Digital diaphanoscopy in the diagnosis of maxillary sinus diseases for patients with different anatomical and gender features

Abstract

Previous studies conducted by digital diaphanoscopy have shown the need to select the optimal value of the radiation source optical power for each individual patient based on their gender and anatomical features.
Within the framework of this study, a group of conditionally healthy volunteers was formed taking into account differences in gender, age and anatomical features. Preliminary experimental studies were carried out on 30 conditionally healthy volunteers, including 20 females and 10 males. At the same time, the study group consisted of persons aged 18 to 27 years with a body mass index (BMI) value within the normal range (18.5-24.9), pre-obesity (25-29.9), and obesity of the 1st degree (30-34.9).
During the study, the values of the brightness control unit were selected sequentially from 0 to 250 a.u. in increments of 50 a.u. Each value of the controller corresponded to the value of the optical power of LEDs of two wavelengths (650 nm and 850 nm), set using the control panel for the PM400 power meter (Thorlabs, Inc.) and the S120VC photodiode sensor (Thorlabs, Inc.).
The obtained results showed the correlation of the selected value of the optical radiation power with the gender of the volunteer. Therefore, in the study of female volunteers, the maxillary sinuses were well visualized at the controller value of 50-100 a.u. (35-40 mW for 650 nm and 25-60 mW for 850 nm) and a camera exposure time value of 40 ms. In the study of male volunteers, the required value of the controller was equal to 200-250 a.u. (55-60 mW for 650 nm and 84 mW for 850 nm wavelength), while increasing the value of the camera exposure time to 60 ms.
For further analysis of the registered patterns of light scattering, a literary review was conducted in the field of medical image processing, in particular, analysis of algorithms for the classification of sinus pathologies in digital diaphanoscopy. Thus, an algorithm for the analysis of registered scattering patterns of light was formed, consisting of 2 stages.
At the first stage, as a decisive sign for dividing the group of subjects into two classes (healthy, with pathology), the intensity parameter (in % ratio) was used, characterizing the amount of light reaching the camera detector after absorption by biological layers and various pathologies.
At the moment, 30 conditionally healthy volunteers and 4 patients with cystic fluid were studied, for whom this parameter was determined. Such analysis showed that in the case of a sinus without pathology, the intensity parameter was from 87% to 100%, for sinuses with cystic fluid – from 10% to 44%.
The second stage of the analysis was to calculate the coefficient K, which contributes to the classification of patients into groups depending on the type of pathological change (pus, cystic fluid, or tumor tissue). This coefficient is the percentage ratio of the intensity in the eye socket area to the intensity of the maxillary sinus. At the same time, the intensity values were obtained using an open-source program for image analysis and processing ImageJ.
This stage revealed that in the case of conditionally healthy volunteers, the coefficient K ranged from 0.2 to 2.9%, and in the case of a sinus with a cystic fluid – from 5.5 to 28.2%.
The further set of experimental data on patients with various maxillary sinuses diseases will allow us to clarify the limits of the calculated parameters characteristic of each specific disease of the maxillary sinuses.
Thus, based on the developed classification algorithm, which will allow for a quantitative assessment of pathological changes, methodological recommendations will be developed for the diagnosis of maxillary sinus pathologies in all groups of patients.
The reported study was funded by RFBR according to the research project No. 20-32-90147.

File with abstract

Speaker

Ekaterina Bryanskaya
Orel State University named after I.S. Turgenev
Russia

Discussion

Ask question