Dear Puja According to the information you provided, there are two possibilities:

1 - The images in a different planes refer to a specific sample liver, then you can use the 3D segmentation instead of the 2D segmentation,specially when 3D volume measurement is required . Of course, this requires image fusion. The whole process can be done both manually and (semi-)automated. thesis [1] has a suitable review about them. if your goal is to segment the particular part of the liver, then the algorithm should be directed in that context.

There are several software that performs it semi-automatically, like 3D-Doctor and ...

2. The images in different planes refer to multiple samples , in which case two-dimensional segmentation methods can work(manual/automated). texture-based Approaches were reviewed in[2].

In this case, the auxiliary softwares can do this process (semi-)automatically.

but, nowadays manually and semi-automated framework are more reliable in clinical used( [1] ).

[1] Childs, Jessie. "Development of a simple technique of measuring liver volume using 2D ultrasound including development of a reference range." (2016).

http://search.ror.unisa.edu.au/record/UNISA_ALMA11144425100001831/media/digital/open/9916128907301831/12144425090001831/13144425080001831/pdf

[2] Rathore, Saima, et al. "Texture analysis for liver segmentation and classification: a survey." Frontiers of Information Technology (FIT), 2011. IEEE, 2011.

I hope it can be useful.

Regards