- Email: [email protected]
Simple microcomputer system for mapping tissue sections with the light microscope
Journal of Neuroscience Methods, 25 (1988) 165 173
165
Elsevier NSM 00858
Simple microcomputer system for mapping tissue sections with the light microscope Arturo Alvarez-Buylla and David S. Vicario The Rockefeller Unit,er~itv, New York, N Y 10021 (U.S.A.)
Received 8 December 1987) (Revised 7 March 1988) (Accepted 12 May 1988)
K e y words: C o m p u t e r - m i c r o s c o p e ; D r a w i n g - t u b e :
Pantograph;
M i c r o c o m p u t e r : Plotter: M o r p h o m e t r y :
Distribution; Neuroanatomy We describe a method for two-dimensional mapping of tissue sections that makes use of a drawing tube. microscope stage encoders and a microcomputer. The drawing tube views the graphics monitor and superimposes the image of the screen cursor and on-screen menus on the specimen image. Thus, the position of every landmark in each microscopic field can be mapped without stage movement while directly viewing the specimen through the microscope. A mouse is used for data entry and program control. Fields mapped in this way are then assembled into a complete map, which can include line drawings as well as up to 20 landmark types. The coordinate values of all landmarks mapped are stored and remain accessible for editing and analysis. High resolution plots are produced. Specialized functions include grain counting, area and perimeter calculations as well as a perimeter limiter that predefmes the area to be mapped. The system uses general purpose hardware that is widely available. Man 3' hitherto time-consuming tasks, such as detailed mapping of cell positions, regions of immunocytochemical staining, degenerating fibers, neuronal connections or any other anatomical feature, can be done in a fraction of the time and effort previously involved. These labor savings can be realized while maintaining the highest resolution and enabling statistical analysis since the data are already in digital form.
Introduction E x a m i n a t i o n of tissue sections u n d e r the light m i c r o s c o p e frequently requires the use of highp o w e r objectives to visualize details of interest (e.g. cell types, a u t o r a d i o g r a p h i c grains, i m m u n o histoc h em i cal stains or d e g e n e r a t i n g fibers). T h e spatial relationship of such details to each o th e r and to the rest of the section is often lost due to the small field size. It is i m p o r t a n t to be able to m a p fine details (referred to here as ' l a n d m a r k s ' ) in relation to the whole section, especially in studies of regional specialization of the brain.
Correspondence: A. Alvarez-Buylla, The Rockefeller Univer-
sity, 1230 York Ave., New York, NY 10021, U.S.A.
C o m m o n but expensive and t i m e - c o n s u m i n g m e t h o d s for m a p p i n g include the c o n s t r u c t i o n of p h o t o g r a p h i c m o n t a g e s or the p o s i t i o n i n g of l a n d m a r k s by h a n d on a low m a g n i f i c a t i o n picture or d r a w i n g of the region to be m ap p ed . It is easier and m o r e efficient to use the position of the m i c r o s c o p e stage to drive the X and Y axes of a plotter. Several such systems have been developed, such as the electronic plotters (Boivie et al., 1968: G r a n t and Boivie, 1970: Eidelberg and Davis, 1977) or the light p a n t o g r a p h (Patterson et al., 1976). R e c e n t c o m p u t e r - a s s i s t e d m a p p i n g systems (Curcio an d Sloan, 1981; reviewed in Mize 1984, 1985) o f t en m a k e use of specialized h a r d w a r e and frequently require that each feature of interest be aligned u n d e r a cross-hair in the m i c r o s c o p e eyepiece ( F o r b e s and Petry, 1979: Williams and Elde, 1982; Mize, 1985). L a n d m a r k s may bc
0165-0270/88/$03.50 (~" 1988 Elsevier Science Publishers B.V. (Biomedical Division)
166 visualized throughout the microscope field, but, because they are only recorded by the computer when they fall under the cross-hair, an excessive amount of stage movement is necessary. In an ideal system, the operator would position a pointer visible through the microscope on a feature of interest anywhere in the field, and enter its position in the computer. While this can be accomplished with digitized video images displayed on a graphics terminal (DuVarney and DuVarney, 1985), such systems are expensive and do not approach the resolution of direct visualization through the microscope. This principle has been better realized in the form of image-combining systems that use a camera lucida to superimpose the tissue image with an image of a computer-controlled graphics device (Glaser et al., 1979, 1983; Glaser and Van der Loos, 1980). In this report we describe a simple system for two-dimensional mapping based on this principle. It makes use of common microcomputer hardware and a drawing tube to view a mouse-driven cursor on the computer screen, while two digital indicators relay the microscope stage position to the computer. Software developed for this system thus allows rapid entry, display and plotting of the positions of cells and fibers (or any landmark) throughout the section while looking through the microscope.
Methods and Materials
Hardware The standard hardware consisted of an IBM PC-XT or PS/2, Model 30 with two serial ports, one parallel port, Hercules graphics card, monochrome monitor, mouse (Microsoft) and plotter (Hewlett-Packard models 7470A or ColorPro). A related system is under the development for Macintosh microcomputers (C. Ghez, personal communication). Specialized equipment consisted of two digital linear position sensors (Mitutoyo Digimatic) with a serial interface (Mitutoyo Multiplexer 10) (Fig. 1). These provided X and Y positional coordinates of the microscope stage to the computer. The indicators were mounted on the arm of a Nikon Labophot microscope or directly
Fig. 1. Mapping system components. The X stage encoder ~s hidden behind the Y encoder (y). The encoders ~ransmit the stage position to the interface (i) that relays this information to the computer. The drawing tube (arrow) is oriented towards the computer screen, allowing direct observation through the microscope of the mouse driven cursor and menus. onto a Zeiss microscope stage with custom-made brackets (Fig. 1). The X indicator has a range of 2.5 cm and reads the short axis of a 2.5 x 7.5 cm slide. The Y indicator has a 5 cm range, sufficient to cover the non-frosted portion of the slide. A camera lucida (drawing tube; Nikon) was modified to visualize the computer screen superimposed on the image of the tissue section. The end mirror was oriented horizontally instead of in the manufacturer's fixed vertical position (looking down at the table). Camera lucidas l'rom other companies are equally easy to adapt. With the drawing tube and monitor aligned, the screen and mouse cursor were clearly visible on the specimen image, enabling interactive data entry while scanning sections (Glaser et al., 1983). In cases where the camera lucida optics attenuated too much of the light emitted from the tissue, e.g. viewing fluorescently-labeled landmarks, positioning was accomplished by matching the location of landmarks in a reticle grid to corresponding positions in an identical grid displayed on the terminal.
