. The

orthotic arch height could be made somewhere between the heights, making a more tolerable orthotic for sensitive

patients. In the future we may routinely take a non weight bearing neutral cast and a weight bearing cast. The

weight bearing expansions on the border of the foot could then be stripped off and applied to the neutral cast,

making expansions that are more specific for the patient's foot. Another use of this combined scan is to mark a

painful metatarsal head on the non weight bearing and weight bearing cast and compare them when overlaid to

see exactly how much forward movement is produced on weight bearing with different foot types. Semi weight

bearing scans can also be taken for pre and postsurgical evaluation of metatarsal head position. In addition this

casting method is helpful in determining the depth needed when adding accommodations on orthotics used

Verify foot

    It is important to first verify that you have gotten all of the foot when you scan. We found that a common error

doctors make is to assume that the lab personnel would see all of the foot even though they didn't. If you do not

see part of the foot on the graphical image, believe me, it is not there. We will not be able to see it any better than

you. So it is important to verify that all of the foot is present. The cross section view is an excellent view to

examine the foot, because it presents a topographical map where you can see if you have captured all of the heel.

You should see a nice rounded heel in this view. If the heel is cut off then you need to take another scan (fig 4).

You have either scanned the foot too low or off to the side of the scan area.

.

Rotation of image to neutral

    After you have verified the image is complete, you will need to position the image in the neutral position. This is

accomplished by bringing up the neutral or bisection view on the computer. If you have marked the bisection of

the heel with a flair-type pen it will be seen in this view.5,7 Next you select the protractor on the computer menu.

The protractor is used to rotate the foot so the bisection of the heel is vertical to the screen. You can also bisect

the whole volume or align the protractor line tangent to the deepest part of the curvature on the lateral side of the

heel below the lateral malleolus. Doing this will put the foot in the proper alignment to evaluate it biomechanically.

Three-dimensional color height contour map of the foot

    After the foot is scanned, the computer analyzes the data and produces a varicolored three-dimensional graphic

image on the computer monitor. The plantar aspect of the foot is facing upward. The highest points on the forefoot

and heel are represented by the color red and they graduate downward into other colors (fig 5). Each change in

color represents a height difference of 1/10 of an inch. The color variations and their height differences help to

define greater depth perception of the image.1 The colors also help us to instantly see what basic type of forefoot

deformity is present. For example, if you have a forefoot varus you will see red on the fifth and fourth metatarsal

heads and then graduate into other colors as you look across the rest of the metatarsal heads. If you have a

forefoot valgus you will see red on the first metatarsal head, graduating into other colors on the rest of the

metatarsal heads. You will also notice that the graphical image is not just of the plantar aspect of the foot but it

also goes around the foot to display the sides of the foot to the height of the lateral and medial malleolus. The

default view is looking from the heel to the forefoot, but the foot can basically be spun around and viewed from

any angle desired. The foot can also be enlarged if you need to look at a specific area.

Gray scale view

    Besides the three-dimensional graphic view there is also a three-dimensional gray scale view. This view is like

looking at a black and white video. It is used to see markings on the foot, such as, for platform boundaries,

markings for metatarsal head accommodations or other landmarks. Any area marked with a flair type pen can be

seen in this view.  Unlike a two-dimensional video, this three-dimensional view can be rotated to see the foot from

other angles. The video image can also be enlarged to look at specific areas.

Forefoot difference and cross section

    When you use the forefoot difference command you will see the forefoot pop up in front of you and the

computer will calculate the forefoot varus or valgus if present.2 This forefoot difference can be calculated up to a

half (0.5) of a degree. If you do not see the forefoot degree that you found on your clinical exam you may want to

rescan the patient's foot. The image of the foot can also be sectioned transversely or longitudinally to study any

section of the foot as to its curvature. This view is very helpful in comparing the overlays of weight bearing and

non weight bearing images of the foot.

Height plot

    Another feature is the height plot which allows you to see the height of one area in relation to another. For

example, if you had a patient with a plantar-flexed third metatarsal head, you could mark the metatarsal heads on

the foot with a flair-type marking pen before scanning the patient. They will then show up on the topographical

view of the foot in the height plot screen.2 The arrow keys are used to move across this topographical map and

the metatarsal heads are then remarked. These data samplings will display as a color bar height chart, which in our

example would show the third metatarsal head to be higher than the other bars in the chart. It also would display

the difference in angle and depth of one metatarsal head to the other (fig 6). Another use of height plot is to show

the change of declination angle of the first metatarsal between a non weight bearing and a weight bearing scan of

the foot. To find the difference, mark the shaft of the first metatarsal proximal and distal on the non weight bearing

and the weight bearing scans and compare the difference between two angles.

