CT has excellent anatomic resolution and distortion-free images. It is still used for brain anatomy and disease diagnosis, although it has largely been replaced by MRI.

Several early studies used CT to assess differences between adult sexual offenders and nonviolent, nonsexual control subjects (7–12). The results have been mixed. In about one-half the studies, no significant group differences were found. Those that found differences tended to find greater temporal lobe abnormalities in sex offenders, especially if violent. One study that did not use a nonoffender control group found that 35% of violent incest offenders had temporal lobe abnormalities, compared with only 13% of nonviolent incest offenders (13). Sample sizes in all these studies were somewhat small, and actual intergroup differences would have been hard to identify. Even when CT has been combined with neuropsychological testing, results have been inconsistent. Only 2 studies found significant differences between adult sex-offender groups and control subjects (7,8); however, only 8% of subjects actually demonstrated pathology on both CT and neuropsychological testing (8).

Other CT studies have focused on aggressive subjects. Herzberg and Fenwick reported no significant differences between aggressive patients with temporal lobe epilepsy and the nonaggressive group (14). In 1991, Tonkonogy found that 5 out of 14 violent patients with organic mental syndromes had a local lesion in their anterior-inferior temporal lobes, although the other 9 did not (15). The author suggested that destruction of amygdaloid nuclei or adjacent limbic structures may play a role in violence, along with kindling of preserved limbic areas. In summary, there have been few CT studies, and even fewer using neuropsychological testing concurrently. Results vary, and samples are usually small and limited to the prison population. Thus, conclusions are tentative at best. Further, CT has the disadvantages of using ionizing radiation, and generating “bone artifacts” where tissue meets the skull; as well, it is limited to 2 dimensions only. Hence, we now turn to MRI.

Magnetic Resonance Imaging (MRI)

MRI is based on the principles of nuclear magnetic resonance imaging, invented in 1971.

Atoms within the human body possess magnetic charge, especially when subjected to the large magnetic field of the MRI apparatus. After being brought into alignment by the MRI electromagnetic waves, tissue atoms are then exposed to radio waves. Each active nucleus resonates at a particular radio frequency. The release of radio waves as the nuclei return to their normal state is captured by detectors, and the information is manipulated mathematically by computer to produce the image. MRI has excellent anatomic resolution and involves no radiation. Functional MRI uses the differences in intensity during activation or task periods and relaxation periods to look at brain function.

The temporal lobe has been implicated in antisocial behaviour since the early MRI studies. In addition to the 1991 Tonkonogy study (15), Chesterman found that 6 out of 10 violent male psychiatric inpatients had mesial temporal atrophy (16). This, however, was a small, heterogeneous population comprising 6 men with schizophrenia and 4 with personality disorders. Recently, patients with temporal lobe epilepsy and intermittent explosive disorder were also found, more often than were control subjects, to have amygdala atrophy or periamygdaloid lesions (17). Again, however, over 70% of the patients did not exhibit these abnormalities on MRI. The advantage of this study, though, was its examination of only affective-impulsive aggression, leading to a more homogeneous group. Most studies have by now found abnormal brain morphological changes, particularly in murderers (18) and violent sex offenders (19), especially in frontotemporal areas. Again, most of these studies have not been quantitative, had heterogeneous samples (for example, murderers and nonmurderers), used different methodologies to clinically measure variables such as violence, and due to technological limits, could not further localize and specify precise neuroanatomic regions.

There have, however, been a few quantitative studies. Raine and others found that 21 men with antisocial personality disorder (APD), living in the community, had at least 11% less prefrontal gray matter when compared with nonclinical control subjects, men with substance dependence, or male psychiatric control subjects (20).