Advanced search
Start date

Equipment to diagnosis the corneal neovascularization through its image captured by mobile phone

Grant number: 18/08344-4
Support type:Research Grants - Innovative Research in Small Business - PIPE
Duration: May 01, 2019 - October 31, 2020
Field of knowledge:Health Sciences - Medicine - Medical Clinics
Principal Investigator:Tais Hitomi Wakamatsu
Grantee:Tais Hitomi Wakamatsu
Company:Vision Serviços Médicos Ltda - ME
CNAE: Atividades de serviços de complementação diagnóstica e terapêutica
Atividades de apoio à gestão de saúde
Atividades de atenção à saúde humana não especificadas anteriormente
City: São Paulo
Assoc. researchers:José Alvaro Pereira Gomes ; Ugo Ibusuki
Associated scholarship(s):19/17932-0 - Equipment for cornea neovascularization diagnosis utilizing images from mobile phone, BP.TT


This project considers several challenges in the study and monitoring of the treatment of corneal neovascularization (NV). The cornea is by nature avascular, and thus kept constantly in healthy condition so that it presents its total transparency. The cornea consists of an outer layer of non-keratinized stratified squamous epithelium, a collagen stroma with keratocytes and, surrounding the anterior chamber, a cuboid endothelium (Rodrigues, 2015) with approximately 1mm of thickness at the periphery and 0.5mm at the center. The cornea, in addition to acting as a protective barrier for intraocular tissues, functions essentially as a transmitting element and refraction of incident light as a camera lens. For efficient light transmission it is essential to maintain the transparency of the cornea, a process in which avascularity is a key factor. In abnormal conditions of inflamed ocular surfaces, the cornea presents a strong tendency to vascularization, which predisposes to its opacity and consequent low vision. Corneal neovascularization becomes a priority study because it consists of a disorder that affects 1% of the general population and can cause blindness in almost 12% of patients with the symptom (Dastjerdi et al., 2009). The use of contact lenses, whose popularity has been increasing in recent decades, has become one of the main causes of NV. Contact lenses stimulate neovascularization, either causing mechanical irritation of the limbic sulcus and epithelial erosion, or through the induction of hypoxia and hypercapnia, which leads to limbic inflammation and release of mediators of angiogenesis leading to NV. Corneal neovascularization can lead to an irreversible loss of visual acuity, which implies an early and aggressive therapeutic approach through topical eye drops that should be monitored by medical specialists periodically (Wakamatsu, 2014). For treatment options to be adequately evaluated, corneal images should be taken at periodic intervals and the vascularization monitored longitudinally. However, several problems enter into the analysis, since the images may be of variable quality. Factors that may lead to differences or deterioration in the images are: (i) differences in illumination, between examinations; (ii) inter- or intra-operator variability in image taking; (iii) lack of sharpness in corneal images; (iv) lack of knowledge about the frequency-absorbing characteristics of eye tissues. The objective of this project is to develop an imaging equipment for the neovascularized area of the cornea through the development of a more economically viable national technology to enable images with adequate clarity for a better clinical diagnosis through the automatic extraction and calculation of the vascular area. The proposal is to develop a handheld device coupled with smartphones, making it easier for poorer and underserved populations to access eye exams, as well as queuing in public hospitals in large centers, and reducing the cost of purchasing sophisticated equipment for this purpose. The image extracted by this device will be processed by an artificial intelligence software that will calculate the neovascularized area of the cornea, compare with exams previously performed by the same patient and generate a diagnosis in an automated way. (AU)