Deep Leaning-Based Adjunctive Therapy Tool for Mitochondrial Dysfunction

Handong Li1,Email

Xiangyu Zhang2

Ishita Gulati2

​​​​​​​Satnam Dlay2

1School of Mathematics, Physics and Electrical Engineering, Northumbria University, Newcastle Upon Tyne, NE1 8ST, UK.
2School of Electrical and Electronic Engineering, Newcastle University, Newcastle Upon Tyne, NE1 7RU, UK.

Abstract

Mitochondrial myopathy is a maternally inherited metabolic disease, the root cause of which is mitochondrial structure and dysfunction due to genetic mutations in mitochondrial DNA (mtDNA) or nuclear DNA (nDNA). Due to complex classification and overlapping symptoms of mitochondrial myopathy, this paper introduces a non-invasive diagnostic tool to help examine patients with mitochondrial dysfunction. This paper first introduces the facial image processing of patients with mitochondrial dysfunction. The Viola-Jones algorithm and the improved Canny Algorithm are used to detect and extract the features of the eyebrows, and then the curvature of the eyebrows is calculated mathematically to describe the facial changes of the patients. The results of our experiments will aid in the patient's treatment record and may help patients detect symptoms earlier. Researchers can remotely and long-term improve the odds of treatment by taking pictures of patients. In addition, experimental data will help to mathematically document treatment improvements.

Deep Leaning-Based Adjunctive Therapy Tool for Mitochondrial Dysfunction