Software Computer programs were written in BASICA (IBM) for graphics equipped computers. One version takes advantage of text cursor positioning commands to provide a pseudo-graphics display and can run on computers without graphics
167 h a r d w a r e . T h e p r o g r a m s w e r e d e s i g n e d to a i d in t h e s y s t e m a t i c s c r e e n i n g o f t i s s u e s e c t i o n s . U p to 20 d i f f e r e n t l a n d m a r k t y p e s o r l i n e d r a w i n g s c a n be selected from a screen menu by means of the mouse. The landmarks or lines are then displayed on the screen, superimposed on the image of the section. In this way, the data entered on the image o f a s i n g l e field c a n b e e d i t e d b e f o r e b e i n g s t o r e d a n d p l o t t e d . T h e X - Y c o o r d i n a t e s of t h e c e n t e r o f t h e field a r e t r a n s m i t t e d t o t h e c o m p u t e r b y pressing the 'load' button of the multiplexer interface. T h i s also s i g n a l s t h e p r o g r a m t o s t o r e a n d p l o t t h e i n f o r m a t i o n in t h e c u r r e n t field a t t h e appropriate X Y location. To insure that fields or l a n d m a r k s a r e n o t c o u n t e d twice, s e v e r a l c h e c k points are built into the program. Erroneously
e n t e r e d p o i n t s c a n b e e r a s e d . T h e f i n a l o u t p u t is a plot of the mapped histological section, showing all t r a c i n g s a n d l a n d m a r k s e n t e r e d . All t h e d a t a a r e s t o r e d o n t h e h a r d d i s k , a n d c a n b e r e c a l l e d t o m a k e p l o t s of s e l e c t e d l a n d marks. With a separate program, the map can be r e d r a w n o n t h e c o m p u t e r s c r e e n : t h e m o u s e is t h e n u s e d t o select s c r e e n p o s i t i o n s , a l l o w i n g dist a n c e s , a r e a s a n d p e r i m e t e r s to b e c a l c u l a t e d .
Procedure The mapping procedure consists of using the m o u s e to e n t e r t h e l o c a t i o n o f l a n d m a r k s o r o t h e r f e a t u r e s in t h e c o r r e c t p o s i t i o n w h i l e v i e w i n g t h e s e c t i o n a t h i g h m a g n i f i c a t i o n ( F i g . 2). P o s i t i o n w i t h i n e a c h field is d e t e r m i n e d as t h e m o u s e ' s
Fig. 2. Superimposed computer screen (screen field) and specimen image. The screen field and menus are presented here unrotatcd, though in the microscope they appear to be rotated by 90 ° clockwise. A: the positions of two neurons and a glial cell with autoradiographic grains overlaying their nuclei are marked by mnemonic screen characters (faintly seen in this photograph). Fhc mouse-driven cursor (arrow) serves as a pointer to designate the position of cells to be marked by pressing one of the two mou.',c buttons. In this case, each mouse button is loaded with neuron (NEU) or gila (GEl) as indicated in the highlighted landmark menu to the right of the field. The menu to the left contains different editing functions. Del, clears the field: Dpf, deletes previous field: Ins, inserts into previous field: Rvi, recovers previous field, chk, activates double landmark checking subroutines; sss. silences computer codes: S:P, skip plotting for this field. The menu at the bottom contains: 1, enters border in center of the field: 2, activates subroutine to find previous field; 3 and 4, pantograph trace functions; 5, records position where photo was taken. B: line drawing of radial gila fibers (see Fig. 4). The mouse is loaded with the 'LIN' function (highlighted on the menu to the right) and the cursor ix used to trace the fibers on the field. The second button on the mouse was loaded with "ERA', used to erase incorrect landmarks from the screen
168
15 A2 :
A
08-09-1987, 23= ;)9 /^
/
o
o
o
/
o~-..--_
/-'-
-.
•
(
X
..-
/
."
I××
/
o
-
~
/
Y
o NEURO* ;)5 • GLIA * 36 x P]CNOT ]B
..y/
o:, o
SPY-MIG
~
0
~...
xl
~-
,
D
. v
B
~
:
y
~ . ' :-:'-" "-" •
°
oO,
°
° •
,,
•
°
•
,s.,
o
,0a.
." '.~ , 0 % o ,
t 0.Sin
~j
•"
°
"
-s.
;
•
o
""'~A",.'". ~ 0°'s
v•', ,..
,°
;"
°
° *
..........
°
°°~
°%
• , .,....:....:..-,o.~. :..~. -7, °
~
.o. FLUI~O 55 • f.ot-nr 4gl]
%
•
.
°
°oa
°
;.
°
°
".-",,"~: :. ..':_,-..": ;.;y
~ "
".:": " . " '
' " ": ¢/
PIFRIIETER
E5211
.356 m =
Fig. 3. Examples of program output. A: m a p of radiolabeled cells in the canary telencephalon (see example 1 of text). The upper right corner is the legend with file name (A152A2W), date, time, program name, landmark counts and scale bar. The small rectangle to the right of the m a p illustrates the proportional field size as viewed through the microscope while mapping. B: maps of fluorogold-labeled cells (upper panel) and total neurons (lower panel) in nucleus HVc of the canary brain (see example 2 of text). The dashed line is the surface of the brain and the dotted line is the ventricle. The region to be mapped was pre-established using the perimeter-limiter and demarcates the borders of HVc (solid line). This area and the perimeter length are shown in an additional plot legend.
169 cursor screen position. T h e a b s o l u t e l o c a t i o n of each field is d e t e r m i n e d from the p o s i t i o n a l indicators which r e c o r d m o v e m e n t of the stage a n d slide. In o r d e r to m a p an entire section or region, successive fields need to be m a p p e d a c c o r d i n g to a s c a n n i n g strategy. The p r o g r a m allows a choice b e t w e e n s c a n n i n g parallel to the X axis, the Y axis or a r a n d o m trail. T h e best choice d e p e n d s on the size, shape a n d o r i e n t a t i o n of the s p e c i m e n as well as on the preference of the user. Initiation. A f t e r zeroing the indicators, the m i n i m u m a n d m a x i m u m X- a n d Y c o o r d i n a t e s are loaded. This a u t o m a t i c a l l y sets the best scale to m a x i m i z e the size of the plot. A reference p o i n t at
a k n o w n l o c a t i o n in the section is entered. This can be used later, if s u p e r i m p o s e d m a p p i n g is desired. Scale, field size, relative s y m b o l size, l a n d m a r k n a m e s a n d p l o t t i n g status are a u t o m a t i cally set b y the p r o g r a m , b u t can be m o d i f i e d by the user. A b o u n d a r y ( p e r i m e t e r limiter) can be e n t e r e d with a lower p o w e r objective that limits m a p p i n g to that area; this p e r i m e t e r limiter is especially useful when m a p p i n g details within a nucleus for which the b o r d e r s are not distinct at high m a g n i f i c a t i o n (Fig. 3B). A square or a rectangle (screen-field) set to m a t c h the size a n d s h a p e of the eyepiece reticle (eyepiece-field) a p p e a r s on the screen. A menu of
/
J l L..
i.
,
Fiq.
