Research group of Dr. Romero (RETINOPROGRAM)

RETINOPROGRAM is a 10-year collaboration between Dr Pere Romero-Aroca (Director of Ophtamology clinic at University Hospital San Joan de Reus) and Prof Domenech Puig (Head of ITAKA, URV) research groups, resulting in several publications and 4 registered softwares. Our access to clinical expertise and patients meant that our tools were awarded a fast track to innovation program (Innobics) to be adopted in the Catalonia ICS hospitals (accessible to all doctors through EHR database).

Under the IISPV umbrella, both groups have joined efforts work in the diagnosis and prevention of Diabetic Retinopathy (DR). Despite being the fifth cause of blindness and visual impairment (and most common one in working age adults), there is no stratification method based on their risk to develop DR. Furthermore, existing treatments can only stop progression and DR is asymptomatic in its early stages. Prevention plays also a key role in DR development as it can be prevented by controlling some risk factors as glycemia or arterial hypertension.

Researchers from both groups are working on a tool that 1) stratify patients based on their risk of developing DR, 2) calculates their next screening appointment based on the calculated risk, and 3) offers prevention interventions for patients to self-manage their risk. This tool would be the merging of ourexisting and validated softwares (through EyePACS and general population data):

  • A diagnostic aid system through Clinical Decision Support System (CDSS) that allows for screening of patients based on their risk using 8 clinically validated variables.
  • A DR Deep Learning Algorithm (DLA) that stratifies patients in 4 levels of severity and sets up their next appointment based on the resulting risk level.

What distinguishes their work from existing companies in the market (IDx and EyeArt) is our access to patients and clinical expertise. Thus, existing companies work with DLA but they don’t take into account the associated clinical risk factors for DR and consequently, cannot provide personalised prevention and treatment plans. In addition, our product has been validated by EyePACS, MESSIDOR-2 and retinographies from the general population (107.977 patients for CDSS and 18.792 patients for DLA), whilst our competitors have reached the market by only validating their models with EyePACS and/or MESSIDOR.