diff --git a/amypet/__init__.py b/amypet/__init__.py
index ab91b3ee..ec9f7ba9 100644
--- a/amypet/__init__.py
+++ b/amypet/__init__.py
@@ -21,6 +21,6 @@
     except FileNotFoundError:
         __licence__ = "MPL-2.0"
 
-from .aux import *
 from .preproc import *
 from .suvr_tools import *
+from .utils import *
diff --git a/amypet/aux.py b/amypet/aux.py
deleted file mode 100644
index 6adb4ce8..00000000
--- a/amypet/aux.py
+++ /dev/null
@@ -1,559 +0,0 @@
-"""Auxiliary functions for the centiloid project"""
-__author__ = "Pawel Markiewicz"
-__copyright__ = "Copyright 2022"
-import pickle
-import re
-import urllib
-from pathlib import Path
-
-import matplotlib.pyplot as plt
-import numpy as np
-from niftypet import nimpa
-from pkg_resources import resource_filename
-from scipy.stats import linregress
-
-# > regions used in CL project
-rvois = ['wc', 'cg', 'wcb', 'pns']
-
-# > folder name with the default conversion tables after calibration
-cl_anchor_fldr = Path(resource_filename(__name__, 'data/cl_anchor_tables'))
-cl_masks_fldr = Path(resource_filename(__name__, 'data/CL_masks'))
-amypet_dir = Path.home() / '.amypet'
-
-# > region full name strings for plots
-rvoi_str = {
-    'wc': 'WHOLE CEREBELLUM', 'cg': 'CEREBELLUM GM', 'wcb': 'WHOLE CEREBELLUM + BRAIN STEM',
-    'pns': 'PONS'}
-
-
-# ----------------------------------------------------------------------
-def get_atlas(atlas='aal', res=1, outpath=amypet_dir):
-    '''Get a brain atlas from `neuroparc` out of many available in MNI space.
-
-       Options:
-       - atlas:     one of many atlases, e.g., 'aal' which is the default;
-                    the number of atlases available can be extended by entering it
-                    in the dictionary. TODO: move the dictionary to `defs.py`.
-       - res:       the resolution of the atlas in mm.
-    '''
-    atlases = {
-        'ghpath': 'https://github.com/neurodata/neuroparc/raw/master/atlases/label/Human/',
-        'aal': 'AAL_space-MNI152NLin6_res-', 'aal_lbl': 'Anatomical-labels-csv/AAL.csv',
-        'schaefer200': 'Schaefer200_space-MNI152NLin6_res-',
-        'talairach': 'Talairach_space-MNI152NLin6_res-'}
-
-    # > get the GitHub data object
-    ghd = urllib.request.urlopen(atlases['ghpath'] + atlases[atlas] + f'{res}x{res}x{res}.nii.gz')
-    data = ghd.read()
-
-    # > save to file
-    nimpa.create_dir(outpath)
-
-    fatl = outpath / f'atlas-{atlas}_res-{res}mm.nii.gz'
-    with open(fatl, 'wb') as f:
-        f.write(data)
-
-    return fatl
-
-
-# ----------------------------------------------------------------------
-
-
-# ----------------------------------------------------------------------
-def im_check_pairs(fpets, fmris):
-    '''
-    checks visually image by image if the PET and MR have
-    the same orientation.
-    '''
-    fig, ax = plt.subplots(2, 2, figsize=(10, 10))
-    for fp, fm in zip(fpets, fmris):
-        print(f'{fp} : {fm}')
-        p = nimpa.getnii(fp)
-        m = nimpa.getnii(fm)
-
-        ps = p.shape
-        ms = m.shape
-
-        ax[0, 0].imshow(p[ps[0] // 2, ...], cmap='magma')
-        ax[1, 0].imshow(m[ms[0] // 2, ...], cmap='bone')
-
-        ax[0, 1].imshow(p[..., ps[2] // 2], cmap='magma')
-        ax[1, 1].imshow(m[..., ms[2] // 2], cmap='bone')
-
-        plt.draw()
-        plt.waitforbuttonpress(0)
-
-
-# ----------------------------------------------------------------------
-
-
-# ----------------------------------------------------------------------
-def identity_line(ax=None, ls='--', *args, **kwargs):
-    '''
-    plot identity line for any correlation analysis
-    needed for the CL calibration and beyond.
-    '''
-    # see: https://stackoverflow.com/q/22104256/3986320
-    ax = ax or plt.gca()
-    identity, = ax.plot([], [], ls=ls, *args, **kwargs)
-
-    def callback(axes):
-        low_x, high_x = ax.get_xlim()
-        low_y, high_y = ax.get_ylim()
-        low = min(low_x, low_y)
-        high = max(high_x, high_y)
-        identity.set_data([low, high], [low, high])
-
-    callback(ax)
-    ax.callbacks.connect('xlim_changed', callback)
-    ax.callbacks.connect('ylim_changed', callback)
-    return ax
-
-
-# ----------------------------------------------------------------------
-
-
-# ----------------------------------------------------------------------
-def get_clref(fxls):
-    '''
-    get the excel sheets with CL reference data from the CL project
-    excel sheet.