Graphic gait analysis illustrations

    The gait analysis was an outgrowth of the ability to display the foot three-dimensionally. It was developed to

help show patients how their foot functions within their own particular biomechanical framework. For example, if

you had a patient with heel pain due to pronation caused by a forefoot varus of 3.5 degrees and a subtaler varus

of 4.5 degrees and a tibial varum of 5 degrees, the patient could see how the foot compensates4 (pronates)

without an orthotic, by using graphic images of their feet to illustrate the gait cycle. These pictures are clear and

comprehensible, even to someone with no knowledge of biomechanics. Just as important as having your patient

understand what has caused his condition is having him understand what you are going to do to correct it. To

accomplish this forefoot and rearfoot posting is added to their graphic foot images and then they are rotated

through the gait cycle. Without using a single biomechanical term the patient can view the computer images and

see how the foot does not compensate (pronate) as much with the proper orthotic posting. (fig 7). So, without

having to understand biomechanics, the patient can comprehend what is causing the foot problem and what needs

to be done to correct it. Most doctors have found this to be a very important tool for expanding their orthotic

practice. Just as biomechanical terminology gave the orthotic specialist a common language to communicate

with one another, these graphical gait images have broken down the barriers between the doctor and the patient

and given us all a common language, that of pictures. Any of these pictures can be printed out in full color and

given to patients for a reference to their foot problem. Also, when a patient is either referred to you or by you to

another medical specialist, that specialist might have little or no knowledge of biomechanics and these printouts

have become a great adjunct in explaining the patient's biomechanical problem and the treatment in a clear and

concise manner. It's an excellent way of creating better understanding and appreciation of   biomechanics in the

medical community.

Using the scanner to promote

    Since the scanner is compact it can be toted to other locations to promote your presence in the community.

Foot screenings can be done at schools. Each student can be given a printout that they can take home so their

parents can understand their foot problem. We have also developed a foot-sizing program that will calculate the

proper shoe size when taking a weight bearing scan (fig 8). Some orthotic specialist have used this to work in

conjunction with local shoe stores to promote proper shoe fit. The shoe store sells more shoes and the orthotic

specialist has the potential of seeing more patients. Both benefit.

    If you lecture to other medical specialists, the scanned images of the feet can be used to explain the basic

principles of biomechanics. Even if the medical audience had very little knowledge of the foot, the images, such as

the gait cycle, would give you the ability to explain biomechanics in a language all people understand: pictures.

    Another feature is that any of the graphical images that you see on the computer screen can be converted into

pic files that can be imported into desktop publishing for newsletters or into computer software that outputs slides

or overhead transparencies.

Sending the optical image to the lab

    When you have decided to send a case to the lab all you need to do is fill out the prescription on the computer.

At the end of the day if you select "send", the computer will sort out the optical casts you wish to send and bundle

them together for transmission to the lab by modem over your existing phone lines. Completely eliminating your

shipping supplies and packaging time. You can literally cast the patient and have it received by the lab five minutes

latter. Each pair of casts takes about 5 minutes to send over a 24 baud modem. However, the transmission time is

going down drastically as compression programs continue to make files smaller and as the modems get faster.

    Computer aided design or three-dimensional modeling in which you change the shape of a computer image was

originally used only by big companies like the auto or space industry. This was because they were the only ones

that could afford the mainframe computers and programs which ran them. However, as microcomputers and

programs became more powerful they got to the stage that they also were able to handle complex

three-dimensional modeling. This reduced the cost so that smaller companies such as ours could afford to develop

it. Using three-dimensional modeling means that we are adding the cast corrections and expansions over the

uncorrected image of the foot. So, just like plaster, the uncorrected foot is the basis on which you build.

However, unlike plaster, we have developed software which allows us to evaluate the fit of the cast correction

overlay by looking right through to the uncorrected foot (fig 9). All aspects of the cast corrections are adjustable

by simple up and down arrow commands. For example, if you want more cast fill in the medial longitudinal arch

you simply go to that section of the foot image. The display will show several transverse cross sections of the

uncorrected foot in gray and the corrected overlay or fill in pink. To add more fill to the corrected overlay you

simply tap on the up arrow key, each time you tap on it, more correction or fill is added. If you want to reduce the

amount of fill just tap on the down arrow key and you will see the cast fill become less each time you tap on it.

Any of the other areas of fill such as the lateral expansion, the heel seat depth, and the forefoot platform are

controlled the same way. All of the original percentage and slope blends have been reduced to the use of these up

and down arrow keys. This makes a very sophisticated process into one that can be understood and used by

anyone with a basic understanding of cast corrections and virtually no computer knowledge. Orthotic specialists

were so impressed with how user friendly it was that we came out with a version that could be used in their office.

We call it the "doctor design program". This has allowed more people to become involved in the creative process

of designing new and better cast corrections and orthotics.

Re-contouring the foot shape for women's high heel shoes

    The ability to alter the shape of the foot allows us to re-contour the shape of the foot for high heel shoes (fig

10).  Without this ability you would need to cast the patient in the high heel shoe to get the contour. There is a

menu selection which re-contours the optical image to a formula in relation to the heel height of the high heel

shoe.3 This method also works when adding a heel lift to an orthotic for a limb length difference. It also is

effective when making orthotics for cowboy boots. By re-contouring to the heel height, the orthotic is not tipped

downward, the foot doesn't slide forward, and the back of the heel cup does not tip up, which can cause irritation

at the back of the heel. 