2B
Fig. 4. Map of radial gila fibers stained by immunocytochemistry (see example 3 of text) in a frontal hemisection of the rostral telencephalon of the canary brain. The midline is to the right and dorsal is up. The small rectangle inside the map indicates the field from which the photograph in Fig. 2B was taken. The solid line next to the midline is the partially collapsed lateral ventricle.
170
function codes appears to the left of the screen field and a menu of landmark names to the right. By adjusting the relative illumination of the microscope and the screen intensity, one can visualize the microscope field superimposed on the computer screen image (Fig. 2). Due to the optics of the drawing tube, the screen as seen through the microscope appears to be rotated by 90 ° To compensate for this rotation, the mouse is used sideways and serves as a pointer for the desired function or landmark. Mapping. Each button of the mouse is loaded with a landmark (e.g. neuron or glia) selected from the menu (Fig. 2). The cursor is then moved inside the screen-field to point to the desired feature of the superimposed microscope image. The appropriate landmark symbol is placed at this position on the screen by pressing the correct mouse button. In case of error, one can erase individual landmarks or clear the whole field. When all landmarks have been entered, the completed field is loaded, signaling the computer to calculate the absolute coordinates of all the landmarks, plot them and store them on the disk. The stage is then moved to view the adjacent field. The program signals if a field overlaps a previously loaded field or if the perimeter limiter (see above) has been violated. The program checks that all fields are entered at the right locatiom preventing the user from accidentally jumping into adjacent rows. If the landmarks are densely distributed, every contiguous field must be mapped. If they are more scattered, only those fields that contain landmarks are mapped. In addition to entering tissue landmarks which map to a point, more complex structures (e.g. degenerating fibers, axons or dendrites) can be entered as line drawings. This can be done in two ways. (1) Within a field, the mouse can be used to trace the structure of interest, using the ' L I N ' function (Figs. 2B and 4). (2) For larger structures, extending beyond a single field, the tracing mode can be used to draw lines (Figs. 3 and 4). In this mode the microscope and computer work like a pantograph: the plotter follows the movement of the stage, using the color pen and line type selected. This is useful to draw the contour of the section, laminae, long fibers, large blood vessels or the ventricles.
Termination. All plots receive an identifying legend, including the filename, date and time, a scale bar, counts of landmarks, scanned area and. when using the perimeter limiter, the area and perimeter (Fig. 3). All plotting data are stored in a random access file enabling the plot to be repeated, modified or completed.
Examples Example 1: mapping the entire section. Dividing cells were labeled in a one-year-old male canary by two injections of [3H]thymidine. Forty days later the brain was fixed and sectioned at 10 ~m. The sections were incubated for autoradiography, stained with Cresyl violet, dehydrated and coverslipped. A section at the level of the anterior commissure is shown in Fig. 3A; the positions of [3H]thymidine-labeled cells were mapped only within the telencephalon. A 63 × objective (Zeiss) was used to recognize and classify labeled cells. The procedure was as follows. (1). The contour of the brain's lateral ventricle was drawn with the trace function (pantograph), represented by the heavy solid line within the section. Using the same function, the dashed line was drawn to show the hyperstriatal lamina. (2) Mapping began at the top of the section and proceeded to the bottom in a zigzag pattern of horizontal scanning rows. Border landmarks were entered at the ends of each scanning row, and the program automatically joined these points to form the solid line of the section perimeter. (3) Within each scanning row, fields were marked with the position and type of labeled cells encountered there. The program put together the fields to produce the final map (Fig. 3A). Example 2." mapping part of the section. Neurons in the hyperstriatum ventralis pars caudalis (HVc) of an adult male canary were labeled with fluorogold (Schmued and Fallon, 1986) by retrograde transport from area X. Brains were embedded in P E G and 6 /xm thick sagittal sections were cut and mounted onto glass slides. HVc is just below the lateral ventricle of the caudal forebrain and plays an important role in song production (Nottebohm et al., 1976). An easy-to-find reference point was selected for later use. (2) The perimeter limiter was entered around the fluoro-
171
gold-positive region (Fig. 3B, solid line) using a 20 × objective. (3) The position of the individual fluorescent neurons was then mapped systematically with a 63 × objective; a series of fields was completed from left to right, then down one field and from right to left and so on. (4) The lateral ventricle was entered with the tracing function as a dotted line and the surface of the brain as a dashed line (Fig. 3). The section was then stained with Cresyl violet. dehydrated and coverslipped. This procedure destroys the fluorogold fluorescence, but makes it possible to recognize all neurons by their Nissl staining. (5) The stained section was again repositioned under the microscope using the selected reference point and all neurons within the HVc were then mapped (Fig. 3B). After staining, the borders of HVc are difficult to find especially when using a high power objective (63 x ). However, since before staining the perimeter limiter was determined while mapping the fluorescent cells, neurons can now be entered in and around the HVc without having to worry whether or not they are inside of the HVc. The program rejects all neurons found outside the perimeter limits. The output is a map of fluorogold-labeled neurons and total neurons (Fig. 3B). The area and perimeter are calculated and printed in the legend. Example 3: drawing lines. P E G sections (10 > m thick) of canary forebrain were stained with a monoclonal antibody (Alvarez-Buylla et al., 1987). In the adult avian brain this antibody recognizes radial glia with their cell bodies on the walls of the lateral ventricle and thin processes that course laterally into the forebrain. The stained glial fibers are clear only at relatively high magnification (40 x objective), hence a given field contains just a small fraction of the section. A hemisection of anterior canary forebrain is used here to illustrate the procedure (Fig. 4). (1) The ventricle and the section borders were traced (pantograph, see above) using two different pen colors (both shown as black in Fig. 4). (2) Starting at the top, each horizontal row was completed before moving one field down to the next row. Each fiber seen in the microscope field was traced by using the mouse (in ' L I N ' mode) to move the
cursor on the screen field. The program assembles all the fiber drawings into a complete map (Fig. 4). With more conventional methods, i.e. photographic montage or mapping field-by-field by hand into a grid, it often took up to two weeks to complete a map of one hemisection. By using this system the hemisection shown here (Fig. 4) was mapped in one day.