-    '''
-    import openpyxl
-
-    # > Define variable to load the workbook
-    wb = openpyxl.load_workbook(fxls)
-    ws = wb.active
-
-    pib_tbl = {'yc': {}, 'ad': {}}
-
-    # > cell offsets
-    ioff_ad = [5, 50]
-    ioff_yc = [51, 85]
-
-    sbj = [str(i.value) for i in ws['A']]
-    cg = [str(i.value) for i in ws['B']]
-    wc = [str(i.value) for i in ws['C']]
-    wcb = [str(i.value) for i in ws['D']]
-    pns = [str(i.value) for i in ws['E']]
-
-    ccg = [str(i.value) for i in ws['F']]
-    cwc = [str(i.value) for i in ws['G']]
-    cwcb = [str(i.value) for i in ws['H']]
-    cpns = [str(i.value) for i in ws['I']]
-
-    # > young control indices and SUVr/CL values
-    pib_tbl['yc']['id'] = np.array([int(i[3:]) for i in sbj[ioff_yc[0]:ioff_yc[1]]])
-
-    pib_tbl['yc']['suvr'] = {}
-    pib_tbl['yc']['suvr']['cg'] = np.array([float(i) for i in cg[ioff_yc[0]:ioff_yc[1]]])
-    pib_tbl['yc']['suvr']['wc'] = np.array([float(i) for i in wc[ioff_yc[0]:ioff_yc[1]]])
-    pib_tbl['yc']['suvr']['wcb'] = np.array([float(i) for i in wcb[ioff_yc[0]:ioff_yc[1]]])
-    pib_tbl['yc']['suvr']['pns'] = np.array([float(i) for i in pns[ioff_yc[0]:ioff_yc[1]]])
-
-    pib_tbl['yc']['cl'] = {}
-    pib_tbl['yc']['cl']['cg'] = np.array([float(i) for i in ccg[ioff_yc[0]:ioff_yc[1]]])
-    pib_tbl['yc']['cl']['wc'] = np.array([float(i) for i in cwc[ioff_yc[0]:ioff_yc[1]]])
-    pib_tbl['yc']['cl']['wcb'] = np.array([float(i) for i in cwcb[ioff_yc[0]:ioff_yc[1]]])
-    pib_tbl['yc']['cl']['pns'] = np.array([float(i) for i in cpns[ioff_yc[0]:ioff_yc[1]]])
-
-    # > AD indices and SUVr/CL
-    pib_tbl['ad']['id'] = np.array([int(i[3:]) for i in sbj[ioff_ad[0]:ioff_ad[1]]])
-    pib_tbl['ad']['suvr'] = {}
-    pib_tbl['ad']['suvr']['cg'] = np.array([float(i) for i in cg[ioff_ad[0]:ioff_ad[1]]])
-    pib_tbl['ad']['suvr']['wc'] = np.array([float(i) for i in wc[ioff_ad[0]:ioff_ad[1]]])
-    pib_tbl['ad']['suvr']['wcb'] = np.array([float(i) for i in wcb[ioff_ad[0]:ioff_ad[1]]])
-    pib_tbl['ad']['suvr']['pns'] = np.array([float(i) for i in pns[ioff_ad[0]:ioff_ad[1]]])
-
-    pib_tbl['ad']['cl'] = {}
-    pib_tbl['ad']['cl']['cg'] = np.array([float(i) for i in ccg[ioff_ad[0]:ioff_ad[1]]])
-    pib_tbl['ad']['cl']['wc'] = np.array([float(i) for i in cwc[ioff_ad[0]:ioff_ad[1]]])
-    pib_tbl['ad']['cl']['wcb'] = np.array([float(i) for i in cwcb[ioff_ad[0]:ioff_ad[1]]])
-    pib_tbl['ad']['cl']['pns'] = np.array([float(i) for i in cpns[ioff_ad[0]:ioff_ad[1]]])
-
-    return pib_tbl
-    # ----------------------------------------------------------------------
-
-
-def check_suvrs(suvr_yc, suvr_ad, refs):
-    ''' Check/QC the obtained SUVRs in `suvr_yc` and `suvr_ad `dictionaries
-        in relation to the reference in `refs` dictionary for the groups
-        ('ad' or 'yc').
-
-        Returns dictionary with the mean SUVrs and differences.
-    '''
-
-    # > prepare diff dictionary for the differences observed
-    # > in young controls (yc) and AD patients (ad)
-    diff = {'yc': {}, 'ad': {}}
-
-    # > initialise the means for each reference VOI
-    for rvoi in rvois:
-        diff['yc'][rvoi] = {'mean_ref': 0, 'mean': 0, 'N': 0}
-        diff['ad'][rvoi] = {'mean_ref': 0, 'mean': 0, 'N': 0}
-
-    def run_checks(suvr_dct, grp):
-
-        for rvoi in rvois:
-            print('========================================================')
-            for k in suvr_dct:
-
-                if grp == 'yc':
-                    idx = int(k[2:5])
-                elif grp == 'ad':
-                    idx = int(k[2:4])
-                else:
-                    raise ValueError('e> unknown group - only <yc>  or <ad> are accepted')
-
-                i = np.where(refs[grp]['id'] == idx)[0][0]
-
-                suvr = suvr_dct[k]['suvr'][rvoi]
-                suvr_ref = refs[grp]['suvr'][rvoi][i]
-                err = 100 * (suvr-suvr_ref) / suvr_ref
-
-                diff[grp][rvoi]['N'] += 1
-                diff[grp][rvoi]['mean_ref'] += suvr_ref
-                diff[grp][rvoi]['mean'] += suvr
-                diff[grp][rvoi][k] = {'suvr': suvr, 'ref': suvr_ref, 'err': err}
-
-                print(f'refvoi={rvoi}, indx> {idx}, suvr={suvr:.3f},'
-                      f' ref={suvr_ref:.3f}, error={err:.3f}%')
-
-            diff[grp][rvoi]['mean'] /= diff[grp][rvoi]['N']
-            diff[grp][rvoi]['mean_ref'] /= diff[grp][rvoi]['N']
-
-            # relative % mean difference
-            emean = diff[grp][rvoi]['mean']
-            rmean = diff[grp][rvoi]['mean_ref']
-            rmd = (emean-rmean) / rmean
-            diff[grp][rvoi]['mean_diff'] = rmd
-            print('--------------------------------------------------------')
-            print(f'> group % mean difference: {rmd:.3f}% (suvr={emean:.3f}, ref={rmean:.3f})')
-
-    print('========================================================')
-    print('YOUNG CONTROLS (YC)')
-    print('========================================================')
-    run_checks(suvr_yc, 'yc')
-
-    print('========================================================')
-    print('AD PATIENTS (AD)')
-    print('========================================================')
-    run_checks(suvr_ad, 'ad')
-
-    return diff
-
-
-# ----------------------------------------------------------------------
-
-
-# ----------------------------------------------------------------------
-def check_cls(suvr_yc, suvr_ad, diff, refs):
-    ''' check centiloid values relative to the reference in `refs`
-        dictionary, using SUVr dictionaries for YC and AD groups
-        `suvr_yc` and `suvr_ad`.  Also, update the `diff` dictionary
-        with the centiloid results and output it.
-    '''
-
-    for rvoi in rvois:
-        print('=====================================================')
-        suvr_0 = diff['yc'][rvoi]['mean']
-        suvr_100 = diff['ad'][rvoi]['mean']
-
-        grp = 'yc'
-        suvr_dct = suvr_yc
-
-        for k in suvr_dct:
-
-            idx = int(k[2:5])
-            i = np.where(refs[grp]['id'] == idx)[0][0]
-
-            cl = 100 * (suvr_dct[k]['suvr'][rvoi] - suvr_0) / (suvr_100-suvr_0)
-            cl_ref = refs[grp]['cl'][rvoi][i]
-
-            diff[grp][rvoi][k].update(cl=cl, cl_ref=cl_ref)
-
-            print(f'refvoi={rvoi}, indx> {idx}, cl={cl:.3f}, ref={cl_ref:.3f}')
-
-        grp = 'ad'
-        suvr_dct = suvr_ad
-
-        for k in suvr_dct:
-
-            idx = int(k[2:4])
-            i = np.where(refs[grp]['id'] == idx)[0][0]
-
-            cl = 100 * (suvr_dct[k]['suvr'][rvoi] - suvr_0) / (suvr_100-suvr_0)
-            cl_ref = refs[grp]['cl'][rvoi][i]
-
-            diff[grp][rvoi][k].update(cl=cl, cl_ref=cl_ref)
-
-            print(f'refvoi={rvoi}, indx> {idx}, cl={cl:.3f}, ref={cl_ref:.3f}')
-
-    def get_cls(diff, grp, rvoi, cl='cl'):
-        ''' Get centiloid values in a list for any group and reference VOI.
