2. Onchocerciasis patterns characterization using Earth Observation and machine learning

Onchocerciasis patterns characterization using Earth Observation and machine learning

GIMA
M-GEO
STAMP
Staff Involved
Dr. Rosa Aguilar
Dr. Ellen-Wien Agustijn
M-SE Core knowledge areas
Spatial Information Science (SIS)
Additional Remarks

Picture source: Joint Action Forum. Ougagdougou: WHO African Programme for Onchocerciasis Control; 1995

Topic description

Human onchocerciasis (river blindness) is part of the so-called Neglected Tropical Diseases (NTDs) that jointly affect more than one billion people living in impoverished communities. It is a chronic infection transmitted via the bites of black flies (Simulium vectors).  The WHO estimates more than 120 million people are exposed, and at least 1 million with severely impaired vision due to this disease. Most cases are in Africa (99%), Yemen, Brazil, and Venezuela. Climate change is worsening the situation of NTDs in general. Onchocerciasis remains a global health and wicked problem causing disability and stigma. The objective of this research is to predict the prevalence and intensity of human onchocerciasis utilizing geo-artificial intelligence methods applied to heterogeneous environmental and epidemiological data.

The guiding research questions are:

  1. To what extent can machine learning models predict spatial distribution (prevalence and intensity) of onchocerciasis in tropical areas?
  2. To what extent has the spatial pattern of onchocerciasis shifted concerning environmental factors?
Topic objectives and methodology
  • Methods suggested follow a traditional ML workflow:
    • data acquisition
    •  feature engineering,
    • modeling testing, and
    • evaluation.
  • Data:
    • Earth Observation data,
    • weather data, and
    • reference data: epidemiological data at community and individual levels, e.g., infection intensity and age, among others.
  • Challenges: tropical areas so cloud coverage is abundant. RADAR images might be an option to explore.

The developed model will be discussed with experts regarding its interpretation and utilization.

To effectively tackle this subject, a student would benefit from having strong skills in Python and familiarity with Jupyter notebooks​​​​​​.

References for further reading

Elimination of human onchocerciasis: progress report, 2021. https://www.who.int/publications/i/item/who-wer9746-591-598.

Botto, C., Basañez, MG., Escalona, M. et al. Evidence of suppression of onchocerciasis transmission in the Venezuelan Amazonian focus. Parasites Vectors 9, 40 (2016). https://doi.org/10.1186/s13071-016-1313-z