Scaling and Orienting the X-rays


Before I could try to orient the AP and lateral X-rays to each other, I needed to resize them so they were in the same proportion to each other as the actual X-rays are. Not having access to the actual X-ray films, I relied on measurement made by the HSCA radiologists.

G.M. McDonnel measured the 6.5 mm fragment on the rear of the skull as "9.6 cm above the mid portion of the external occipital protuberance. 1 cm above the metallic fragment is a depressed fracture…" He doesn't state if he measures to the bottom edge of the fragment, or the middle, or if the 1 cm to the fracture is from the top of the fragment. Mantik's precise measurements confirmed that the fragment is 6.5 mm high. I found the X-ray could not be scaled so these measurements work out assuming the measurement was to one of the fragment's edges. However, if I assumed the 9.6 cm measurement was to the middle of the fragment, and the depressed fracture was 1 cm above that point, then everything worked out nicely. So, from the tip of the mid-portion EOP to the depressed fracture is 10.6 cm.

On the AP X-ray, David O. Davis measured the 6.5 mm fragment as "2.5 cm to the right of midline." Again, he didn't indicate if that was measured to the edge of the fragment. Mantik's measured this as also exactly 6.5 mm, so it was easy enough to use that as a reference to determine that Davis's 2.5 cm measurement was from the midline to the medial (inner) edge of the fragment.

I scanned enhanced versions of both X-rays from Assassination Science. Unfortunately, a given distance of pixels on the AP that measured 10.6 cm there, did not correspond to the expected 10.6 cm distance on the lateral from the EOP to the fracture. The AP was too large by a factor of 1.06.

Next I had to resize one or the other X-ray so that 1 cm measured on one is 1 cm measured on the other. At the bottom of the HSCA reproduction of the enhanced AP, it states, "SCALE 400 MICRONS PER PIXEL." The lateral X-ray lists, "424 MICRONS PER PIXEL." It would make sense that the lateral was displayed at a smaller scale because the head is larger in profile than the frontal view and the image needed to be shrunk to fit on the screen. And here was the exact scaling factor I was looking for. 424/400 equals 1.06.

I then scaled the lateral image so that is was larger by a factor of 424/400. Now the images were at the same scale so that 152 pixels equaled 10 cm on both.



Finally, I rotated the lateral image until various anatomical features such as the frontal sinuses, the bridge of the nose, and the petrous bone, came into close alignment. I settled on a 23 degree rotation. I then tweaked the vertical alignment until the two large metal fragments aligned perfectly.

The overall alignment is not perfect because the X-rays are not parallel but diverge from the radiation source. My orientation diagram's "A" and "D" lines need to be angled towards each other a bit. But I would never be able to get it perfect because the same principle applies to the lateral X-ray. But it is certainly close enough to run your finger across the page matching one feature to another.