-        '''
-        return [diff[grp][rvoi][k][cl] for k in diff[grp][rvoi] if grp in k.lower()]
-
-    # ------------------------------------------------------------------
-    fig, ax = plt.subplots(2, 2, figsize=(10, 10))
-
-    for a, rvoi in enumerate(rvois):
-        i = a // 2
-        j = a % 2
-
-        # > combine CLs for both groups
-        cl = get_cls(diff, 'yc', rvoi, cl='cl')
-        cl += get_cls(diff, 'ad', rvoi, cl='cl')
-
-        # > combine reference CLs into one list
-        clref = get_cls(diff, 'yc', rvoi, cl='cl_ref')
-        clref += get_cls(diff, 'ad', rvoi, cl='cl_ref')
-
-        # > find the linear regression between the CL and reference CL
-        m, a, r, p, stderr = linregress(clref, cl)
-        r2 = r**2
-
-        # > plot the scatter plots for all 4 reference VOIs
-        ax[i, j].scatter(clref, cl, c='black')
-        identity_line(ax=ax[i, j], ls='--', c='b')
-        ax[i, j].text(60, 20, f'$y = {m:.4f}x + {a:.4f}$', fontsize=12)
-        ax[i, j].text(60, 10, f'$R^2={r2:.4f}$', fontsize=12)
-        ax[i, j].set_title(rvoi_str[rvoi])
-        ax[i, j].set_xlabel('CL ref')
-        ax[i, j].set_ylabel('CL AmyPET')
-        ax[i, j].grid('on')
-
-    fig.tight_layout()
-    # ------------------------------------------------------------------
-
-    return diff
-
-
-# ----------------------------------------------------------------------
-
-# ======================================================================
-# CL CALIBRATION FOR A NEW TRACER
-# ======================================================================
-
-
-def calib_tracer(outpib, outnew, xystr=None, figsize=(10, 10), fontsize=12):
-    '''calibrate CL scale for a new amyloid PET tracer,
-       given the dictionaries from processing the PiB
-       and a new tracer scans, i.e., `outpib` and `outnew`.
-
-       Other arguments:
-       xystr - dictionary of plot string XY locations.
-       figsize - tuple for controlling the figure size
-    '''
-
-    # > get PiB centiloid anchor points
-    clap = get_cl_anchors()
-
-    if xystr is None:
-        xystr = {'wc': [1.5, 1.1], 'cg': [2.0, 1.2], 'wcb': [1.5, 1.0], 'pns': [1.0, 0.75]}
-
-    # > extracting the index from names
-    pp = re.compile(r'(GE_\w*_\w*_(\d*)_NIFTI)|(Y*(\d*)\w*)')
-
-    # > calibration dictionary with SUVr and CL values
-    cal = {}
-    for rvoi in rvois:
-
-        print('------------------------------------------------------------')
-        print(rvoi_str[rvoi])
-
-        # > PiB and F18 SUVrs and PiB CL in one list/array
-        cl_suvr = []
-
-        # > initialise dictionary
-        cal[rvoi] = {'sbj': {}, 'calib': {}}
-
-        for k in outpib:
-
-            idx = pp.match(k)[2] or pp.match(k)[4]
-
-            # > get the key name in the NEW tracer dataset
-            # > (matching using the first 20 characters)
-            kf = [ky for ky in outnew if idx in ky[:20]] # idx+'_'
-            if len(kf) == 1:
-                kf = kf[0]
-            else:
-                raise ValueError('e> problem with F18 index...')
-
-            # > get the SUVr for PiB and FBB
-            suvrp = outpib[k]['suvr'][rvoi]
-            suvrf = outnew[kf]['suvr'][rvoi]
-
-            # > calculate the CL for the calibrating PiB scans
-            cl = 100 * (suvrp - clap[rvoi][0]) / (clap[rvoi][1] - clap[rvoi][0])
-
-            # > put the SUVrs for PiB and FBB into dictionary
-            cal[rvoi]['sbj'][idx] = {'cl': cl, 'suvrp': suvrp, 'suvrf': suvrf}
-
-            # > ... and a list for a future array
-            cl_suvr.append([cl, cal[rvoi]['sbj'][idx]['suvrp'], cal[rvoi]['sbj'][idx]['suvrf']])
-
-            print(f'refvoi={rvoi}, indx> {idx}, cl={cl:.3f},'
-                  f' suvr_pib={suvrp:.3f}, suvr_fbb={suvrf:.3f}')
-
-        # --------------------------------------------------------------
-        # > find the linear relationship for FBB_SUVr = m*PiB_SUVr + b
-        # > calculate the tracer `m_std` and `b_std`
-        # > (Eq. 2.2.3.1a in Klunk et al. 2015)
-
-        cl_suvr = np.array(cl_suvr)
-        m_std, b_std, r, p, stderr = linregress(cl_suvr[:, 1], cl_suvr[:, 2])
-        r2 = r**2
-
-        suvr_pib_calc = (cl_suvr[:, 2] - b_std) / m_std
-
-        cl_std_fbb = 100 * (suvr_pib_calc - clap[rvoi][0]) / (clap[rvoi][1] - clap[rvoi][0])
-
-        cal[rvoi]['calib'] = {
-            'm_std': m_std, 'b_std': b_std, 'r2': r2, 'p': p, 'stderr': stderr, 'cl_suvr': cl_suvr,
-            'suvr_pib_calc': suvr_pib_calc, 'cl_std_fbb': cl_std_fbb}
-        # --------------------------------------------------------------
-
-    # ------------------------------------------------------------------
-    # VISUALISATION
