Collect Death Causal Graphs to Prevent Death
Collecting medical tracebacks from hospitals around the world and computing risk of deeper causes to recommend early diagnostics and lifestyle changes
To prevent deaths and preserve youth, we have to understand their causes, and the idea is that we could do this very efficiently with more details from hospitals, which perform autopsies and determining root causes, and then, trying to reduce the number of events that lead to occurrence of these causes.
In other words, imagine collecting causal trees:
condition A caused by B caused by C
While many doctors in the diagnosis of , maybe the conditional probabilities of P(B|C) and P(A|B) are high, and that means, we have to prevent C.
The final causes (such as A) are generally publicly available, for example, the U.S. Centers for Disease Control and Prevention ("CDC") provides the datasets of all anonymized deaths from 1968 to 2016, with the following kind of columns are available.
The idea could be implemented by collecting comparable datasets from all countries, parsing them into deep causal trees, computing probabilities of causes, a model for early diagnostics recommender (advertising revenue-based diagnostics recommender could pay for itself), as well as life styles changes recommender would be an example of the realization of this idea.