News |  Sitemap |  Contact
PDF Export  | 

N.A. Englevskiy: Automatization of the process of examination of tracheostome's size, using the ImageJ

Abstract

Nowadays there are two main methods used in otorhinolaryngology for forming the non-tube tracheostome. One of them includes different plastic methods, using local tissue and different types of stitches made with traditional suture materials. The other one is based on peritracheal implanting of various prostheses to create a rigid frame of tracheostome's walls.

While choosing the method the size of tracheostomes area is of the most importance. According to methods used in clinical practice there is a certain division: - if tracheostomes area is less than 2 square cm, then the plastic method is used, if it is 2 square cm or more the second method is used.

Earlier to find tracheostome's area the formula was used:

S=πxAxB/2, where S - tracheostome's area (square cm), B - width of a tracheostome (cm). Tracheostome's measurement was made using trammel, so the measurement was dependent on the doctor's or a laboratorian's subjective opinion.

This work's goal was to create macros for the ImageJ program, which could measure the area of tracheostome automatically and based on the measurements would give recommendations on which method of non-tube tracheostomes forming to use.

We were making photographs of tracheostomes together with a special patch next to it. The patch had two circles drawn on it. The space between their centers was 1 cm. After the photo was taken, the pictures were transferred to the computer for special analysis through the ImageJ program. Especially for this procedure we have written macros “AreaMeter.Tracheostomy”, all the demands of SSMU ORT specialists were taken into account so the macros can analyze the area of tracheostome and give recommendations on the choice of method.

Macros was realized with the help of ActionBar plugin, which in this case has five buttons - opening of the image, highlighting the area needed, making a copy of a highlighted area, choice of the Multipoint instrument and general processing of the image.

The algorithm of working with this macros is following: first, one higlights a certain area on the image, one then makes a copy of the area highlighted and smoothes the background using command run(“Subtract Background…”, “rolling#0 light”). It is necessary to do it to allocate the space between dots which are of the most importance when the matching of the space in pixels to the space in centimeters takes place. After matching the measuring scale to the centimeter ruler, the image is transformed into 8-bit, Gauss filter is used and with the help of with threshold «Minimum» the highlight of the certain area is being made. Next is calculation of tracheostome's area.

After the work with macros is finished, the user sees an informative message. If the user thinks that the highlighting of the area was done correctly, he pushes an OK button, and the data is being saved into the data base of the patients. If one sees on the image that the highlighting was done incorrectly, he then pushes Cancel button and the image is not being saved. In this case the user can choose another image of the patient, because in the process of tracheostome's measurement several photos are made.

Conclusions. ORT specialists of SSMU (Tomsk) tested this macros and claimed that it is more précised than manual measurements done with trammel. When measuring tracheostome's area with the help of this macros, there is no subjective doctor's opinion. All the images have the same error index, which leads to more accurate choice of the method non-tube tracheostomes forming.

Keywords

choice of method of operation, non-tube tracheostoma, otorhinolaryngology, ActionBar

Administrative data

Presenting author: Englevskiy N.A.
Organisation: Siberian State Medical University

co-authors: Simonov S.V., Starokha A.V., Shilov B.V., Davydov A.V.

© Luxembourg Institute of Science and Technology | Legal Notice