-
-    fig, ax = plt.subplots(2, 2, figsize=figsize)
-    fig2, ax2 = plt.subplots(2, 2, figsize=figsize)
-
-    for a, rvoi in enumerate(rvois):
-        i = a // 2
-        j = a % 2
-
-        ax[i, j].scatter(cal[rvoi]['calib']['cl_suvr'][:, 1], cal[rvoi]['calib']['cl_suvr'][:, 2],
-                         c='black')
-        identity_line(ax=ax[i, j], ls='--', c='b')
-        m_std = cal[rvoi]['calib']['m_std']
-        b_std = cal[rvoi]['calib']['b_std']
-        r2 = cal[rvoi]['calib']['r2']
-        ax[i, j].text(xystr[rvoi][0], xystr[rvoi][1], f'$y = {m_std:.4f}x + {b_std:.4f}$',
-                      fontsize=fontsize)
-        ax[i, j].text(xystr[rvoi][0], xystr[rvoi][1] - 0.1, f'$R^2={r2:.4f}$', fontsize=fontsize)
-        ax[i, j].set_title(rvoi_str[rvoi], fontweight='bold')
-        ax[i, j].set_xlabel(r'$^\mathrm{PiB}$SUVr$_\mathrm{IND}$', fontsize=fontsize)
-        ax[i, j].set_ylabel(r'$^\mathrm{NEW}$SUVr$_\mathrm{IND}$', fontsize=fontsize)
-        ax[i, j].grid('on')
-
-        low_x, high_x = ax[i, j].get_xlim()
-        x = np.linspace(low_x, high_x)
-        y = m_std*x + b_std
-        ax[i, j].plot(x, y, 'g')
-
-        ax2[i, j].scatter(cal[rvoi]['calib']['cl_suvr'][:, 0], cal[rvoi]['calib']['cl_std_fbb'],
-                          c='black')
-        identity_line(ax=ax2[i, j], ls='--', c='b')
-        ax2[i, j].set_xlabel('CL$^{**}$', fontsize=fontsize)
-        ax2[i, j].set_ylabel(r'$^\mathrm{NEW}$CL$_\mathrm{Std}$', fontsize=fontsize)
-        ax2[i, j].grid('on')
-        ax2[i, j].set_title(rvoi_str[rvoi])
-        # > add the same text about the equation used for converting to PiB SUVrs
-        ax2[i, j].text(
-            40, 10, f'$y_{{\\mathrm{{SUVr}}}}^{{\\mathrm{{NEW}}}}'
-            f' = {m_std:.4f}x_{{\\mathrm{{SUVr}}}}^{{\\mathrm{{PiB}}}} + {b_std:.4f}$',
-            fontsize=fontsize)
-        ax2[i, j].text(40, -5, f'$R^2={r2:.4f}$', fontsize=fontsize)
-
-    fig.tight_layout()
-    fig2.tight_layout()
-    # ------------------------------------------------------------------
-
-    return cal
-
-
-# ----------------------------------------------------------------------
-def save_cl_anchors(diff, outpath=None):
-    """save the CL anchor points for each reference VOI."""
-
-    CLA = {}
-    txt = '# centiloid anchor points for different reference VOIs\n'
-    for rvoi in rvois:
-        suvr_0 = diff['yc'][rvoi]['mean']
-        suvr_100 = diff['ad'][rvoi]['mean']
-        CLA[rvoi] = (suvr_0, suvr_100)
-        txt += f'cla_{rvoi} = ({suvr_0}, {suvr_100})\n'
-    print(txt)
-
-    if outpath is None:
-        outpath = cl_anchor_fldr
-
-    outpath = Path(outpath) / 'CL_PiB_anchors.pkl'
-    nimpa.create_dir(outpath.parent)
-    with open(outpath, 'wb') as f:
-        pickle.dump(CLA, f)
-
-    CLA['path'] = outpath
-    return CLA
-
-
-def get_cl_anchors(path=None):
-    """load the centiloid anchor points"""
-    if path is None:
-        path = cl_anchor_fldr
-    path = Path(path) / 'CL_PiB_anchors.pkl'
-    with open(path, 'rb') as f:
-        return pickle.load(f)
-
-
-# ----------------------------------------------------------------------
-
-
-# ----------------------------------------------------------------------
-def save_suvr2pib(cal, tracer, outpath=None):
-    ''' save the linear transformation parameters, `m_Std` and `b_Std`;
-        the parameters are taken from the bigger calibration dictionary `cal`
-        used here as input.
-
-        The compact dictionary consists the two parameters for all reference regions
-        and for the specific F-18 tracer being used.
-
-        `tracer` can be:
-            - 'fbb' for [18F]florbetaben
-            - 'flute' for [18F]flutemetamol
-            - 'fbp' for [18F]florbetapir
-    '''
-
-    if tracer not in ('fbp', 'fbb', 'flute'):
-        raise ValueError('tracer is unrecognised or not given!')
-
-    # conversion dictionary
-    CNV = {}
-    txt = '# Tracer SUVr conversion to PiB SUVr\n'
-    for rvoi in rvois:
-        mstd = cal[rvoi]['calib']['m_std']
-        bstd = cal[rvoi]['calib']['b_std']
-        CNV[rvoi] = {'m_std': mstd, 'b_std': bstd}
-        txt += f'{rvoi}: m_std, b_std = ({mstd}, {bstd})\n'
-    print(txt)
-
-    if outpath is None:
-        outpath = cl_anchor_fldr
-    outpath = Path(outpath)
-
-    nimpa.create_dir(outpath)
-    pth = outpath / f'suvr_{tracer}_to_suvr_pib__transform.pkl'
-    with open(pth, 'wb') as f:
-        pickle.dump(CNV, f)
-
-    CNV['path'] = pth
-    return CNV
-
-
-def get_suvr2pib(tracer, path=None):
-    """
-    load the linear transformation parameters from a tracer SUVr
-    to PiB SUVr; it's used for converting to CL scale any F-18
-    tracers.
