Profiling Bell’s Palsy based on House-Brackmann Score

Abstract

In this study, we propose to diagnose facial nerve palsy using Support Vector Machines (SVMs) and Emergent Self-Organizing Map (ESOM). This research seeks to analyze facial palsy domain using facial features and grade the degree of nerve damage based on the House-Brackmann score. Traditional diagnostic approaches involve a medical doctor recording a thorough history of a patient and determining the onset of paralysis, rate of progression and so on. The most important step is to assess the degree of voluntary movement of the facial nerves and document the grade of facial paralysis using House- Brackmann score. The significance of the work is the attempt to understand the diagnosis and grading processes using semi-supervised learning with the aim of automating the process. The value of the research is in identifying and documenting the limited literature seen in this area. The use of automated diagnosis and grading greatly reduces the duration of medical examination and increases the consistency, because many palsy images are stored to provide benchmark references for comparative purposes. The proposed automated diagnosis and grading are computationally efficient. This automated process makes it ideal for remote diagnosis and examination of facial palsy. The profiling of a large number of facial images are captured using mobile phones and digital cameras.

Automatic Grading of Palsy Using Asymmetrical Facial Features: A Study Complemented by New Solutions

“…Precision and recall metrics represent the cases that are predicted correctly over all positive predictions and observations, respectively. The precision value of our method is 92.91% compared to 89.11% and 89.23% for [1,34] respectively. This indicates that the prediction accuracy of our classification model is better than the existing models.…”

“…Similarly, the sensitivity of 93.14% implies better performance of the proposed method compared to the existing methods in classifying the facial palsy into one of the 5 grades. Finally, a F1 score of 93.00% is achieved compared to 89.66% and 90.00% for existing methods [1,34], respectively.…”

“…The precision, sensitivity and F1 score metrics for our proposed method are displayed in Table 3, where the performance comparison of VGG-16 net with [1,34] is presented. Precision and recall metrics represent the cases that are predicted correctly over all positive predictions and observations, respectively.…”

Deep Learning for the Assessment of Facial Nerve Palsy: Opportunities and Challenges

“…For example, the House-Brackmann scale only provides physicians a grade of 1 to 6 (from normal to total paralysis) to compare the forehead, eye, and mouth regions between the paralyzed and nonparalyzed regions of a patient's face. 7 Meanwhile, the eFACE evaluates 16 items through a visual analogue scale evaluating synkinesis, dynamic, and static domains, which allows for a more nuanced and specific assessment of palsy severity. Patients are asked to perform certain movements such as eyebrow elevation, eye closure, closed and open mouth smile, and lower lip depression.…”

Study on clinical characteristics and evaluating treatment results of bell’s palsy patients at Can Tho Central General Hospital and Can Tho University of Medicine and Pharmacy Hospital in 2020 – 2022

“…* Các đặc điểm lâm sàng: bất thường nếp nhăn trán; bất thường rãnh mũi má; dấu hiệu Charles Bell; một số triệu chứng khác; phân loại mức độ tổn thương khi vào viện theo thang điểm House -Brackmann [8].…”