A method and apparatus for testing for cancer cells in the human breast is set forth. In the preferred and illustrated embodiment, a pendulant breast exposed to a magnetic field is tested by nuclear magnetic resonant (NMR). The magnetic field is varied to provide the requisite field intensity in selected segmented sections of the pendulant breast within the field. The field intensity is varied to thereby observe all of the breast extending toward the thorax cavity, obtaining test data from section after section aggregating test data from the entire breast within the pole pieces of the magnet and sufficiently therebeyond to obtain full interrogation. The NMR response focuses on hydrogen nuclei in the water in the breast region, determines the voltage proportional to the density of hydrogen nuclei and the binding of the water described by the relaxation times T.sub.1 and T.sub.2 for each binding level for water. The data so obtained in sequence is examined. Difference values from sequentially obtained signals are evaluated, yielding a plot of different signals as a function of breast length, thereby yielding abnormalities. Such abnormalities are shown in the difference signal from NMR interrogation as a result the concentration and relaxation time (T.sub.1 and T.sub.2) changes in tumo
William L. Rollwitz;