-    """
-    if tracer not in ['fbp', 'fbb', 'flute']:
-        raise ValueError(f'tracer ({tracer}) unrecognised')
-
-    if path is None:
-        path = cl_anchor_fldr
-    pth = Path(path) / f'suvr_{tracer}_to_suvr_pib__transform.pkl'
-
-    with open(pth, 'rb') as f:
-        return pickle.load(f)
diff --git a/amypet/backend_centiloid.py b/amypet/backend_centiloid.py
index dc853949..c748a4d6 100644
--- a/amypet/backend_centiloid.py
+++ b/amypet/backend_centiloid.py
@@ -27,7 +27,7 @@
 from miutil.fdio import hasext, nsort
 from niftypet import nimpa
 
-from .aux import cl_anchor_fldr, cl_masks_fldr
+from .utils import cl_anchor_fldr, cl_masks_fldr
 
 log = logging.getLogger(__name__)
 
diff --git a/amypet/preproc.py b/amypet/preproc.py
index abf361c9..ca05deee 100644
--- a/amypet/preproc.py
+++ b/amypet/preproc.py
@@ -18,8 +18,8 @@
 import spm12
 from niftypet import nimpa
 
-from .aux import get_atlas
 from .suvr_tools import r_trimup
+from .utils import get_atlas
 
 log.basicConfig(level=log.WARNING, format=nimpa.LOG_FORMAT)
 
diff --git a/amypet/utils.py b/amypet/utils.py
index dfd33f06..28e03570 100644
--- a/amypet/utils.py
+++ b/amypet/utils.py
@@ -1,3 +1,14 @@
+import pickle
+import re
+import urllib
+from pathlib import Path
+
+import matplotlib.pyplot as plt
+import numpy as np
+from niftypet import nimpa
+from pkg_resources import resource_filename
+from scipy.stats import linregress
+
 try:
     from os import cpu_count
 except ImportError:
@@ -7,3 +18,550 @@
 
         def cpu_count():
             return 4
+
+
+# > regions used in CL project
+rvois = ['wc', 'cg', 'wcb', 'pns']
+
+# > folder name with the default conversion tables after calibration
+cl_anchor_fldr = Path(resource_filename(__name__, 'data/cl_anchor_tables'))
+cl_masks_fldr = Path(resource_filename(__name__, 'data/CL_masks'))
+amypet_dir = Path.home() / '.amypet'
+
+# > region full name strings for plots
+rvoi_str = {
+    'wc': 'WHOLE CEREBELLUM', 'cg': 'CEREBELLUM GM', 'wcb': 'WHOLE CEREBELLUM + BRAIN STEM',
+    'pns': 'PONS'}
+
+
+# ----------------------------------------------------------------------
+def get_atlas(atlas='aal', res=1, outpath=amypet_dir):
+    '''Get a brain atlas from `neuroparc` out of many available in MNI space.
+
+       Options:
+       - atlas:     one of many atlases, e.g., 'aal' which is the default;
+                    the number of atlases available can be extended by entering it
+                    in the dictionary. TODO: move the dictionary to `defs.py`.
+       - res:       the resolution of the atlas in mm.
+    '''
+    atlases = {
+        'ghpath': 'https://github.com/neurodata/neuroparc/raw/master/atlases/label/Human/',
+        'aal': 'AAL_space-MNI152NLin6_res-', 'aal_lbl': 'Anatomical-labels-csv/AAL.csv',
+        'schaefer200': 'Schaefer200_space-MNI152NLin6_res-',
+        'talairach': 'Talairach_space-MNI152NLin6_res-'}
+
+    # > get the GitHub data object
+    ghd = urllib.request.urlopen(atlases['ghpath'] + atlases[atlas] + f'{res}x{res}x{res}.nii.gz')
+    data = ghd.read()
+
+    # > save to file
+    nimpa.create_dir(outpath)
+
+    fatl = outpath / f'atlas-{atlas}_res-{res}mm.nii.gz'
+    with open(fatl, 'wb') as f:
+        f.write(data)
+
+    return fatl
+
+
+# ----------------------------------------------------------------------
+
+
+# ----------------------------------------------------------------------
+def im_check_pairs(fpets, fmris):
+    '''
+    checks visually image by image if the PET and MR have
+    the same orientation.
+    '''
+    fig, ax = plt.subplots(2, 2, figsize=(10, 10))
+    for fp, fm in zip(fpets, fmris):
+        print(f'{fp} : {fm}')
+        p = nimpa.getnii(fp)
+        m = nimpa.getnii(fm)
+
+        ps = p.shape
+        ms = m.shape
+
+        ax[0, 0].imshow(p[ps[0] // 2, ...], cmap='magma')
+        ax[1, 0].imshow(m[ms[0] // 2, ...], cmap='bone')
+
+        ax[0, 1].imshow(p[..., ps[2] // 2], cmap='magma')
+        ax[1, 1].imshow(m[..., ms[2] // 2], cmap='bone')
+
+        plt.draw()
+        plt.waitforbuttonpress(0)
+
+
+# ----------------------------------------------------------------------
+
+
+# ----------------------------------------------------------------------
+def identity_line(ax=None, ls='--', *args, **kwargs):
+    '''
+    plot identity line for any correlation analysis
+    needed for the CL calibration and beyond.
+    '''
+    # see: https://stackoverflow.com/q/22104256/3986320
+    ax = ax or plt.gca()
+    identity, = ax.plot([], [], ls=ls, *args, **kwargs)
+
+    def callback(axes):
+        low_x, high_x = ax.get_xlim()
+        low_y, high_y = ax.get_ylim()
+        low = min(low_x, low_y)
+        high = max(high_x, high_y)
+        identity.set_data([low, high], [low, high])
+
+    callback(ax)
+    ax.callbacks.connect('xlim_changed', callback)
+    ax.callbacks.connect('ylim_changed', callback)
+    return ax
+
+
+# ----------------------------------------------------------------------
+
+
+# ----------------------------------------------------------------------
+def get_clref(fxls):
+    '''
+    get the excel sheets with CL reference data from the CL project
+    excel sheet.