Discussion A powerful feature of this system is that points are mapped while directly viewing the tissue through the microscope at the optimum magnification, using a camera hicida. Other systems, designed mainly for neuronal reconstruction, have used a similar principle (Glaser et al., 1979: Capowski and Sedivec, 1981; Glaser et al., 1983, Capowski, 1985, Freire, 1986). By assembling a complete map from individual fields, the absolute locations of cellular features can be stored at high resolution (1 /,m) for tissue sections up to 2.5 cm in size. These are important advantages when compared to the digitized video images with a superimposed cursor that are commercially available. While recently developed high resolution video systems (4096 x 4096 pixels), using expensive specialized hardware and software, may provide improved resolution (Hillman et al., 1987), the images they generate require large amounts of mass storage space. Our program only records landmarks of interest and their coordinates, since it is not necessary to store the entire image for many common applications. The most complex example illustrated (Fig. 4), needed only 70 kilobytes of disk space. The mapping resolution in our system is determined by two parameters: (1) the stage position indicators have a 1 > m resolution, and (2) the resolution of the screen. The effective size of the pixel image seen through the microscope depends on the objective used. This means that at high magnification, the effective screen resolution is equal to or better than the stage resolution of 1 /~m, which then sets the limit. At lower magnification, the screen resolution is the limiting factor, reducing resolution to about 2 or 4 ~tm at 40 x or
172 20 x , respectively. The m a p resolution is limited by the n u m b e r of a d d r e s s a b l e p o i n t s in the screen-field of the c o m p u t e r m o n i t o r a n d the plotter. In our experience, the graphics m o d e of the screen exceeds w h a t is required for most a p p l i cations. A slight error is i n t r o d u c e d because of the curvature of the screen, m a k i n g the sides of the screen-field a p p e a r slightly curved when o b s e r v e d through the m i c r o s c o p e (Fig. 2B). The c a l c u l a t e d m a x i m u m d e v i a t i o n is 0.8% of the size o f the field. This could only b e of significance when w o r k i n g with a very low m a g n i f i c a t i o n objective. F l a t screen m o n i t o r s which would solve this p r o b l e m have recently b e c o m e available a n d a f f o r d a b l e for personal computers. A n e n o r m o u s a d v a n t a g e of c o m p u t e r - b a s e d p l o t t i n g m e t h o d s is the a b i l i t y to edit a n d analyze this m a p during a n d after d a t a collection (Mize, 1985). This includes calculating areas, distances a n d d i s t r i b u t i o n statistics, r e p l o t t i n g only p a r t of the file or m o d i f y i n g the plot. M a p p i n g of a u t o r a d i o g r a p h i c a l l y l a b e l e d cells, a c o m m o n a p p l i c a t i o n in our l a b o r a t o r y , was originally d o n e b y d r a w i n g c o l o r p o i n t s into a previously taken picture (or pictures) of the tissue section. This was tedious a n d time consuming, taking f r o m 4 to 5 days to c o m p l e t e a section of a p p r o x i m a t e l y 5 × 5 ram. W i t h the c o m p u t e r - a i d e d m a p p i n g system described here, one such m a p can be c o m p l e t e d in less t h a n a day, T h e same c o m p u t e r can also be used for other tasks. The specialized h a r d w a r e connects to a s t a n d a r d serial port, a n d costs less than $1000 at this time. Once installed, this system can be used by l a b o r a t o r y p e r s o n n e l with no specialized c o m p u t e r knowledge. T h e p r o g r a m s m a y be o b t a i n e d by c o n t a c t i n g the authors. The software is written in B A S I C A and is easily c u s t o m i z e d for new applications.
Acknowledgements T h e s u p p o r t o f P H S grants M H 18343 (to F e r n a n d o N o t t e b o h m ) a n d M H 40900 (to D.V.) is gratefully a c k n o w l e d g e d . The w o r k was also supp o r t e d in p a r t b y N I H g r a n t B R S G S07 RR07065, a w a r d e d to Rockefeller University. W e t h a n k
F e r n a n d o N o t t e b o h m for his enthusiastic s u p p o r t for this project, a n d for his helpful c o m m e n t s ~n this m a n u s c r i p t . W e also t h a n k M a r g a Theelen for her help with the m a p p i n g a n d Jud~ Dye for secretarial assistance.
References Alvarez-Buylla, A., Buskirk, D.R. and Nottebohm, F. (1987) Monoclonal antibody reveals radial glia in adult avian brain, J. Comp. Neurol., 264: 159-170, Boivie. J., Grant, G. and Utfendahl, H. (1968) The X-Y recorder used for mapping under the microscope, Acta Physiol. Stand., 74: 1A-2A. Capowski, J.J. (1985) The reconstruction, display, and analysis of neuronal structure using a computer. In R.R. Mize (Ed.), The Microcomputer in Cell and Neurobiology Research, Elsevier, Amsterdam, pp. 85-109. Capowski, J.J. and Sedivec, M.J. (1981) Accurate computer reconstruction and graphics display of complex neurons utilizing state-of-the-art interactive techniques, Comp. Biomed. Res., 14: 518-532. Curcio. C.A. and Sloan, K.R. (1981) A computer system for combined neuronal mapping and morphomet~, J. Neurosci. Methods, 4: 267-276. DuVarney, D. and DuVarney, R.C. (1985) A computer-based video microscope for cell measurement. In R,R. Mize (Fxt.). The Microcomputer in Cell and Neurobiology Research, Elsevier, Amsterdam, pp. 233-246. Eidelberg, E. and Davis, F. (1977) An improved electronic pantograph, J. Histochem. Cytochem.. 25: 1016-1018. Forbes, D.J. and Petry, R.W. (1979) Computer-assisted mapping with the light microscope, J. Neurosci. Methods, 1: 77-94. Freire, M. (1986) An inexpensive and interactive microcomputer system for codifying Golgi-impregnated neuronal morphology, J. Neurosci. Methods, 16: 103-117. Glaser, E.M., Gissler, M. and Van der Loos, H. (1979) An interactive camera lucida computer microscope, Soc. Neurosci. Abstr., 5 503. Glaser, E.M., Tagamets, M., McMullen, N.T. and Van der Loos. H. (1983) The image-combining computer microscope - an interactive instrument for morphometry of the nervous system, J. Neurosci. Methods, 8: 17-32. Glaser, E.M. and Van der Loos, H. (1980) Computer microscope apparatus and method for superimposing an electronically produced image from the computer memory upon the microscope field of view. U.S. Pat. 4,202.037, U.S. Patent Office, Washington, D.C. Grant, G. and Boivie, J. (1970) The charting of degenerative changes in nervous tissue with the aid of the electronic pantograph device, Brain Res., 21: 439-442. Hillman, D.E., Canaday, M., Mahoney, G. and Crank. R.
173 (1987) High resolution brain mapping, Soc. Neurosci. Abstr.. 13: 568. Mize, R.R. (1984) Computer applications in cell and Neurology: a review, Int. Rev. Cytol., 90: 83-124. Mize, R.R. (1985) A microcomputer plotter for use with light and electron microscope. In R,R. Mize (Ed.), The Microcomputer in Cell and Neurobiology Research, Elsevier, Amsterdam, pp. 177-215. Nottebohm. F.. Stokes, T.M. and Leonard, C.M. (1976) Central control of song in the Canary, Serinus canarius, J. Comp. Neurol., 165:457 486.
Patterson, H.A., Warr, W.B. and Kleinmann, A.J. (1976) A mapping device for attachment to the light microscope. Technical note, Brain Res., 102: 323-328. Schmued, L. and Fallon, J.H. (1986) Fluoro-gold: a new retrograde axonal tracer with numerous unique properties, Brain Res., 377: 147-154. Williams, F.G. and Elde, R. (1982) A microcomputer-aided system for the graphic reproduction of neurohistochemical maps, Comput. Programs Biomed.. 15: 93-102.
