Important advances in neurology often begin with an intellectually curious physician observing an interesting patient. Such was the case that led Dr. Aaron Boes to the discovery of lesion network mapping.
Aaron was a resident in child neurology at Massachusetts General Hospital when he encountered an unusual teenager. The patient was a previously healthy 17-year-old girl, who had the sudden onset of visual hallucinations. She was previously a straight-A student and captain of her soccer team, but she was suddenly impaired by the presence of vivid visual symptoms, including the perception of objects coming in and out of focus “like a zoom lens” and of the scenery around her being drawn in by crayon. A toxicology screen and head CT scan were negative, but an MRI scan of the brain revealed a small stroke in the thalamus.
Aaron realized that the MRI results could explain the etiology of the patient’s symptoms, but could not adequately explain their mechanism. Why would a small stroke in the non-visual region of the thalamus induce visual hallucinations? Aaron hypothesized that this condition—peduncular hallucinosis—is due to brain dysfunction not restricted to the lesion site itself, but to neuronal networks connected to the lesion.
With the help of his colleagues, Aaron investigated the lesions of 23 patients with peduncular hallucinosis. He utilized the 3D volume of each lesion in a large normative database of functional connectivity MRI to examine the networks associated with each lesion location. In this way, he was able to infer the remote sites impacted by the lesions.
Dr. Boes found that there was much more overlap in the networks associated with the lesions than in the lesion locations themselves. He further found that 22 of the 23 studied lesions in peduncular hallucinosis fell along a single network with connectivity to the ventral extrastriate visual cortex, a brain region long-hypothesized to be involved in the generation of hallucinations.
Currently, neurologists’ ability to accurately predkt the clinical outcome of stroke and other focal lesions is poor. Dr. Boes hypothesizes that information regarding lesion location and the networks associated with the lesion can improve the accuracy of outcome predictions.
To conduct these studies, Dr. Boes developed a novel method to assess the effects of cerebral lesions on remote brain regions and networks. The technique involves mapping a brain lesion from a clinical scan onto a reference brain, then using the lesion volume as a seed region of interest for a resting state functional connectivity MRI analysis that uses normative data from a large representative cohort (see figure). Aaron named this technique lesion network mapping and published his findings in Brain (138: 3061-3075).
Now an Assistant Professor of Pediatrics at the University of Iowa, Dr. Boes utilizes lesion network mapping in several exciting projects. In one project, he is examining the neural basis for posterior fossa syndrome (PFS). Following surgery of the cerebellum, many pediatric patients develop dramatic neuropsychiatric symptoms, including mutism and other behavioral problems. With the use of lesion network mapping, Dr. Boes is examining the neuroanatomical and neural network basis for this syndrome. Elucidation of the involved structures could guide neurosurgical approaches to avoid PFS. In addition, knowledge of the lesion-associated networks could augment rehabilitation for those who develop PFS.
In a second project, Dr. Boes is attempting to improve the accuracy of prognosis following the onset of focal brain lesions. Currently, neurologists’ ability to accurately predict the clinical outcome of stroke and other focal lesions is poor. Dr. Boes hypothesizes that information regarding lesion location and the networks associated with the lesion can improve the accuracy of outcome predictions.
In a third set of projects, Dr. Boes is studying TMS. He is working with neurosurgical colleagues, who implant intracranial electrodes, to examine the remote effects of TMS on network dynamics and to identify the basic mechanisms through which TMS improves outcome in some patients with depression. He hopes that understanding the basic mechanisms of TMS will expand its clinical indications.
Dr. Boes’ research idea—utilizing network localization to explain symptoms—originated from a single patient with visual hallucinations. The technique is proving versatile and durable, as it is now being used in a wide range of applications, and may, someday, become a standard tool in neurology.
Figure Legends: Lesion network mapping involves two steps. First, a brain lesion from a clinical scan is mapped onto a reference brain (columns 1 & 2). Second, the lesion volume is used as a seed region of interest for a resting state functional connectivity MRI analysis that uses normative data (column 3). The lesion-associated networks derived from the latter step can be used for research or clinical applications, providing more information than is available from the lesion location alone.
Dr. Boes directs the Noninvasive Brain Stimulation Clinical Program at the University of Iowa and the Iowa Neuroimaging and Noninvasive Brain Stimulation Laboratory.