+    '''
+    import openpyxl
+
+    # > Define variable to load the workbook
+    wb = openpyxl.load_workbook(fxls)
+    ws = wb.active
+
+    pib_tbl = {'yc': {}, 'ad': {}}
+
+    # > cell offsets
+    ioff_ad = [5, 50]
+    ioff_yc = [51, 85]
+
+    sbj = [str(i.value) for i in ws['A']]
+    cg = [str(i.value) for i in ws['B']]
+    wc = [str(i.value) for i in ws['C']]
+    wcb = [str(i.value) for i in ws['D']]
+    pns = [str(i.value) for i in ws['E']]
+
+    ccg = [str(i.value) for i in ws['F']]
+    cwc = [str(i.value) for i in ws['G']]
+    cwcb = [str(i.value) for i in ws['H']]
+    cpns = [str(i.value) for i in ws['I']]
+
+    # > young control indices and SUVr/CL values
+    pib_tbl['yc']['id'] = np.array([int(i[3:]) for i in sbj[ioff_yc[0]:ioff_yc[1]]])
+
+    pib_tbl['yc']['suvr'] = {}
+    pib_tbl['yc']['suvr']['cg'] = np.array([float(i) for i in cg[ioff_yc[0]:ioff_yc[1]]])
+    pib_tbl['yc']['suvr']['wc'] = np.array([float(i) for i in wc[ioff_yc[0]:ioff_yc[1]]])
+    pib_tbl['yc']['suvr']['wcb'] = np.array([float(i) for i in wcb[ioff_yc[0]:ioff_yc[1]]])
+    pib_tbl['yc']['suvr']['pns'] = np.array([float(i) for i in pns[ioff_yc[0]:ioff_yc[1]]])
+
+    pib_tbl['yc']['cl'] = {}
+    pib_tbl['yc']['cl']['cg'] = np.array([float(i) for i in ccg[ioff_yc[0]:ioff_yc[1]]])
+    pib_tbl['yc']['cl']['wc'] = np.array([float(i) for i in cwc[ioff_yc[0]:ioff_yc[1]]])
+    pib_tbl['yc']['cl']['wcb'] = np.array([float(i) for i in cwcb[ioff_yc[0]:ioff_yc[1]]])
+    pib_tbl['yc']['cl']['pns'] = np.array([float(i) for i in cpns[ioff_yc[0]:ioff_yc[1]]])
+
+    # > AD indices and SUVr/CL
+    pib_tbl['ad']['id'] = np.array([int(i[3:]) for i in sbj[ioff_ad[0]:ioff_ad[1]]])
+    pib_tbl['ad']['suvr'] = {}
+    pib_tbl['ad']['suvr']['cg'] = np.array([float(i) for i in cg[ioff_ad[0]:ioff_ad[1]]])
+    pib_tbl['ad']['suvr']['wc'] = np.array([float(i) for i in wc[ioff_ad[0]:ioff_ad[1]]])
+    pib_tbl['ad']['suvr']['wcb'] = np.array([float(i) for i in wcb[ioff_ad[0]:ioff_ad[1]]])
+    pib_tbl['ad']['suvr']['pns'] = np.array([float(i) for i in pns[ioff_ad[0]:ioff_ad[1]]])
+
+    pib_tbl['ad']['cl'] = {}
+    pib_tbl['ad']['cl']['cg'] = np.array([float(i) for i in ccg[ioff_ad[0]:ioff_ad[1]]])
+    pib_tbl['ad']['cl']['wc'] = np.array([float(i) for i in cwc[ioff_ad[0]:ioff_ad[1]]])
+    pib_tbl['ad']['cl']['wcb'] = np.array([float(i) for i in cwcb[ioff_ad[0]:ioff_ad[1]]])
+    pib_tbl['ad']['cl']['pns'] = np.array([float(i) for i in cpns[ioff_ad[0]:ioff_ad[1]]])
+
+    return pib_tbl
+    # ----------------------------------------------------------------------
+
+
+def check_suvrs(suvr_yc, suvr_ad, refs):
+    ''' Check/QC the obtained SUVRs in `suvr_yc` and `suvr_ad `dictionaries
+        in relation to the reference in `refs` dictionary for the groups
+        ('ad' or 'yc').
+
+        Returns dictionary with the mean SUVrs and differences.
+    '''
+
+    # > prepare diff dictionary for the differences observed
+    # > in young controls (yc) and AD patients (ad)
+    diff = {'yc': {}, 'ad': {}}
+
+    # > initialise the means for each reference VOI
+    for rvoi in rvois:
+        diff['yc'][rvoi] = {'mean_ref': 0, 'mean': 0, 'N': 0}
+        diff['ad'][rvoi] = {'mean_ref': 0, 'mean': 0, 'N': 0}
+
+    def run_checks(suvr_dct, grp):
+
+        for rvoi in rvois:
+            print('========================================================')
+            for k in suvr_dct:
+
+                if grp == 'yc':
+                    idx = int(k[2:5])
+                elif grp == 'ad':
+                    idx = int(k[2:4])
+                else:
+                    raise ValueError('e> unknown group - only <yc>  or <ad> are accepted')
+
+                i = np.where(refs[grp]['id'] == idx)[0][0]
+
+                suvr = suvr_dct[k]['suvr'][rvoi]
+                suvr_ref = refs[grp]['suvr'][rvoi][i]
+                err = 100 * (suvr-suvr_ref) / suvr_ref
+
+                diff[grp][rvoi]['N'] += 1
+                diff[grp][rvoi]['mean_ref'] += suvr_ref
+                diff[grp][rvoi]['mean'] += suvr
+                diff[grp][rvoi][k] = {'suvr': suvr, 'ref': suvr_ref, 'err': err}
+
+                print(f'refvoi={rvoi}, indx> {idx}, suvr={suvr:.3f},'
+                      f' ref={suvr_ref:.3f}, error={err:.3f}%')
+
+            diff[grp][rvoi]['mean'] /= diff[grp][rvoi]['N']
+            diff[grp][rvoi]['mean_ref'] /= diff[grp][rvoi]['N']
+
+            # relative % mean difference
+            emean = diff[grp][rvoi]['mean']
+            rmean = diff[grp][rvoi]['mean_ref']
+            rmd = (emean-rmean) / rmean
+            diff[grp][rvoi]['mean_diff'] = rmd
+            print('--------------------------------------------------------')
+            print(f'> group % mean difference: {rmd:.3f}% (suvr={emean:.3f}, ref={rmean:.3f})')
+
+    print('========================================================')
+    print('YOUNG CONTROLS (YC)')
+    print('========================================================')
+    run_checks(suvr_yc, 'yc')
+
+    print('========================================================')
+    print('AD PATIENTS (AD)')
+    print('========================================================')
+    run_checks(suvr_ad, 'ad')
+
+    return diff
+
+
+# ----------------------------------------------------------------------
+
+
+# ----------------------------------------------------------------------
+def check_cls(suvr_yc, suvr_ad, diff, refs):
+    ''' check centiloid values relative to the reference in `refs`
+        dictionary, using SUVr dictionaries for YC and AD groups
+        `suvr_yc` and `suvr_ad`.  Also, update the `diff` dictionary
+        with the centiloid results and output it.
+    '''
+
+    for rvoi in rvois:
+        print('=====================================================')
+        suvr_0 = diff['yc'][rvoi]['mean']
+        suvr_100 = diff['ad'][rvoi]['mean']
+
+        grp = 'yc'
+        suvr_dct = suvr_yc
+
+        for k in suvr_dct:
+
+            idx = int(k[2:5])
+            i = np.where(refs[grp]['id'] == idx)[0][0]
+
+            cl = 100 * (suvr_dct[k]['suvr'][rvoi] - suvr_0) / (suvr_100-suvr_0)
+            cl_ref = refs[grp]['cl'][rvoi][i]
+
+            diff[grp][rvoi][k].update(cl=cl, cl_ref=cl_ref)
+
+            print(f'refvoi={rvoi}, indx> {idx}, cl={cl:.3f}, ref={cl_ref:.3f}')
+
+        grp = 'ad'
+        suvr_dct = suvr_ad
+
+        for k in suvr_dct:
+
+            idx = int(k[2:4])
+            i = np.where(refs[grp]['id'] == idx)[0][0]
+
+            cl = 100 * (suvr_dct[k]['suvr'][rvoi] - suvr_0) / (suvr_100-suvr_0)
+            cl_ref = refs[grp]['cl'][rvoi][i]
+
+            diff[grp][rvoi][k].update(cl=cl, cl_ref=cl_ref)
+
+            print(f'refvoi={rvoi}, indx> {idx}, cl={cl:.3f}, ref={cl_ref:.3f}')
+
+    def get_cls(diff, grp, rvoi, cl='cl'):
+        ''' Get centiloid values in a list for any group and reference VOI.
