You may try to exploit so called autofluorescence (AF) of endogenous fluorophores. Many endogenous molecules, such as proteins, DNA, collagen etc. display their spontaneous emission under proper excitation. You may try to correlate this with biological processes inside : see http://www.ncbi.nlm.nih.gov/pubmed/21959354. You'll not get number of stationary macrophages, limphocytes etc. from these measurements, but the AF may be possibly applied for liver metabolism assessment.

I agree with Artur. If you're interested in metabolism, quantification of NADH, FADH would be a top priority. You may be interested in this work as well:

http://www.ncbi.nlm.nih.gov/pubmed/15148320

I would like to give you some suggestions in terms of MRI technology. I suggest you use the birdcage coil for transmission and surface arrays for reception. Please pay attention to the base line calibration and water suppression in the MR Spectroscopy.

I have skimmed read it and from what I understand this is method that can be easily used in vitro but not in patients?! I will explain what my idea is and we can take it from here.

I have a patient with liver cancer, primary liver cancer, no metastasis as of now. I want to do metabolic map of liver, oxygenation map, vascularisation map, local immunology map. Ideally it would be each hepatocyte described with above data, if not hepatocyte then some bigger area, the better resolution the better. I wait for metastases. Patient either dies of metastatic diseases or bigger part of the liver is resected along with 'meta'. I design algorithm based on gather data, algorithm helps me predict places of future metastatic niches. I check specificity and sensitivity of the algorithm by doing histopathological analysis of metastatic niches and predicted by algorithm future metastatic niches but this is another part of the project. Lets focus on metabolism imaging but in living human.

It's difficult to implement fluorescence techniques in vivo in humans. The best you could do would be to utilize a somewhat invasive approach, which wouldn't help you much. In this case, you may be better served with a high-end MR spectroscopy approach, although access to this type of instrumentation may be limited (and I'm not sure if it's FDA approved or not).

Your goal of multi parametric studies of liver diseases is current under investigation by a number of teams. An good model to follow with magnetic resonance spectroscopy and imaging has been taken by collaboration at Oxford. I recommend that you look at their study as a roadmap for the use of MRI.

Multiparametric magnetic resonance for the non-invasive diagnosis of liver disease. J Hepatol. 2014 Jan;60(1):69-77. doi: 10.1016/j.jhep.2013.09.002 The paper is available as a free download from PMC at the US National Library of Medicine.

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3865797/

I agree with Victor. The MRS technique, would not allow you to obtain a result on the cellular level, but only on the tissue level. Still, it might give you reliable result on the changes in tissue metabolism. MRS is often use to investigate the tumor progression level in the brain.