Challenges and opportunities for artificial intelligence in oncological imaging

Imaging plays a key role in oncology, including the diagnosis and detection of cancer, determining clinical management, assessing treatment response, and complications of treatment or disease. The current use of clinical oncology is predominantly qualitative in nature with some relatively crude size-based measurements of tumours for assessment of disease progression or treatment response; however, it is increasingly understood that there may be significantly more information about oncological disease that can be obtained from imaging that is not currently utilized. Artificial intelligence (AI) has the potential to harness quantitative techniques to improve oncological imaging. These may include improving the efficiency or accuracy of traditional roles of imaging such as diagnosis or detection. These may also include new roles for imaging such as risk-stratifying patients for different types of therapy or determining biological tumour subtypes. This review article outlines several major areas in oncological imaging where there may be opportunities for AI technology. These include (1) screening and detection of cancer, (2) diagnosis and risk stratification, (3) tumour segmentation, (4) precision oncology, and (5) predicting prognosis and assessing treatment response. This review will also address some of the potential barriers to AI research in oncological imaging.

Cortical responses before 6 months of life associate with later autism

Autism spectrum disorder (ASD) is a common, highly heritable, developmental disorder and later-born siblings of diagnosed children are at higher risk of developing ASD than the general population. Although the emergence of behavioural symptoms of ASD in toddlerhood is well characterized, far less is known about development during the first months of life of infants at familial risk. In a prospective longitudinal study of infants at familial risk followed to 36 months, we measured functional near-infrared spectroscopy (fNIRS) brain responses to social videos of people (i.e. peek-a-boo) compared to non-social images (vehicles) and human vocalizations compared to non-vocal sounds. At 4-6 months, infants who went on to develop ASD at 3 years (N = 5) evidenced-reduced activation to visual social stimuli relative to low-risk infants (N = 16) across inferior frontal (IFG) and posterior temporal (pSTS-TPJ) regions of the cortex. Furthermore, these infants also showed reduced activation to vocal sounds and enhanced activation to non-vocal sounds within left lateralized temporal (aMTG-STG/pSTS-TPJ) regions compared with low-risk infants and high-risk infants who did not develop ASD (N = 15). The degree of activation to both the visual and auditory stimuli correlated with parent-reported ASD symptomology in toddlerhood. These preliminary findings are consistent with later atypical social brain responses seen in children and adults with ASD, and highlight the need for further work interrogating atypical processing in early infancy and how it may relate to later social interaction and communication difficulties characteristic of ASD.