I think ultrasound is an excellent way of characterising Mullarian duct variations.
You do need to do it with a transvaginal scan.
You do need the technology to reconstruct a C-plane image of the uterus. This can be done with a manual sweep or even better if you have a proper 3D transvaginal transducer with a mechanical transducer in the end. Many manufacturers have this technology and it provides you with a calibrated 3D volume to work with.
There are well established criteria for characterising uterine shapes and personally I don’t find it difficult.
To give you the exact details of how to characterise uterine cavity shapes would require a long essay so I will email you a ew journal articles and I will record a webinar on the topic and add it to the website. I have a nice lecture on the topic already written so I will record it this weekend.