how can i start coding for that.

Well, like you would start any other (smaller) project - break the functionality into small tasks that can be handled separately (hardware initialization, adc data acquisition, signal processing, communication, etc) and implement and test them. Then put it all together.

However, you have probably picked a rather complex task there. The "biomedical signal monitor" by itself is far from trivial (and if it's not a study/hobby project, you'll also have to deal with a truckload of regulatory issues), neither on the hardware side (signal quality, safety (!!!!!!) if you're going to hook this thing up to anything that's more alive than an ECG simulator, EMI issues, defibrillator safety, etc) nor on the software side. Coupling that with "telemetry", which also is far from trivial, you get a fairly complex project.

If you need to ask the question "how can i start coding for that.", you might be better of starting with a smaller project, and tackle the telemetry ECG once you've had opportunity to experiment with the smaller project.

I also doubt that a '51 is the right processor for anything but a very, very simple (3 lead) ECG. ECG requires significant CPU horsepower for signal processing.

Also, do you already know what vital parameter you are going to measure (ECG, (non-)invasive blood pressure, generic pressure, temperature, blood oxygen saturation, etc ) ?

Maybe the first issue to tackle would be deciding on one. There's complex measurements (SpO2, ECG, non-invasive blood pressure), and then there's those that can be easily handled by a '51 (pressure, temperature, simple plethysmography to derive a heart rate).