How the brain predicts the face: Teaching the brain as a co‐architect of the face

As educators of future clinicians, how do we encourage learners to appreciate the predictive power and beauty of anatomy? One persuasive approach is to teach anatomy as relationships that explain disease, guide diagnosis and treatment, and inspire new discoveries. This is well illustrated in the intimate relationship between the development of the forebrain and midface. As clinicians have long noted,1 “the face predicts the brain,” a clinico-radiologic principle crystallized in holoprosencephaly (HPE), where degrees of failed hemispheric cleavage parallel a spectrum of midline facial anomalies. HPE offers a compact story that first-year learners can grasp.2 During the third to the fourth week of development (days ~18–28), Sonic hedgehog (SHH) signaling from the prechordal plate patterns the ventral forebrain and establishes the midline. When this ventral induction is reduced, prosencephalic cleavage can be incomplete. The phenotypic spectrum ranges from alobar HPE with severe facial changes to lobar and middle-interhemispheric (syntelencephaly) variants with subtler facial differences.3 Classic severe changes include cyclopia with a single midline orbit and a proboscis superior to the eye; subtler findings include hypotelorism or a solitary median maxillary central incisor. These variations show how even modest shifts in midline patterning are mirrored in facial morphology. By naming the core mechanism and asking students to predict facial outcomes from the brain's midline status, embryology becomes a generative exercise rather than a list of derivatives. But why teach the brain and face together in undergraduate medical education (UME)? First, curricular time for embryology is constrained (approximately 14 h in US programs4 and about 7 h in Canadian programs5), and learners often encounter neuroanatomy and craniofacial development in different courses or blocks, which can make their shared mechanisms less visible. Integrating forebrain patterning with midface development gives learners a single explanatory thread that makes the shared mechanism explicit and helps limited instructional time work harder. Second, the integration aligns anatomy with clinical reasoning by using a single annotated fetal MRI to anchor mechanistic understanding, build basic image interpretation skills, and link findings to diagnostic and management decisions. In practice, this integration can be delivered in a brief, evidence-informed sequence. Begin with one image and one mechanism: present an axial fetal MRI showing incomplete interhemispheric separation and sketch the SHH gradient from the prechordal plate, linking an underspecified midline to expected facial findings such as hypotelorism, a median cleft, or a solitary median maxillary central incisor. Before displaying the facial image or 3D rendering, ask learners to predict the findings from the mechanism, then compare their reasoning with the reveal. To ease cognitive load and strengthen spatial reasoning, add a second modality, either a short radiology-anchored module6 or a 10- to 15-min clay-based modeling7 of the frontonasal and maxillary prominences. Conclude by connecting the mechanism to clinical decisions, including patterns of variable expressivity within families8 and the increased anomaly risk associated with pregestational diabetes.9 Impact can be gauged by using a five-item predict-then-explain probe, adding one NBME-style item keyed to SHH and prechordal-plate timing, or including a mini-OSCE that asks learners to interpret a single fetal MRI slice and justify the facial findings mechanistically. Because these checks are brief and scalable, they fit easily within integrated curricula. While HPE is a classic example of dysmorphology, contemporary discoveries can also be used to encourage mechanistic and holistic thinking. For example, many genetic variants that shape the brain are also involved in shaping the face,10 indicating a quasi-shared genetic program. This may have direct clinical implications for disorders affecting both of these sets of tissues. Further, facial shape is often distinct in people with syndromic disease to the point of being a potential diagnostic tool,11 and this can be used to demonstrate anatomical relationships across the entire body. Finally, variation is often a hallmark of disease phenotypes, and understanding mechanisms that contribute to variation may shed light on potential therapeutic options.12, 13 More importantly, discussions of how dysmorphology may arise can contribute to a deeper understanding of embryology of the brain and face. The result of these approaches is not more embryology but better-aligned anatomy that brings current developmental science into day-to-day teaching, gives learners a predictive model they can apply at the bedside, and cultivates curiosity that fuels future discovery. Jennifer M. McBride: Conceptualization; writing – original draft; writing – review and editing. Ralph Marcucio: Conceptualization; writing – original draft; writing – review and editing. Jennifer M. McBride, PhD, FAAA, is a Professor in the Section of Anatomy in the Department of Surgery at UT Southwestern Medical Center (Dallas, TX). She serves as Director of Advanced Anatomy, co-Director of Human Anatomy in the School of Health Professions, and as faculty in the Human Structure course for medical students. Her scholarly interests include the effectiveness of educational technologies for knowledge acquisition, clinically oriented anatomical studies, and long-term knowledge retention.. Ralph Marcucio, PhD, FAAA, is a Professor in the Orthopaedic Trauma Institute in the Department of Orthopaedic Surgery at the University of California, San Francisco. His research focuses on morphogenesis of the craniofacial complex as well as how variation in disease and evolution arise. In addition, He has a translational research program that focuses on regeneration of the skeleton after traumatic injury.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.