165
Elsevier NSM 00858
Simple microcomputer system for mapping tissue sections with the light microscope Arturo Alvarez-Buylla and David S. Vicario The Rockefeller Unit,er~itv, New York, N Y 10021 (U.S.A.)
Received 8 December 1987) (Revised 7 March 1988) (Accepted 12 May 1988)
K e y words: C o m p u t e r - m i c r o s c o p e ; D r a w i n g - t u b e :
Pantograph;
M i c r o c o m p u t e r : Plotter: M o r p h o m e t r y :
Distribution; Neuroanatomy We describe a method for two-dimensional mapping of tissue sections that makes use of a drawing tube. microscope stage encoders and a microcomputer. The drawing tube views the graphics monitor and superimposes the image of the screen cursor and on-screen menus on the specimen image. Thus, the position of every landmark in each microscopic field can be mapped without stage movement while directly viewing the specimen through the microscope. A mouse is used for data entry and program control. Fields mapped in this way are then assembled into a complete map, which can include line drawings as well as up to 20 landmark types. The coordinate values of all landmarks mapped are stored and remain accessible for editing and analysis. High resolution plots are produced. Specialized functions include grain counting, area and perimeter calculations as well as a perimeter limiter that predefmes the area to be mapped. The system uses general purpose hardware that is widely available. Man 3' hitherto time-consuming tasks, such as detailed mapping of cell positions, regions of immunocytochemical staining, degenerating fibers, neuronal connections or any other anatomical feature, can be done in a fraction of the time and effort previously involved. These labor savings can be realized while maintaining the highest resolution and enabling statistical analysis since the data are already in digital form.
Introduction E x a m i n a t i o n of tissue sections u n d e r the light m i c r o s c o p e frequently requires the use of highp o w e r objectives to visualize details of interest (e.g. cell types, a u t o r a d i o g r a p h i c grains, i m m u n o histoc h em i cal stains or d e g e n e r a t i n g fibers). T h e spatial relationship of such details to each o th e r and to the rest of the section is often lost due to the small field size. It is i m p o r t a n t to be able to m a p fine details (referred to here as ' l a n d m a r k s ' ) in relation to the whole section, especially in studies of regional specialization of the brain.
Correspondence: A. Alvarez-Buylla, The Rockefeller Univer-
sity, 1230 York Ave., New York, NY 10021, U.S.A.
C o m m o n but expensive and t i m e - c o n s u m i n g m e t h o d s for m a p p i n g include the c o n s t r u c t i o n of p h o t o g r a p h i c m o n t a g e s or the p o s i t i o n i n g of l a n d m a r k s by h a n d on a low m a g n i f i c a t i o n picture or d r a w i n g of the region to be m ap p ed . It is easier and m o r e efficient to use the position of the m i c r o s c o p e stage to drive the X and Y axes of a plotter. Several such systems have been developed, such as the electronic plotters (Boivie et al., 1968: G r a n t and Boivie, 1970: Eidelberg and Davis, 1977) or the light p a n t o g r a p h (Patterson et al., 1976). R e c e n t c o m p u t e r - a s s i s t e d m a p p i n g systems (Curcio an d Sloan, 1981; reviewed in Mize 1984, 1985) o f t en m a k e use of specialized h a r d w a r e and frequently require that each feature of interest be aligned u n d e r a cross-hair in the m i c r o s c o p e eyepiece ( F o r b e s and Petry, 1979: Williams and Elde, 1982; Mize, 1985). L a n d m a r k s may bc
0165-0270/88/$03.50 (~" 1988 Elsevier Science Publishers B.V. (Biomedical Division)
166 visualized throughout the microscope field, but, because they are only recorded by the computer when they fall under the cross-hair, an excessive amount of stage movement is necessary. In an ideal system, the operator would position a pointer visible through the microscope on a feature of interest anywhere in the field, and enter its position in the computer. While this can be accomplished with digitized video images displayed on a graphics terminal (DuVarney and DuVarney, 1985), such systems are expensive and do not approach the resolution of direct visualization through the microscope. This principle has been better realized in the form of image-combining systems that use a camera lucida to superimpose the tissue image with an image of a computer-controlled graphics device (Glaser et al., 1979, 1983; Glaser and Van der Loos, 1980). In this report we describe a simple system for two-dimensional mapping based on this principle. It makes use of common microcomputer hardware and a drawing tube to view a mouse-driven cursor on the computer screen, while two digital indicators relay the microscope stage position to the computer. Software developed for this system thus allows rapid entry, display and plotting of the positions of cells and fibers (or any landmark) throughout the section while looking through the microscope.
Methods and Materials
Hardware The standard hardware consisted of an IBM PC-XT or PS/2, Model 30 with two serial ports, one parallel port, Hercules graphics card, monochrome monitor, mouse (Microsoft) and plotter (Hewlett-Packard models 7470A or ColorPro). A related system is under the development for Macintosh microcomputers (C. Ghez, personal communication). Specialized equipment consisted of two digital linear position sensors (Mitutoyo Digimatic) with a serial interface (Mitutoyo Multiplexer 10) (Fig. 1). These provided X and Y positional coordinates of the microscope stage to the computer. The indicators were mounted on the arm of a Nikon Labophot microscope or directly
Fig. 1. Mapping system components. The X stage encoder ~s hidden behind the Y encoder (y). The encoders ~ransmit the stage position to the interface (i) that relays this information to the computer. The drawing tube (arrow) is oriented towards the computer screen, allowing direct observation through the microscope of the mouse driven cursor and menus. onto a Zeiss microscope stage with custom-made brackets (Fig. 1). The X indicator has a range of 2.5 cm and reads the short axis of a 2.5 x 7.5 cm slide. The Y indicator has a 5 cm range, sufficient to cover the non-frosted portion of the slide. A camera lucida (drawing tube; Nikon) was modified to visualize the computer screen superimposed on the image of the tissue section. The end mirror was oriented horizontally instead of in the manufacturer's fixed vertical position (looking down at the table). Camera lucidas l'rom other companies are equally easy to adapt. With the drawing tube and monitor aligned, the screen and mouse cursor were clearly visible on the specimen image, enabling interactive data entry while scanning sections (Glaser et al., 1983). In cases where the camera lucida optics attenuated too much of the light emitted from the tissue, e.g. viewing fluorescently-labeled landmarks, positioning was accomplished by matching the location of landmarks in a reticle grid to corresponding positions in an identical grid displayed on the terminal.
