Feedback

[brynamo]

Feed Back For EKG Team

Real World

This is really an interesting problem that I feel has some amazing and under realized implications. Obviously heart health is a huge concern in the US but access to affordable health-care is a consistent problem especially in areas that are either poor or rural (worst case both). The goal of making EKG technology more ubiquitous will allow for proactive medicine which has an amazing impact of both quality of life but also cost savings (i.e. a win-win).

Considering the fact that this technology has the potential for wide impact the "target" group is quite wide, ranging from doctors interested in working with hard to reach populations to individuals interested in maintaining their health.

With this in mind, depending on your goal (i.e. developing a consumer product vs. a health-care product), I feel it would be good to consider what aspects you can tweak in order to achieve the maximum desired outcome.

An example being:

You are trying to make a consumer product. It would be good to focus on how to get the user "hooked" on the product to ensure that the data is as complete as possible, therefor having the best chance for preventative medicine. There is obviously a huge drop off in usage once the initial "OH COOL!" factor has worn off, so thinking about ways to keep the user engaged will be important (maybe gamification, maybe dataviz is really important, etc.)

Data

Frankly I think you have a better idea than I do what kind of related data is important so I'm not going to touch that. The biggest thing I would think about is security around the BLE you plan to implement. Obviously health-care data is highly sensitive so privacy and security really should not be an after thought.

Approach

I think it is great that you are building an iOS application, this helps with the security problem I was identifying (as opposed to sending it to the phone, then up to the cloud, around in the cloud, etc.). The one caution I would say is to take a look around and see what people have developed for dataviz in iOS. Building from scratch is great for learning if you have time but you can also waste a lot of time building something that already exists (or is close enough for you to take and then modify). Really it boils down to your intent, are you trying to learn everything about iOS development or are you trying to 'just get it done'. (And for the record I think both are totally fair)

Project Management

I think you have a realistic plan for the end of class, in terms of have a working prototype and it also helps that you want to work on this even after the class concludes. I think the working iOS application is a reach, but it would be awesome if you got it done!!

[davclark]

So this is a bit odd... as this was the team @ProximaDas was supposed to give feedback to. This trio seems to have been very problematic. Was the table hard to understand?

In any case, I'll contact @ProximaDas on Gitter.

[brynamo]

@davclark, the reason it happened this way was we were the only ones of the trio who were in class and so we met up to give each other feedback. @juanshishido met with @ProximaDas to make sure she was not left out in terms of feedback (as you saw in the other issue on save-seistool).

I can also meet with @ProximaDas if you would like to complete the circle.

[davclark]

Formally for the class, I'd just like @ProximaDas to chime in here. But I think I mentioned that the reason I wanted y'all to chat with each other was because there are lots of opportunities for you to share with each other!

So, ideally, she'll be in touch with @Jorjorbinx and @HanZen13. Thanks for chiming in, though!

[davclark]

Also the above feedback is still great (and hopefully appreciated by more than just me!)

[ProximaDas]

I did give feedback on EKG, all I did was upload a feedback file under the projects folder - as there was another team's feedback in there. My bad, I didn't realize I was supposed to open an issue!

[ProximaDas]

Real-world challenge

This project aims at making heart-rate monitoring products more affordable for the public. Another goal of the project is to make such devices portable, since the current devices are large and can mostly be housed in hospitals or other health-care institutions. EKG devices in hospitals allow patients to monitor their heart rate for 10-15 minutes at max. However, heart failures don't happen during this time period. Making heart rate monitors that can give continuous feedback to the user would help immensely in better monitoring of heart related ailments.
This product could be especially benefitical to heart patients, but could also be of use to other people who perform physically intensive activities.

Data / Materials

An ArduinoPro mini was used in conjunction with SparkFun's heart rate monitor to gather heart rate data of the team members. Currently, the data is stored into a computer using a USB drive that is connected to the Arduino.
The challenge in this method of data collection is that the recording is a series of electrovoltage changes. However, a heart beat is a much more complex phenomenon than just a time series of electrovoltage data. There are known to be atleast three waveforms constituting of a heart beat. Next steps in the data collection process is to figure out how can machines identify the intervals in heartbeats. Machine learning methods can be used here to train systems to calculate/glean the information of interest to the users.

Approach

Python is used as the primary tool for data analysis in the project. The Arduino used in the project was also programmed using the Arduino programming language.

Project Management

The first milestone of the project was to get the Arduino to start capturing heart rate information. The final goal is to be able to upload live heart rate information to a cellphone using Bluetooth technology. The data analysis of this data would be done in a smartphone in real-time.

Other

At this stage, figuring out what kind of information is expected to be gleaned out of heart rate data is important. Setting a simple and straightforward expectation from the data would help in getting the final deliverable up and running in less time, and would give more time for testing and debugging of the application.