| KEYWORD |
Automated Analysis of Retinal Images for early Diagnosis of Diabetic Retinopathy
keywords BIOENGINEERING, E-HEALTH, IMAGE PROCESSING
Reference persons PAOLO ERNESTO PRINETTO
External reference persons Prof. Massimo PORTA (Centro Retinopatia Diabetica, Università di Torino – Ospedale San Giovanni Antica Sede)
Marco INDACO (PhD, Turin node of the CINI National Lab on Assistive Technologies)
Giuseppe AIRO’ FARULLA (PhD candidate, Politecnico di Torino)
Research Groups TESTGROUP - TESTGROUP
Thesis type EXPERIMENTAL
Description Diabetic retinopathy is the leading cause of new blindness in persons aged 25-74 years in the industrialized world.
In the initial to fairly advanced stages of diabetic retinopathy, patients are generally asymptomatic; in the more advanced stages of the disease, however, they may experience symptoms that include floaters, blurred vision, distortion, and sudden or progressive loss of visual acuity. Retinal signs of early diabetic retinopathy include microaneurysms: the earliest clinical sign of diabetic retinopathy, these occur secondary to capillary wall outpouching consequent to pericyte loss. They appear as small, red dots in the superficial retinal layers. Subsequent lesions include larger blot and flame-shaped haemorrhages, “hard exudates” (bright yellow deposits of cholesterol) and “cotton wool spots” (areas of micro-infarcts) At present, such lesions are detected by trained personnel that evaluates retinal images: such screening is effective but time-consuming. Although no software can replace a comprehensive eye examination in terms of overall ocular evaluation, it is clear that the development of a reliable tool would increase the rate of retinal assessment in various diabetic populations, increase access to care in remote or underserved areas, and identify clinically important diseases while reducing the immediate need for retinal evaluation of those with minimal or no retinopathy.
The thesis aims at developing an Image Processing environment to automatically screen retinal images in order to evaluate the presence (or absence) of pre-symptomatic lesions with the ultimate purpose of identifying those patients in most need of prompt ophthalmic assessment and treatment.
See also http://emedicine.medscape.com/article/1225122-overview
Required skills OpenCV or Matlab
Notes The candidate will acquire a deep knowledge of Image Processing, with particular emphasis on Object Detection techniques.
The research activities and the thesis are carried out in cooperation with:
- Department of Medical Sciences, University of Turin
- Centro Retinopatia Diabetica – Ospedale San Giovanni Antica Sede (Turin, Italy)
- Turin node of the CINI National Lab on Assistive Technologies
Deadline 02/04/2016
PROPONI LA TUA CANDIDATURA