+        '''
+        return [diff[grp][rvoi][k][cl] for k in diff[grp][rvoi] if grp in k.lower()]
+
+    # ------------------------------------------------------------------
+    fig, ax = plt.subplots(2, 2, figsize=(10, 10))
+
+    for a, rvoi in enumerate(rvois):
+        i = a // 2
+        j = a % 2
+
+        # > combine CLs for both groups
+        cl = get_cls(diff, 'yc', rvoi, cl='cl')
+        cl += get_cls(diff, 'ad', rvoi, cl='cl')
+
+        # > combine reference CLs into one list
+        clref = get_cls(diff, 'yc', rvoi, cl='cl_ref')
+        clref += get_cls(diff, 'ad', rvoi, cl='cl_ref')
+
+        # > find the linear regression between the CL and reference CL
+        m, a, r, p, stderr = linregress(clref, cl)
+        r2 = r**2
+
+        # > plot the scatter plots for all 4 reference VOIs
+        ax[i, j].scatter(clref, cl, c='black')
+        identity_line(ax=ax[i, j], ls='--', c='b')
+        ax[i, j].text(60, 20, f'$y = {m:.4f}x + {a:.4f}$', fontsize=12)
+        ax[i, j].text(60, 10, f'$R^2={r2:.4f}$', fontsize=12)
+        ax[i, j].set_title(rvoi_str[rvoi])
+        ax[i, j].set_xlabel('CL ref')
+        ax[i, j].set_ylabel('CL AmyPET')
+        ax[i, j].grid('on')
+
+    fig.tight_layout()
+    # ------------------------------------------------------------------
+
+    return diff
+
+
+# ----------------------------------------------------------------------
+
+# ======================================================================
+# CL CALIBRATION FOR A NEW TRACER
+# ======================================================================
+
+
+def calib_tracer(outpib, outnew, xystr=None, figsize=(10, 10), fontsize=12):
+    '''calibrate CL scale for a new amyloid PET tracer,
+       given the dictionaries from processing the PiB
+       and a new tracer scans, i.e., `outpib` and `outnew`.
+
+       Other arguments:
+       xystr - dictionary of plot string XY locations.
+       figsize - tuple for controlling the figure size
+    '''
+
+    # > get PiB centiloid anchor points
+    clap = get_cl_anchors()
+
+    if xystr is None:
+        xystr = {'wc': [1.5, 1.1], 'cg': [2.0, 1.2], 'wcb': [1.5, 1.0], 'pns': [1.0, 0.75]}
+
+    # > extracting the index from names
+    pp = re.compile(r'(GE_\w*_\w*_(\d*)_NIFTI)|(Y*(\d*)\w*)')
+
+    # > calibration dictionary with SUVr and CL values
+    cal = {}
+    for rvoi in rvois:
+
+        print('------------------------------------------------------------')
+        print(rvoi_str[rvoi])
+
+        # > PiB and F18 SUVrs and PiB CL in one list/array
+        cl_suvr = []
+
+        # > initialise dictionary
+        cal[rvoi] = {'sbj': {}, 'calib': {}}
+
+        for k in outpib:
+
+            idx = pp.match(k)[2] or pp.match(k)[4]
+
+            # > get the key name in the NEW tracer dataset
+            # > (matching using the first 20 characters)
+            kf = [ky for ky in outnew if idx in ky[:20]] # idx+'_'
+            if len(kf) == 1:
+                kf = kf[0]
+            else:
+                raise ValueError('e> problem with F18 index...')
+
+            # > get the SUVr for PiB and FBB
+            suvrp = outpib[k]['suvr'][rvoi]
+            suvrf = outnew[kf]['suvr'][rvoi]
+
+            # > calculate the CL for the calibrating PiB scans
+            cl = 100 * (suvrp - clap[rvoi][0]) / (clap[rvoi][1] - clap[rvoi][0])
+
+            # > put the SUVrs for PiB and FBB into dictionary
+            cal[rvoi]['sbj'][idx] = {'cl': cl, 'suvrp': suvrp, 'suvrf': suvrf}
+
+            # > ... and a list for a future array
+            cl_suvr.append([cl, cal[rvoi]['sbj'][idx]['suvrp'], cal[rvoi]['sbj'][idx]['suvrf']])
+
+            print(f'refvoi={rvoi}, indx> {idx}, cl={cl:.3f},'
+                  f' suvr_pib={suvrp:.3f}, suvr_fbb={suvrf:.3f}')
+
+        # --------------------------------------------------------------
+        # > find the linear relationship for FBB_SUVr = m*PiB_SUVr + b
+        # > calculate the tracer `m_std` and `b_std`
+        # > (Eq. 2.2.3.1a in Klunk et al. 2015)
+
+        cl_suvr = np.array(cl_suvr)
+        m_std, b_std, r, p, stderr = linregress(cl_suvr[:, 1], cl_suvr[:, 2])
+        r2 = r**2
+
+        suvr_pib_calc = (cl_suvr[:, 2] - b_std) / m_std
+
+        cl_std_fbb = 100 * (suvr_pib_calc - clap[rvoi][0]) / (clap[rvoi][1] - clap[rvoi][0])
+
+        cal[rvoi]['calib'] = {
+            'm_std': m_std, 'b_std': b_std, 'r2': r2, 'p': p, 'stderr': stderr, 'cl_suvr': cl_suvr,
+            'suvr_pib_calc': suvr_pib_calc, 'cl_std_fbb': cl_std_fbb}
+        # --------------------------------------------------------------
+
+    # ------------------------------------------------------------------
+    # VISUALISATION
+
+    fig, ax = plt.subplots(2, 2, figsize=figsize)
+    fig2, ax2 = plt.subplots(2, 2, figsize=figsize)
+
+    for a, rvoi in enumerate(rvois):
+        i = a // 2
+        j = a % 2
+
+        ax[i, j].scatter(cal[rvoi]['calib']['cl_suvr'][:, 1], cal[rvoi]['calib']['cl_suvr'][:, 2],
+                         c='black')
+        identity_line(ax=ax[i, j], ls='--', c='b')
+        m_std = cal[rvoi]['calib']['m_std']
+        b_std = cal[rvoi]['calib']['b_std']
+        r2 = cal[rvoi]['calib']['r2']
