Diabetic retinopathy

[dehall]

This PR adds modules for Diabetic Retinopathy progression and treatment. These were originally developed for a synthetic dataset for the National Eye Institute on the diabetic_retinopathy branch. Because some "experimental" features were used, we don't necessarily want to merge that to master as-is.

For a full description of the modules, see the writeup at https://github.com/synthetichealth/synthea/blob/diabetic_retinopathy/docs/diabetic_retinopathy_module_guide.md
Also, some synthetic patients are available at:

1. https://synthetichealth.github.io/spt/#/?patient=github/synthetichealth/synthea/diabetic_retinopathy/samples/Eve68_Bahringer146.json
2. https://synthetichealth.github.io/spt/#/?patient=github/synthetichealth/synthea/diabetic_retinopathy/samples/Ken316_Hermann103.json
3. https://synthetichealth.github.io/spt/#/?patient=github/synthetichealth/synthea/diabetic_retinopathy/samples/Maryjane289_White193.json
4. https://synthetichealth.github.io/spt/#/?patient=github/synthetichealth/synthea/diabetic_retinopathy/samples/Eve68_Pfannerstill264.json
5. https://synthetichealth.github.io/spt/#/?patient=github/synthetichealth/synthea/diabetic_retinopathy/samples/Janet238_Marquardt819.json

In addition to the new modules, this adds the following minor tweaks:

• Adds a "note" field to HealthRecord.Entry which will be included in the generated notes in a couple places. Nothing currently sets this field (it was used in the original branch by a Java submodule that generated unstructured text) but I think it could be useful in the future.
• Fix for a bug that creates duplicate ImagingStudy instances

This PR should be squashed prior to merging so that the files that were added but then removed don't take up space

[dehall]

Quick test of the prevalence of the new conditions before merging:
My notes say this should be aiming for 28.5% prevalence of diabetic retinopathy among adults age 40+ with diabetes, and 4.4% prevalence of vision threatening diabetes, which looks like it's based on this survey https://pmc.ncbi.nlm.nih.gov/articles/PMC2945293/

I ran a population of 10,000 and got the following counts:
Diabetes - 630
Nonproliferative Diabetic Retinopathy - 174, 174/630 = 27.6%
Proliferative Diabetic Retinopathy - 12, 12/630 = 2%

So PDR is a little low but overall this is about where I want it, so I'm going to go ahead and merge