During this work three goals have to be taken in mind. First to understand if the VHD-MMS site is quite enough known and at which hits rate it accomplish. Second if the visitors are comnig from everywhere or are located in the Continental Europe. Last how this traffic can impact on that on the NLM site.

      The first graph generated (Figure 2) show that, after a logical rapid growth at the beginning of its life, hits can be considered static, inside limited variations, during all the last 9 months. This data is very important, because show that people read one time general sections of a site, and further ones go to the specific sections, in our case go to use the FTP download facility. It means that we must build new services to invite people to visit our web site more than once. This kind of work has begun putting on line a Visible Human Project hypertextual tutorial (Figure 3), and some more.

     On the other hand, more than 14,000 hits reached in 13 months, on a total of 26,000, come from Continental European countries, and about 4,000 from no geographic domains, i.e. .net, .gov, .com, .mil and .org. If we consider that other 6,000 are undefined, the percentual of extra european users that have visited the "VISIBLE HUMAN DATASET - MILANO MIRROR SITE^®" Web is very little (Figure 4).

     We have also monitored how many people has gone to the VHP web site starting from our pages. The graph (Figure 5) seems very similar to that showing the VHD-MMS accesses. This could mean that at least 1 VHD-MMS web site visitor on a total of 3 go to the NLM VHP one.

     Another data we have analyzed is, month by month, the number of downloads of the VHD licence request form, we have put on our site. We have thought to do it to best understand if the people visiting us are really interested in the project, i.e. like to download images and working with them on some local projects, or are only Internet casual navigators. From these data we have report a medium of more than 35 downloads one month (Figure 6), on a total this year of 497. This number, that could seem very little, has to be put in front of the number of researchers really needed these kind of images. When this rate is done, the number seems not more so little.

     Our research focus then has been put on FTP downloads.

     Some ore the questions we have put to ourselves. Have we reached an acceptable total number of single image download? And from which countries downloads are more frequents? Which kind of data have been downloaded? More fullcolor on radiological ones? More from male dataset or from the female one? How much is decreased the download traffic generated by Continental Europe countries users on NLM Ftp site?

     To give an answer to these questions we have generated some graphs based on FTP logs data collection between January and September 1998.

     Before all we have collected together the number of single slice download performed on the VHD-MMS FTP site (Figure 7). This graph show some aspects of this kind of collection. If we analyze the number of downloads in each month, we can annotate a linear growth of this number, with some high single peaks. If we analyze the total amount growth of downloads from January to September the growth is quite exponential. This one reach quote of 14,000 single image downloads.

      A further specific exam of the data (Figure 8) show us that these downloads have been done first of all from European Continental countries, with Italy (6046), France (3585), Germany (1219), Spain (841), and Netherlands (180) on the first five places. This graph show two other interesting data: some downloads from Japan (676) and from .us domain (38).

      Another graph built using NLM FTP site data (Figure 9) show how many download has been done from that side during a limited period, i.e. five months. In this graph is relevant that only few european countries are present on the top, and those are Germany, Italy and France (Figure 10) the most important users of VHD-MMS FTP service.

     Coming back to the period January-September 1998 and to the data concerning the VHD-MMS FTP site, we can show (Figure 11) how the male dataset have had about six times more downloads than the female one.

      Then we can also show (Figure 12) that radiological images are more downloaded that the fullcolor ones.

     Detail of the data can give us an overview about which kind of slices are downloaded more frequently. So we found that inside the Female Dataset Normal CT and MRI images are the higher downloaded ones, followed by the Fullcolor Abdomen (Figure 13 and Figure 13.1). And inside Male Dataset the higher (Figure 14 and Figure 14.1) are Radiological Normal CT and Fullcolor Head images.