In brain imaging, two complementary but technologically contradicting techniques are magnetoencephalography (MEG) and magnetic resonance imaging (MRI). MEG examines the function of the brain by measuring very weak magnetic fields, produced as a result of neuronal activity, with sensors based on superconducting quantum interference device (SQUID). MRI employs large magnetic fields and enables imaging of the structure of matter. The recent advances in ultra-low-field (ULF) MRI have made a medical instrument incorporating MEG and ULF MRI functionalities an attractive topic of research. The contradictions become evident when comparatively high fields of ULF MRI are subjected to SQUID magnetic field sensors, degrading their performance. In this thesis, the field tolerance of the sensors was improved. Special attention was paid to sensor response recovery and operation after a magnetic pulse. A hybrid MEG-ULF MRI instrument was constructed with the aid of new sensors. The instrument operation was verified, and results indicate that including ULF MRI in a MEG device is a viable concept. In addition, a new type of magnetometer was developed, taking advantage of the nonlinear kinetic inductance of superconducting material. The experimental data, together with the theory, demonstrate a device with low noise and intrinsically high dynamic range. Furthermore, the kinetic inductance magnetometer is suitable for biomagnetic multichannel measurements, as only one amplifier is needed in the readout of multiple sensors. The simple design reduces costs in fabrication and enables higher tolerance of magnetic fields than achievable with SQUID sensors. A new superconducting transformer design is introduced as a final step. Connecting to a SQUID results in a highly sensitive current detector. The device is a candidate for closing the quantum metrology triangle (QMT) experiment, a long-standing goal in metrology. The aim of the QMT is to improve confidence in the planned revision of the SI unit system by comparing the quantum standards of current, voltage and resistance. The device was characterized for the purpose by using it as a null current detector in the simulation of a QMT experiment. Results disclose the potential of the device and provide insight into some of the practical challenges relevant to null detection.