Software Computer programs were written in BASICA (IBM) for graphics equipped computers. One version takes advantage of text cursor positioning commands to provide a pseudo-graphics display and can run on computers without graphics
167 h a r d w a r e . T h e p r o g r a m s w e r e d e s i g n e d to a i d in t h e s y s t e m a t i c s c r e e n i n g o f t i s s u e s e c t i o n s . U p to 20 d i f f e r e n t l a n d m a r k t y p e s o r l i n e d r a w i n g s c a n be selected from a screen menu by means of the mouse. The landmarks or lines are then displayed on the screen, superimposed on the image of the section. In this way, the data entered on the image o f a s i n g l e field c a n b e e d i t e d b e f o r e b e i n g s t o r e d a n d p l o t t e d . T h e X - Y c o o r d i n a t e s of t h e c e n t e r o f t h e field a r e t r a n s m i t t e d t o t h e c o m p u t e r b y pressing the 'load' button of the multiplexer interface. T h i s also s i g n a l s t h e p r o g r a m t o s t o r e a n d p l o t t h e i n f o r m a t i o n in t h e c u r r e n t field a t t h e appropriate X Y location. To insure that fields or l a n d m a r k s a r e n o t c o u n t e d twice, s e v e r a l c h e c k points are built into the program. Erroneously
e n t e r e d p o i n t s c a n b e e r a s e d . T h e f i n a l o u t p u t is a plot of the mapped histological section, showing all t r a c i n g s a n d l a n d m a r k s e n t e r e d . All t h e d a t a a r e s t o r e d o n t h e h a r d d i s k , a n d c a n b e r e c a l l e d t o m a k e p l o t s of s e l e c t e d l a n d marks. With a separate program, the map can be r e d r a w n o n t h e c o m p u t e r s c r e e n : t h e m o u s e is t h e n u s e d t o select s c r e e n p o s i t i o n s , a l l o w i n g dist a n c e s , a r e a s a n d p e r i m e t e r s to b e c a l c u l a t e d .
Procedure The mapping procedure consists of using the m o u s e to e n t e r t h e l o c a t i o n o f l a n d m a r k s o r o t h e r f e a t u r e s in t h e c o r r e c t p o s i t i o n w h i l e v i e w i n g t h e s e c t i o n a t h i g h m a g n i f i c a t i o n ( F i g . 2). P o s i t i o n w i t h i n e a c h field is d e t e r m i n e d as t h e m o u s e ' s
Fig. 2. Superimposed computer screen (screen field) and specimen image. The screen field and menus are presented here unrotatcd, though in the microscope they appear to be rotated by 90 ° clockwise. A: the positions of two neurons and a glial cell with autoradiographic grains overlaying their nuclei are marked by mnemonic screen characters (faintly seen in this photograph). Fhc mouse-driven cursor (arrow) serves as a pointer to designate the position of cells to be marked by pressing one of the two mou.',c buttons. In this case, each mouse button is loaded with neuron (NEU) or gila (GEl) as indicated in the highlighted landmark menu to the right of the field. The menu to the left contains different editing functions. Del, clears the field: Dpf, deletes previous field: Ins, inserts into previous field: Rvi, recovers previous field, chk, activates double landmark checking subroutines; sss. silences computer codes: S:P, skip plotting for this field. The menu at the bottom contains: 1, enters border in center of the field: 2, activates subroutine to find previous field; 3 and 4, pantograph trace functions; 5, records position where photo was taken. B: line drawing of radial gila fibers (see Fig. 4). The mouse is loaded with the 'LIN' function (highlighted on the menu to the right) and the cursor ix used to trace the fibers on the field. The second button on the mouse was loaded with "ERA', used to erase incorrect landmarks from the screen
168
15 A2 :
A
08-09-1987, 23= ;)9 /^
/
o
o
o
/
o~-..--_
/-'-
-.
•
(
X
..-
/
."
I××
/
o
-
~
/
Y
o NEURO* ;)5 • GLIA * 36 x P]CNOT ]B
..y/
o:, o
SPY-MIG
~
0
~...
xl
~-
,
D
. v
B
~
:
y
~ . ' :-:'-" "-" •
°
oO,
°
° •
,,
•
°
•
,s.,
o
,0a.
." '.~ , 0 % o ,
t 0.Sin
~j
•"
°
"
-s.
;
•
o
""'~A",.'". ~ 0°'s
v•', ,..
,°
;"
°
° *
..........
°
°°~
°%
• , .,....:....:..-,o.~. :..~. -7, °
~
.o. FLUI~O 55 • f.ot-nr 4gl]
%
•
.
°
°oa
°
;.
°
°
".-",,"~: :. ..':_,-..": ;.;y
~ "
".:": " . " '
' " ": ¢/
PIFRIIETER
E5211
.356 m =
Fig. 3. Examples of program output. A: m a p of radiolabeled cells in the canary telencephalon (see example 1 of text). The upper right corner is the legend with file name (A152A2W), date, time, program name, landmark counts and scale bar. The small rectangle to the right of the m a p illustrates the proportional field size as viewed through the microscope while mapping. B: maps of fluorogold-labeled cells (upper panel) and total neurons (lower panel) in nucleus HVc of the canary brain (see example 2 of text). The dashed line is the surface of the brain and the dotted line is the ventricle. The region to be mapped was pre-established using the perimeter-limiter and demarcates the borders of HVc (solid line). This area and the perimeter length are shown in an additional plot legend.
169 cursor screen position. T h e a b s o l u t e l o c a t i o n of each field is d e t e r m i n e d from the p o s i t i o n a l indicators which r e c o r d m o v e m e n t of the stage a n d slide. In o r d e r to m a p an entire section or region, successive fields need to be m a p p e d a c c o r d i n g to a s c a n n i n g strategy. The p r o g r a m allows a choice b e t w e e n s c a n n i n g parallel to the X axis, the Y axis or a r a n d o m trail. T h e best choice d e p e n d s on the size, shape a n d o r i e n t a t i o n of the s p e c i m e n as well as on the preference of the user. Initiation. A f t e r zeroing the indicators, the m i n i m u m a n d m a x i m u m X- a n d Y c o o r d i n a t e s are loaded. This a u t o m a t i c a l l y sets the best scale to m a x i m i z e the size of the plot. A reference p o i n t at
a k n o w n l o c a t i o n in the section is entered. This can be used later, if s u p e r i m p o s e d m a p p i n g is desired. Scale, field size, relative s y m b o l size, l a n d m a r k n a m e s a n d p l o t t i n g status are a u t o m a t i cally set b y the p r o g r a m , b u t can be m o d i f i e d by the user. A b o u n d a r y ( p e r i m e t e r limiter) can be e n t e r e d with a lower p o w e r objective that limits m a p p i n g to that area; this p e r i m e t e r limiter is especially useful when m a p p i n g details within a nucleus for which the b o r d e r s are not distinct at high m a g n i f i c a t i o n (Fig. 3B). A square or a rectangle (screen-field) set to m a t c h the size a n d s h a p e of the eyepiece reticle (eyepiece-field) a p p e a r s on the screen. A menu of
/
J l L..
i.
,
Fiq.
2B
Fig. 4. Map of radial gila fibers stained by immunocytochemistry (see example 3 of text) in a frontal hemisection of the rostral telencephalon of the canary brain. The midline is to the right and dorsal is up. The small rectangle inside the map indicates the field from which the photograph in Fig. 2B was taken. The solid line next to the midline is the partially collapsed lateral ventricle.
