Wijbenga, Peter (2010) Het ritme van Benchmark Ultra : Algoritme ontwikkeling voor het optimaal vullen van immunokleuring machines bij de afdeling Pathologie. thesis, Other studies (UMCG).
![[img]](https://umcg.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
Scriptie_P._Wijbenga.pdf Download (518kB) |
Abstract
As task for my graduation at the NHL University Leeuwarden, I have tried to find a worthy solution for a problem that arose at the Pathology department from the Universal Medical Centre Groningen. This department had purchased for colouring machines called Benchmark® Ultra, just before I was assigned to this task. These colouring machines can automatically do a part of immunohystological research by bringing a colour to some tissues. The problem of these machines is that they have to be classified by hand. A problem occurs because of the limited space for dispensers and coupes. A dispenser is a small tube with reagents, the necessarily substance with antibodies to do the colouring. A coupe is a piece of glass containing a sliced tissue (2 to 3 micrometers thin), made for histological research. Each machine has place for 30 coupes and 35 dispensers. Standard reagents take 8 of these 35 dispenser positions. Besides this, some colours take more than one position. It depends on these positions how many places remain to place other dispensers to realize more colours at the same time. I succeeded to write a well working algorithm, taking all requirements for the classification into account. Anyhow I could not fulfil all wishes of the Pathology department. The pathologists have to count the coupes of every colour for themselves if my algorithm is included in a computer application, because the application has to know of each colour how many coupes are presented. They can read the classification from the display by pressing a button, after they have typed the number of coupes per colour in the application. Counting these coupes takes a lot of time. A solution therefore is to scan each barcode from every coupe. It would save a lot of time if the pathologists could use an external scanner to read the number of coupes per colour into the application. They can save even more time by sending the number of coupes to the application from the first request for colouring. For this the application will have to remember how many coupes of each colour are sent. This requires still a lot of work, but is certainly feasible for the near future! The pathologists would like to have as many as colours on one given machine, to avoid dragging the dispensers over and over again. This will also ease loading the coupes after each other, what is more likely for the future. I have not calculated on which machine the colours that appear frequently can be classified best, because there was not enough relevant data presented. Since there are no existing solutions for a similar classification problem, I had to write the whole algorithm by myself. With smart thinking and logical reasoning, I have made a well working algorithm. The algorithm was tested with normal / actual values and the result shows a good classification of the machines! The final algorithm is both written in storyline and in algorithm language. You can find this algorithm in Chapter 5. The algorithm in storyline is written to support the real algorithm. The real algorithm is written in a sort of intermediate language; a language that is generated from programming source code, but that cannot be directly executed by the computer. I can recommend the Pathology department to let my algorithm be processed in a computer application by a programmer. If this works, then I recommend to try using a scanner to read the coupes into the program, because this will save the employees a lot of time . This will mostly satisfy the needs of the department. Next will be to calculate a (more) fixed classification for each machine. In the future, it may be even possible to merge a good computer application with the user interface of the machines!
| Item Type: | Thesis (Thesis) |
|---|---|
| Supervisor name: | Goudswaard, ir. A.P. |
| Supervisor name: | Wiering, K.J. and Litjens, M. |
| Faculty: | Mathematics ∧ Natural Sciences |
| Keywords: | instellingen, Kleuringsapparatuur, weefselcoupes, Pathologie |
| Date Deposited: | 25 Jun 2020 10:46 |
| Last Modified: | 25 Jun 2020 10:46 |
| URI: | https://umcg.studenttheses.ub.rug.nl/id/eprint/717 |
Actions (login required)
 |
View Item |