Adding accommodations to orthotics

    Local accommodations for metatarsal head relief, metatarsal raises and heel spur accommodations are also

added on top of the uncorrected image of the foot.1 We have stored many of the original handmade shapes we

have always used at our lab. With complete control over the digitized image it also allows us to individually design

these corrections for each foot. The foot needs to be marked in order for us to transfer these corrections to the

digitized image.

Retaining doctor preferences

    We have evaluated three-dimensional models for over six years and have found certain types of cast

expansions ideal for various types of orthotics and foot types such as sports orthotics, pump orthotics and foot

shapes such as cavus or high arched feet. These are available as generic corrections for the lab and the "doctor

design program". If a doctor has certain particular corrections that he/she likes, they can be saved and brought up

every time his account number is input. In this way the system can be more specific than the unautomated system.

It is very hard for a lab technician to remember more than a few basic cast designs, but the computer has no limit.

It can store hundreds of thousands of cast expansion designs.

    Computer aided manufacturing consists of milling out the designed part, in this case, the orthotic or cast

correction. For those not familiar with milling machines, the best way to describe it is to compare it to an electric

hand drill. Imagine moving a drill side to side, up and down, and forward and backward all at the same time. This

is what a milling machine does. Now if you add a computer to the system to direct its path, speed of direction,

and revolutions per minute, you now have a CNC milling machine. If you mill out an uncorrected cast you will

create a tool path that will follow dot for dot the exact shape of the foot that has been digitized. If you mill out a

corrected cast it will follow the foot except where you have changed it with expansions. Molding wax is used as

the material of choice by our lab for milling casts. It is commonly used in industry to mill out prototypes. It is easily

milled and after the orthotic is fabricated over it the wax can be melted down and reused, making it economical

and also environmentally responsible. If you are going to mill an orthotic out directly you have to mill both sides of

the material, one side for the top surface of the orthotic and the other for the bottom surface. In a plastic orthotic

you are milling out the posts, including the grind off and the top surface.1 In a soft orthotic such as korex you are

milling out the bottom wedging and then the top surface which may also include metatarsal raises or pockets.

    Our lab has found that directly milled orthotics although a part of the spectrum of orthotics, are not the

majority. Most practitioners see a wide range of foot problems in their practice that need a wide range of orthotic

devices. We have found that the majority of cases that require sophisticated orthotic control need a combination

of materials that can only be provided at this time by thermoformed or cold formed laminated materials over a

computerized wax cast. Our micothotic which is intrinsically posted in the forefoot and heel orthotic can either be

directly milled or thermoformed. Both give the same control and wearablity, however the thermoformed orthotic is

more economical to produce and does not waste as much material which is more environmentally responsible.

New types of injection molding now underway will eliminate many of these production problems. As stated

earlier, although the milling machine is accurate up to a tenth of a thousandths (0.00010) of an inch, the milled out

orthotic or cast correction will not be because the initial capture was one four hundredths (0.040) of an inch.

Therefore, an orthotic can only be as accurate as its least accurate manufacturing process.

Cast storage

The corrected and uncorrected image is able to be broken down and saved. It is backed up to a tape drive and

can be recalled at any time to make a new pair of orthotics or to be used to compare to a new image to see if the

foot has changed. It also can be used if the orthotic needs to be adjusted. The advantage with using the image for

the adjustment is, since you can look through the correction to the underlying uncorrected image, to see exactly

how much correction has been added.3 For example, if the arch height needs to be lowered, it can be redesign to

add the exact amount of fill addition desired, then save the correction in addition to the original. This way, you will

always know exactly what you have added. Unlike the storage of plaster casts which physically takes up a good

amount of space, digitized images take up virtually none. A two by five inch tape will store several hundred pair of

feet at a time. This has given our orthotic lab the ability to store the images not just six months , but indefinitely.

This data bank of hundreds of thousands of foot images should prove to be a valuable source of research for

various biomechanical studies. Unusual foot types can be saved and used for teaching purposes.

In conclusion

The word automation can sometimes be misleading. I think some feel that automation of orthotics means a system

where orthotics are massed produced. Also, that the main reasons for automation is to make orthotics quicker

and cheaper, not individualized but kind of a blur of the sameness. As far as our lab is concerned, nothing could

be farther from the truth. We think of automation as the use of a computer technology to make orthotics that are

more accurate, and ones that will be more specialized as to the doctor's specific requirements. This is a method

which takes the expertise of the biomechanist and the lab technician and gives them a tool that can be used to

more accurately reflect the perfection in orthotic fabrication they are trying to attain. In testing out new types of

cast correction designs, computer designed images will give the repeatability needed in scientific research not

found with manual or hand made corrections.