170
function codes appears to the left of the screen field and a menu of landmark names to the right. By adjusting the relative illumination of the microscope and the screen intensity, one can visualize the microscope field superimposed on the computer screen image (Fig. 2). Due to the optics of the drawing tube, the screen as seen through the microscope appears to be rotated by 90 ° To compensate for this rotation, the mouse is used sideways and serves as a pointer for the desired function or landmark. Mapping. Each button of the mouse is loaded with a landmark (e.g. neuron or glia) selected from the menu (Fig. 2). The cursor is then moved inside the screen-field to point to the desired feature of the superimposed microscope image. The appropriate landmark symbol is placed at this position on the screen by pressing the correct mouse button. In case of error, one can erase individual landmarks or clear the whole field. When all landmarks have been entered, the completed field is loaded, signaling the computer to calculate the absolute coordinates of all the landmarks, plot them and store them on the disk. The stage is then moved to view the adjacent field. The program signals if a field overlaps a previously loaded field or if the perimeter limiter (see above) has been violated. The program checks that all fields are entered at the right locatiom preventing the user from accidentally jumping into adjacent rows. If the landmarks are densely distributed, every contiguous field must be mapped. If they are more scattered, only those fields that contain landmarks are mapped. In addition to entering tissue landmarks which map to a point, more complex structures (e.g. degenerating fibers, axons or dendrites) can be entered as line drawings. This can be done in two ways. (1) Within a field, the mouse can be used to trace the structure of interest, using the ' L I N ' function (Figs. 2B and 4). (2) For larger structures, extending beyond a single field, the tracing mode can be used to draw lines (Figs. 3 and 4). In this mode the microscope and computer work like a pantograph: the plotter follows the movement of the stage, using the color pen and line type selected. This is useful to draw the contour of the section, laminae, long fibers, large blood vessels or the ventricles.
Termination. All plots receive an identifying legend, including the filename, date and time, a scale bar, counts of landmarks, scanned area and. when using the perimeter limiter, the area and perimeter (Fig. 3). All plotting data are stored in a random access file enabling the plot to be repeated, modified or completed.
Examples Example 1: mapping the entire section. Dividing cells were labeled in a one-year-old male canary by two injections of [3H]thymidine. Forty days later the brain was fixed and sectioned at 10 ~m. The sections were incubated for autoradiography, stained with Cresyl violet, dehydrated and coverslipped. A section at the level of the anterior commissure is shown in Fig. 3A; the positions of [3H]thymidine-labeled cells were mapped only within the telencephalon. A 63 × objective (Zeiss) was used to recognize and classify labeled cells. The procedure was as follows. (1). The contour of the brain's lateral ventricle was drawn with the trace function (pantograph), represented by the heavy solid line within the section. Using the same function, the dashed line was drawn to show the hyperstriatal lamina. (2) Mapping began at the top of the section and proceeded to the bottom in a zigzag pattern of horizontal scanning rows. Border landmarks were entered at the ends of each scanning row, and the program automatically joined these points to form the solid line of the section perimeter. (3) Within each scanning row, fields were marked with the position and type of labeled cells encountered there. The program put together the fields to produce the final map (Fig. 3A). Example 2." mapping part of the section. Neurons in the hyperstriatum ventralis pars caudalis (HVc) of an adult male canary were labeled with fluorogold (Schmued and Fallon, 1986) by retrograde transport from area X. Brains were embedded in P E G and 6 /xm thick sagittal sections were cut and mounted onto glass slides. HVc is just below the lateral ventricle of the caudal forebrain and plays an important role in song production (Nottebohm et al., 1976). An easy-to-find reference point was selected for later use. (2) The perimeter limiter was entered around the fluoro-
171
gold-positive region (Fig. 3B, solid line) using a 20 × objective. (3) The position of the individual fluorescent neurons was then mapped systematically with a 63 × objective; a series of fields was completed from left to right, then down one field and from right to left and so on. (4) The lateral ventricle was entered with the tracing function as a dotted line and the surface of the brain as a dashed line (Fig. 3). The section was then stained with Cresyl violet. dehydrated and coverslipped. This procedure destroys the fluorogold fluorescence, but makes it possible to recognize all neurons by their Nissl staining. (5) The stained section was again repositioned under the microscope using the selected reference point and all neurons within the HVc were then mapped (Fig. 3B). After staining, the borders of HVc are difficult to find especially when using a high power objective (63 x ). However, since before staining the perimeter limiter was determined while mapping the fluorescent cells, neurons can now be entered in and around the HVc without having to worry whether or not they are inside of the HVc. The program rejects all neurons found outside the perimeter limits. The output is a map of fluorogold-labeled neurons and total neurons (Fig. 3B). The area and perimeter are calculated and printed in the legend. Example 3: drawing lines. P E G sections (10 > m thick) of canary forebrain were stained with a monoclonal antibody (Alvarez-Buylla et al., 1987). In the adult avian brain this antibody recognizes radial glia with their cell bodies on the walls of the lateral ventricle and thin processes that course laterally into the forebrain. The stained glial fibers are clear only at relatively high magnification (40 x objective), hence a given field contains just a small fraction of the section. A hemisection of anterior canary forebrain is used here to illustrate the procedure (Fig. 4). (1) The ventricle and the section borders were traced (pantograph, see above) using two different pen colors (both shown as black in Fig. 4). (2) Starting at the top, each horizontal row was completed before moving one field down to the next row. Each fiber seen in the microscope field was traced by using the mouse (in ' L I N ' mode) to move the
cursor on the screen field. The program assembles all the fiber drawings into a complete map (Fig. 4). With more conventional methods, i.e. photographic montage or mapping field-by-field by hand into a grid, it often took up to two weeks to complete a map of one hemisection. By using this system the hemisection shown here (Fig. 4) was mapped in one day.