+        ax[i, j].text(xystr[rvoi][0], xystr[rvoi][1], f'$y = {m_std:.4f}x + {b_std:.4f}$',
+                      fontsize=fontsize)
+        ax[i, j].text(xystr[rvoi][0], xystr[rvoi][1] - 0.1, f'$R^2={r2:.4f}$', fontsize=fontsize)
+        ax[i, j].set_title(rvoi_str[rvoi], fontweight='bold')
+        ax[i, j].set_xlabel(r'$^\mathrm{PiB}$SUVr$_\mathrm{IND}$', fontsize=fontsize)
+        ax[i, j].set_ylabel(r'$^\mathrm{NEW}$SUVr$_\mathrm{IND}$', fontsize=fontsize)
+        ax[i, j].grid('on')
+
+        low_x, high_x = ax[i, j].get_xlim()
+        x = np.linspace(low_x, high_x)
+        y = m_std*x + b_std
+        ax[i, j].plot(x, y, 'g')
+
+        ax2[i, j].scatter(cal[rvoi]['calib']['cl_suvr'][:, 0], cal[rvoi]['calib']['cl_std_fbb'],
+                          c='black')
+        identity_line(ax=ax2[i, j], ls='--', c='b')
+        ax2[i, j].set_xlabel('CL$^{**}$', fontsize=fontsize)
+        ax2[i, j].set_ylabel(r'$^\mathrm{NEW}$CL$_\mathrm{Std}$', fontsize=fontsize)
+        ax2[i, j].grid('on')
+        ax2[i, j].set_title(rvoi_str[rvoi])
+        # > add the same text about the equation used for converting to PiB SUVrs
+        ax2[i, j].text(
+            40, 10, f'$y_{{\\mathrm{{SUVr}}}}^{{\\mathrm{{NEW}}}}'
+            f' = {m_std:.4f}x_{{\\mathrm{{SUVr}}}}^{{\\mathrm{{PiB}}}} + {b_std:.4f}$',
+            fontsize=fontsize)
+        ax2[i, j].text(40, -5, f'$R^2={r2:.4f}$', fontsize=fontsize)
+
+    fig.tight_layout()
+    fig2.tight_layout()
+    # ------------------------------------------------------------------
+
+    return cal
+
+
+# ----------------------------------------------------------------------
+def save_cl_anchors(diff, outpath=None):
+    """save the CL anchor points for each reference VOI."""
+
+    CLA = {}
+    txt = '# centiloid anchor points for different reference VOIs\n'
+    for rvoi in rvois:
+        suvr_0 = diff['yc'][rvoi]['mean']
+        suvr_100 = diff['ad'][rvoi]['mean']
+        CLA[rvoi] = (suvr_0, suvr_100)
+        txt += f'cla_{rvoi} = ({suvr_0}, {suvr_100})\n'
+    print(txt)
+
+    if outpath is None:
+        outpath = cl_anchor_fldr
+
+    outpath = Path(outpath) / 'CL_PiB_anchors.pkl'
+    nimpa.create_dir(outpath.parent)
+    with open(outpath, 'wb') as f:
+        pickle.dump(CLA, f)
+
+    CLA['path'] = outpath
+    return CLA
+
+
+def get_cl_anchors(path=None):
+    """load the centiloid anchor points"""
+    if path is None:
+        path = cl_anchor_fldr
+    path = Path(path) / 'CL_PiB_anchors.pkl'
+    with open(path, 'rb') as f:
+        return pickle.load(f)
+
+
+# ----------------------------------------------------------------------
+
+
+# ----------------------------------------------------------------------
+def save_suvr2pib(cal, tracer, outpath=None):
+    ''' save the linear transformation parameters, `m_Std` and `b_Std`;
+        the parameters are taken from the bigger calibration dictionary `cal`
+        used here as input.
+
+        The compact dictionary consists the two parameters for all reference regions
+        and for the specific F-18 tracer being used.
+
+        `tracer` can be:
+            - 'fbb' for [18F]florbetaben
+            - 'flute' for [18F]flutemetamol
+            - 'fbp' for [18F]florbetapir
+    '''
+
+    if tracer not in ('fbp', 'fbb', 'flute'):
+        raise ValueError('tracer is unrecognised or not given!')
+
+    # conversion dictionary
+    CNV = {}
+    txt = '# Tracer SUVr conversion to PiB SUVr\n'
+    for rvoi in rvois:
+        mstd = cal[rvoi]['calib']['m_std']
+        bstd = cal[rvoi]['calib']['b_std']
+        CNV[rvoi] = {'m_std': mstd, 'b_std': bstd}
+        txt += f'{rvoi}: m_std, b_std = ({mstd}, {bstd})\n'
+    print(txt)
+
+    if outpath is None:
+        outpath = cl_anchor_fldr
+    outpath = Path(outpath)
+
+    nimpa.create_dir(outpath)
+    pth = outpath / f'suvr_{tracer}_to_suvr_pib__transform.pkl'
+    with open(pth, 'wb') as f:
+        pickle.dump(CNV, f)
+
+    CNV['path'] = pth
+    return CNV
+
+
+def get_suvr2pib(tracer, path=None):
+    """
+    load the linear transformation parameters from a tracer SUVr
+    to PiB SUVr; it's used for converting to CL scale any F-18
+    tracers.
+    """
+    if tracer not in ['fbp', 'fbb', 'flute']:
+        raise ValueError(f'tracer ({tracer}) unrecognised')
+
+    if path is None:
+        path = cl_anchor_fldr
+    pth = Path(path) / f'suvr_{tracer}_to_suvr_pib__transform.pkl'
+
+    with open(pth, 'rb') as f:
+        return pickle.load(f)
diff --git a/cl_scripts/centi_fbp.py b/cl_scripts/centi_fbp.py
index 7a37a146..5c12dc1b 100644
--- a/cl_scripts/centi_fbp.py
+++ b/cl_scripts/centi_fbp.py
@@ -220,4 +220,4 @@
 
 fig, ax = plt.subplots()
 ax.scatter(cal[rvoi]['calib']['cl_suvr'][:, 1], cal[rvoi]['calib']['cl_suvr'][:, 2], c='black')
-amypet.aux.identity_line(ax=ax, ls='--', c='b')
+amypet.utils.identity_line(ax=ax, ls='--', c='b')
diff --git a/cl_scripts/centi_pib.py b/cl_scripts/centi_pib.py
index 69219a9b..aff8ab0d 100644
--- a/cl_scripts/centi_pib.py
+++ b/cl_scripts/centi_pib.py
@@ -10,8 +10,8 @@
 from scipy.stats import linregress
 
 import amypet
-from amypet.aux import cl_anchor_fldr
 from amypet.centiloid import run as centiloid_run
+from amypet.utils import cl_anchor_fldr
 
 # > input paths
 drv = Path('/data/AMYPET')