Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2012 Apr 10;2(4):e100.
doi: 10.1038/tp.2012.10.

Use of Machine Learning to Shorten Observation-Based Screening and Diagnosis of Autism

Free PMC article

Use of Machine Learning to Shorten Observation-Based Screening and Diagnosis of Autism

D P Wall et al. Transl Psychiatry. .
Free PMC article


The Autism Diagnostic Observation Schedule-Generic (ADOS) is one of the most widely used instruments for behavioral evaluation of autism spectrum disorders. It is composed of four modules, each tailored for a specific group of individuals based on their language and developmental level. On average, a module takes between 30 and 60 min to deliver. We used a series of machine-learning algorithms to study the complete set of scores from Module 1 of the ADOS available at the Autism Genetic Resource Exchange (AGRE) for 612 individuals with a classification of autism and 15 non-spectrum individuals from both AGRE and the Boston Autism Consortium (AC). Our analysis indicated that 8 of the 29 items contained in Module 1 of the ADOS were sufficient to classify autism with 100% accuracy. We further validated the accuracy of this eight-item classifier against complete sets of scores from two independent sources, a collection of 110 individuals with autism from AC and a collection of 336 individuals with autism from the Simons Foundation. In both cases, our classifier performed with nearly 100% sensitivity, correctly classifying all but two of the individuals from these two resources with a diagnosis of autism, and with 94% specificity on a collection of observed and simulated non-spectrum controls. The classifier contained several elements found in the ADOS algorithm, demonstrating high test validity, and also resulted in a quantitative score that measures classification confidence and extremeness of the phenotype. With incidence rates rising, the ability to classify autism effectively and quickly requires careful design of assessment and diagnostic tools. Given the brevity, accuracy and quantitative nature of the classifier, results from this study may prove valuable in the development of mobile tools for preliminary evaluation and clinical prioritization-in particular those focused on assessment of short home videos of children--that speed the pace of initial evaluation and broaden the reach to a significantly larger percentage of the population at risk.


Figure 1
Figure 1
Receiver operating characteristic curves mapping sensitivity versus specificity for the 16 different machine-learning algorithms tested on the ADOS Module 1 training data. We identified the best classifiers as those closest to the point (1, 0) on the graph indicating perfect sensitivity (true positive rate) and one specificity (false positive rate). The best performing models were the ADTree and functional tree (FT). The ADTree was chosen over the FT because it used less items. See Table 2 for a summary of the 16 machine-learning algorithms used in our analysis.
Figure 2
Figure 2
Diagrammatic representation of the classifier generated by the ADTree algorithm. The ADTree was found to perform best out of the 16 different machine-learning approaches (Figure 1, Table 2). The resulting tree enables one to follow each path originating from the top node and increment (+) or decrement (−) prediction variables accordingly. In our case, variables with a negative sign yielded the classification of autism, whereas those with a positive sign resulted in a classification of non-spectrum. The magnitude of the score corresponded to confidence in the class prediction.
Figure 3
Figure 3
The ADTree scores of individuals in the AGRE, Boston AC and SSC data sets plotted against age in years (range from 13 months to 49 years). A majority of the ADTree scores were large, indicating confidence in the class predictions, and uncorrelated with the ages of the individuals.

Similar articles

See all similar articles

Cited by 31 articles

See all "Cited by" articles


    1. Bailey A, Le Couteur A, Gottesman I, Bolton P, Simonoff E, Yuzda E, et al. Autism as a strongly genetic disorder: evidence from a British twin study. Psychol Med. 1995;25:63–77. - PubMed
    1. Lord C, Risi S, Lambrecht L, Cook EH, Jr, Leventhal BL, DiLavore PC, et al. The autism diagnostic observation schedule-generic: a standard measure of social and communication deficits associated with the spectrum of autism. J Autism Dev Disord. 2000;30:205–223. - PubMed
    1. Shattuck PT, Durkin M, Maenner M, Newschaffer C, Mandell DS, Wiggins L, et al. Timing of identification among children with an autism spectrum disorder: findings from a population-based surveillance study. J Am Acad Child Adolesc Psychiatry. 2009;48:474–483. - PMC - PubMed
    1. Wiggins LD, Baio J, Rice C. Examination of the time between first evaluation and first autism spectrum diagnosis in a population-based sample. J Dev Behav Pediatr. 2006;27 (2 Suppl:S79–S87. - PubMed
    1. Bernier R, Mao A, Yen J. Psychopathology, families, and culture: autism. Child Adolesc Psychiatr Clin N Am. 2010;19:855–867. - PubMed

MeSH terms