Discussion A powerful feature of this system is that points are mapped while directly viewing the tissue through the microscope at the optimum magnification, using a camera hicida. Other systems, designed mainly for neuronal reconstruction, have used a similar principle (Glaser et al., 1979: Capowski and Sedivec, 1981; Glaser et al., 1983, Capowski, 1985, Freire, 1986). By assembling a complete map from individual fields, the absolute locations of cellular features can be stored at high resolution (1 /,m) for tissue sections up to 2.5 cm in size. These are important advantages when compared to the digitized video images with a superimposed cursor that are commercially available. While recently developed high resolution video systems (4096 x 4096 pixels), using expensive specialized hardware and software, may provide improved resolution (Hillman et al., 1987), the images they generate require large amounts of mass storage space. Our program only records landmarks of interest and their coordinates, since it is not necessary to store the entire image for many common applications. The most complex example illustrated (Fig. 4), needed only 70 kilobytes of disk space. The mapping resolution in our system is determined by two parameters: (1) the stage position indicators have a 1 > m resolution, and (2) the resolution of the screen. The effective size of the pixel image seen through the microscope depends on the objective used. This means that at high magnification, the effective screen resolution is equal to or better than the stage resolution of 1 /~m, which then sets the limit. At lower magnification, the screen resolution is the limiting factor, reducing resolution to about 2 or 4 ~tm at 40 x or
172 20 x , respectively. The m a p resolution is limited by the n u m b e r of a d d r e s s a b l e p o i n t s in the screen-field of the c o m p u t e r m o n i t o r a n d the plotter. In our experience, the graphics m o d e of the screen exceeds w h a t is required for most a p p l i cations. A slight error is i n t r o d u c e d because of the curvature of the screen, m a k i n g the sides of the screen-field a p p e a r slightly curved when o b s e r v e d through the m i c r o s c o p e (Fig. 2B). The c a l c u l a t e d m a x i m u m d e v i a t i o n is 0.8% of the size o f the field. This could only b e of significance when w o r k i n g with a very low m a g n i f i c a t i o n objective. F l a t screen m o n i t o r s which would solve this p r o b l e m have recently b e c o m e available a n d a f f o r d a b l e for personal computers. A n e n o r m o u s a d v a n t a g e of c o m p u t e r - b a s e d p l o t t i n g m e t h o d s is the a b i l i t y to edit a n d analyze this m a p during a n d after d a t a collection (Mize, 1985). This includes calculating areas, distances a n d d i s t r i b u t i o n statistics, r e p l o t t i n g only p a r t of the file or m o d i f y i n g the plot. M a p p i n g of a u t o r a d i o g r a p h i c a l l y l a b e l e d cells, a c o m m o n a p p l i c a t i o n in our l a b o r a t o r y , was originally d o n e b y d r a w i n g c o l o r p o i n t s into a previously taken picture (or pictures) of the tissue section. This was tedious a n d time consuming, taking f r o m 4 to 5 days to c o m p l e t e a section of a p p r o x i m a t e l y 5 × 5 ram. W i t h the c o m p u t e r - a i d e d m a p p i n g system described here, one such m a p can be c o m p l e t e d in less t h a n a day, T h e same c o m p u t e r can also be used for other tasks. The specialized h a r d w a r e connects to a s t a n d a r d serial port, a n d costs less than $1000 at this time. Once installed, this system can be used by l a b o r a t o r y p e r s o n n e l with no specialized c o m p u t e r knowledge. T h e p r o g r a m s m a y be o b t a i n e d by c o n t a c t i n g the authors. The software is written in B A S I C A and is easily c u s t o m i z e d for new applications.
Acknowledgements T h e s u p p o r t o f P H S grants M H 18343 (to F e r n a n d o N o t t e b o h m ) a n d M H 40900 (to D.V.) is gratefully a c k n o w l e d g e d . The w o r k was also supp o r t e d in p a r t b y N I H g r a n t B R S G S07 RR07065, a w a r d e d to Rockefeller University. W e t h a n k
F e r n a n d o N o t t e b o h m for his enthusiastic s u p p o r t for this project, a n d for his helpful c o m m e n t s ~n this m a n u s c r i p t . W e also t h a n k M a r g a Theelen for her help with the m a p p i n g a n d Jud~ Dye for secretarial assistance.
References Alvarez-Buylla, A., Buskirk, D.R. and Nottebohm, F. (1987) Monoclonal antibody reveals radial glia in adult avian brain, J. Comp. Neurol., 264: 159-170, Boivie. J., Grant, G. and Utfendahl, H. (1968) The X-Y recorder used for mapping under the microscope, Acta Physiol. Stand., 74: 1A-2A. Capowski, J.J. (1985) The reconstruction, display, and analysis of neuronal structure using a computer. In R.R. Mize (Ed.), The Microcomputer in Cell and Neurobiology Research, Elsevier, Amsterdam, pp. 85-109. Capowski, J.J. and Sedivec, M.J. (1981) Accurate computer reconstruction and graphics display of complex neurons utilizing state-of-the-art interactive techniques, Comp. Biomed. Res., 14: 518-532. Curcio. C.A. and Sloan, K.R. (1981) A computer system for combined neuronal mapping and morphomet~, J. Neurosci. Methods, 4: 267-276. DuVarney, D. and DuVarney, R.C. (1985) A computer-based video microscope for cell measurement. In R,R. Mize (Fxt.). The Microcomputer in Cell and Neurobiology Research, Elsevier, Amsterdam, pp. 233-246. Eidelberg, E. and Davis, F. (1977) An improved electronic pantograph, J. Histochem. Cytochem.. 25: 1016-1018. Forbes, D.J. and Petry, R.W. (1979) Computer-assisted mapping with the light microscope, J. Neurosci. Methods, 1: 77-94. Freire, M. (1986) An inexpensive and interactive microcomputer system for codifying Golgi-impregnated neuronal morphology, J. Neurosci. Methods, 16: 103-117. Glaser, E.M., Gissler, M. and Van der Loos, H. (1979) An interactive camera lucida computer microscope, Soc. Neurosci. Abstr., 5 503. Glaser, E.M., Tagamets, M., McMullen, N.T. and Van der Loos. H. (1983) The image-combining computer microscope - an interactive instrument for morphometry of the nervous system, J. Neurosci. Methods, 8: 17-32. Glaser, E.M. and Van der Loos, H. (1980) Computer microscope apparatus and method for superimposing an electronically produced image from the computer memory upon the microscope field of view. U.S. Pat. 4,202.037, U.S. Patent Office, Washington, D.C. Grant, G. and Boivie, J. (1970) The charting of degenerative changes in nervous tissue with the aid of the electronic pantograph device, Brain Res., 21: 439-442. Hillman, D.E., Canaday, M., Mahoney, G. and Crank. R.
173 (1987) High resolution brain mapping, Soc. Neurosci. Abstr.. 13: 568. Mize, R.R. (1984) Computer applications in cell and Neurology: a review, Int. Rev. Cytol., 90: 83-124. Mize, R.R. (1985) A microcomputer plotter for use with light and electron microscope. In R,R. Mize (Ed.), The Microcomputer in Cell and Neurobiology Research, Elsevier, Amsterdam, pp. 177-215. Nottebohm. F.. Stokes, T.M. and Leonard, C.M. (1976) Central control of song in the Canary, Serinus canarius, J. Comp. Neurol., 165:457 486.
Patterson, H.A., Warr, W.B. and Kleinmann, A.J. (1976) A mapping device for attachment to the light microscope. Technical note, Brain Res., 102: 323-328. Schmued, L. and Fallon, J.H. (1986) Fluoro-gold: a new retrograde axonal tracer with numerous unique properties, Brain Res., 377: 147-154. Williams, F.G. and Elde, R. (1982) A microcomputer-aided system for the graphic reproduction of neurohistochemical maps, Comput. Programs Biomed.. 15: 93-102.