Aim: To produce an accurate model of BCVA changes of postpterygium surgery according to its morphological characteristics by using the machine learning technique. Methodology. A retrospective of the secondary dataset of 93 samples of pterygium patients with different pterygium attributes was used and imported into four different machine learning algorithms in RapidMiner software to predict the improvement of BCVA after pterygium surgery.
Results: The performance of four machine learning techniques were evaluated, and it showed the support vector machine (SVM) model had the highest average accuracy (94.44% ± 5.86%), specificity (100%), and sensitivity (92.14% ± 8.33%).
Conclusion: Machine learning algorithms can produce a highly accurate postsurgery classification model of BCVA changes using pterygium characteristics.
MATERIALS AND METHODS: Two hundred retinal samples of right eye [57.0% females (n = 114) and 43.0% males (n = 86)] were selected from baseline visit. A custom-written software was used for vessel segmentation. Vessel segmentation is the process of transforming two-dimensional color images into binary images (i.e. black and white pixels). The circular area of approximately 2.6 optic disc radii surrounding the center of optic disc was cropped. The non-vessels fragments were removed. FracLac was used to measure the fractal dimension and vessel density of retinal vessels.
RESULTS: This study suggested that 14.1% of the region of interest (i.e. approximately 2.6 optic disk radii) comprised retinal vessel structure. Using correlation analysis, vessel density measurement and fractal dimension estimation are linearly and strongly correlated (R = 0.942, R(2) = 0.89, p
MATERIALS AND METHODS: Six sounds (broadband noise, rain, ocean, waterfall, Quranic chapters Al-Fatihah, and Yasin recitations) were calibrated at the intensity levels of 45, 50, 55, 60, 65, 70, 75, and 80dBA. The sounds were delivered through a pair of Sennheiser HD 280 Pro headphones connected to the Sound Blaster X-Fi Surround 5.1 Pro sound card. The long-term average of the sound pressure level over the time of recording (LAseq) was recorded using the 3M SoundPro Class 1 1/3 Octave RTA sound level meter (SLM). The desired intensity levels were obtained by making adjustments to the sound files via the Audacity® software.
RESULTS: All sound files were calibrated at the targeted levels as verified by the value of LAseq.
CONCLUSIONS: Calibration of audio files can be done using a free/open-source software, as all six sound files were successfully calibrated at the targeted levels of 45, 50, 55, 60, 65, 70, 75, and 80dBA. The calibration steps provided in this paper can be easily applied by other researchers for similar purposes, with precautions when calibrating at low levels.
METHODS: A cross-sectional observational study was conducted using a convenience sample technique. The translation procedure included five stages: forward translation, revision of translation, backward translation, refinement of translation, and a final test of the pre-final version. The final sets of questionnaires were constructed using an online JotForm platform. The online platform was chosen to automatically calculate the questionnaire's final overall score. Overall, 260 participants were instructed to fill out the English and the Arab-OSDI version twice to conduct the reliability of the translated version and repeatability evaluation.
RESULTS: The mean age of the participants was 33.45 ± 11.74 years old. Cronbach's alpha for all items was greater than 0.80, except for the "blurred vision" and "deteriorating vision" items (0.77 and 0.74, respectively). The mean overall score difference between the English-OSDI and Arab-OSDI was 0.86 based on the Bland-Altman chart. For repeatability, no significant difference in the overall scores between the two repeats of the Arab-OSDI (p = 0.632). The Arab-OSDI overall score (sessions 1 and 2) has a clinical difference (bias) of 0.21. Using the varimax rotation method, only three factors (ocular symptoms, vision-related function, and environmental triggers) had eigenvalues greater than one in the structure of the Arab-OSDI.
CONCLUSION: The Arab-OSDI is an appropriate, reliable, and repeatable tool for the determination of dry eye symptoms, ocular discomfort, and quality of life in the Gazan population. This version could remove the language barrier in answering OSDI items more easily.
METHODS: A cross-sectional study was carried out between March and August 2022 in Gaza governorates using a proportional stratified sampling technique. Only Gazan individuals ≥ 18 years old and able to follow the instructions were included. The Ocular Surface Disease Index (OSDI) questionnaire, which has previously been translated into Arabic and validated, was applied to evaluate DED symptoms. Subjective clinical tests for DED conducted were tear meniscus height (TMH), meibomian gland dysfunctions (MGDs), Marx line (ML), conjunctival Lissamine green staining (LGS), tear film break-up time test (TBUT), corneal fluorescein staining (CFS) and Schirmer II tear test (STT). DED was defined based on an Arab-OSDI score ≥ 13 and at least one positive clinical sign.
RESULTS: A total of 426 participants were assessed from four areas (North Gaza Strip, 82; Gaza City, 147; Mid-Zone Gaza Strip, 62; South Gaza Strip, 135). The prevalence of DED in the present study was 31.5% (95% CI: 27.1, 36.1). Age > 50 years old (odds ratio [OR] = 10.45; 95% CI: 2.95, 37.05; P < 0.001), female gender (OR = 3.24; 95% CI: 1.40, 7.52, P = 0.006), menopause or pregnancy (OR = 2.59; 95% CI: 1.25, 5.35; P = 0.03) and pharmacotherapy (artificial tears; OR = 9.91; 95% CI: 2.77, 35.46; P < 0.001) were each associated with DED symptoms. South Gaza Strip (OR = 0.04; 95% CI: 0.01, 0.12; P < 0.001), unemployed (OR = 11.67; 95% CI: 1.43, 95.44; P = 0.02), non-consumption of caffeine (OR = 0.40; 95% CI: 0.19, 0.88; P = 0.02) and TMH < 0.2 (OR = 1.80; 95% CI: 1.02, 3.19; P = 0.04) were associated with TBUT < 5 s. LGS was associated with those > 50 years old (OR = 2.70; 95% CI: 1.38, 5.28; P = 0.004), previous refractive or ocular surface surgeries (OR = 2.97; 95% CI: 1.34, 6.59; P = 0.008) and CFS ≥ 1 (OR = 1.91; 95% CI: 1.07, 3.44; P = 0.03).
CONCLUSION: Various aspects of DED were linked with different risk factors, suggesting that DED subtypes have different underlying pathophysiologies.