1. Dengue Prediction
Dengue fever is an infectious disease caused by Flavivirus transmitted by the Aedes mosquito. This disease predominantly occurs in the tropical and subtropical regions. With no specific treatment, the most effective way to prevent dengue is vector control. The dependence of the Aedes mosquito population on meteorological variables make the prediction of dengue infection possible using conventional statistical and epidemiologic models. However, with increasing average global temperature, the predictability of these models may be lessened employing the need for artificial neural network. We created an artificial neural network to predict dengue incidence with humidity, rainfall, and temperature as input variables. You can read the paper in this link https://doi.org/10.25929/e3bk-a370
2. Melanoma Prediction
Melanoma is a type of skin cancer that affects many people worldwide. Although it only consists 1% of the total skin cancer incidence, 80% of deaths due to skin cancer is related to melanoma. Dermatologists are healthcare specialist that has a big role in the diagnosis of melanoma. However, these experts are not sufficient enough to cover a wide population. To augment this problem, MelanomaENV2, is created that can predict melanoma based on dermoscopy images.
*This is the master thesis of Bryan Zafra, the founder of Kinase AI. If you wish to have copy of this paper please send him an email corporate@kinase-ai.com
This model is available in GitHub via this link: https://github.com/baypogiMD/MelanomaENV2
3. Drug Discovery or Re-purposing
Some of the common diseases have no direct treatment modalities; most uncommon and rare diseases have no treatment at all. Using AI, we can hasten the drug discovery to address this unmet need in therapeutic aspect of these diseases. The existing drug molecule database can be use to create an machine learning pipeline for new drug discovery. With lists of FDA-approved drugs, another pipeline can be used to re-purpose certain drugs to be use to treat other diseases.
Are you an INVESTOR?
Send us an email corporate@kinase-ai.com