BACKGROUND Twenty-thirty percent of individuals with autism stand, and/or walk, and/or run on their tiptoes. To our knowledge, studies about quantitative assessment of Tip-toe behavior (TTB) are scarce relating to walking and running and absent relating to standing. In a previous cohort study, using a qualitative assessment we described three mutually exclusive clinical functional classes of TTB of increasing severity: TTB only during running (TTB1), TTB only during walking and running (TTB2) and TTB during standing, walking and running (TTB3). OBJECTIVES The aims of this cross-sectional cohort study are: 1) to quantify TTB during both a static and a dynamic test in our ASD sample; 2) to compare the intensity of TTB in the three TTB clinical functional classes and in the NO-TTB group. METHODS Our study sample included 45 ASD subjects (mean age: 13,15 years – 4,63 SD; 40 males) diagnosed according to the DSM V criteria and under observation at our Institute. The confirmation and the severity of autism was established through ADOS-2. A therapist assessed the presence/absence of TTB during standing, walking and running using direct observation and a structured interview of the main caregiver living with the child. According to this assessment, 25 ASD subjects resulted not TTB, 3 resulted in TTB class 1, 10 in TTB class 2 and 7 in TTB class 3. The intensity of TTB expression during static and dynamic tests was quantified as percentage of time spent on the tip toes and as the percentage of toe steps, respectively. Both tests were conducted using standardized video recordings reviewed independently by two expert therapists. RESULTS The overall ADOS calibrated severity score (CSS) of all the subjects was 7.56 (1.71 SD). The overall ADOS CCS was 6,92 (1,55 SD) in NO-TTB, 8,15 (1,73 SD) in TTB1+2 and 8,71 (1,38 SD) in TTB3 (p = 0.02). The TTB time percentage values of the NO-TTB group during the static quantitative test was 0.25% (0.37 SD), while the time percentage values of the TTB1+2 was 1,82% (2,82 SD) vs 32,34% (31,82 SD) in TTB3 (table 1). We found a significant difference between each of the 3 groups (p < 0,02). In the NO-TTB group, during the dynamic quantitative test, the mean percentage of the TTB steps was 0.66% (1.48 SD), while it was 7,91% (5,71 SD) in TTB1+2 vs 60,93% (28,29 SD) in TTB3 (table 1). We found a significant difference between each of the 3 groups (p= 0.000) (fig. 1). Moreover, we found a significant correlation (r = 0.702) between the quantity of TTB in the static and the dynamic test. Finally, we also found a significant correlation between the severity of TTB during both the static and the dynamic tests and the ADOS-2 CSS (r = 0.305 and r = 0.406 respectively). CONCLUSIONS We quantified TTB using a new structured static and dynamic assessment test in our ASD sample. We found significant difference between NO-TTB and the 2 TTB subgroups. The TTB quantity in the static test is correlated with TTB quantity in the dynamic test.

BACKGROUND Evaluating treatments outcome in children with low-functioning autistic disorder requires the utilization of specific but manageable instruments, both for patients and for their environment (parents, educators, doctors). Few studies so far have focused how use multidimensional data for outcome assessment in residential health settings. OBJECTIVE The aim of this study is to highlight the possible outcome prediction of personalized plans of intervention for low-functioning ASD subjects using innovative machine learning systems enabling also to understand which treatment factors are significantly involved. METHODS In this pilot observational study, twelve consecutive new patients with low-functioning autism (range of age 3-13 years) have been enrolled between November 2015 and October 2016. Four complementary assessment instruments (Vineland Adaptive Behavior Scales, a 540-item questionnaire which evaluate personal and social autonomy, communication and motoric competences; SDQ-Strenghts and Difficulties Questionnaire-, a 25-item questionnaire useful to screen emotional, behavioral and social problems in children aged 4-16 years; The HoNOSCA-Health of Nation Outcome Scale for Children and Adolescents-, a 15-item clinical assessment scale used as part of the routine outcome monitoring in mental health services, which measures global functioning in patients aged 3-18 years through 4 different areas: behavioral, impairment, symptoms, social functioning; DC-GAS -Disability Child Global Assessment Scale-, a dimensional scale used by the clinician to evaluate global functioning in disabled children and adolescents, have been used at the patients first access in the neuropsychiatric clinic and after 6 months of intensive personalized treatment. Vineland Scales and SDQ questionnaire have been completed by educators and parents; HoNOSCA and DC-GAS have been completed by the clinician. The most important outcome measure has been individuated in the DC-GAS total score, in particular considering differences between baseline evaluation (T1) and the assessment after 6 months (T2). Ninety four variables related to demographic, familiar, therapeutic, pharmacological, medical and checkup information have been used to represent the input for preprocessing. An evolutionary algorithm (a TWIST system based on KNN algorithm) has been used to subdivide the dataset into training and testing set and select features yielding the maximal amount of information. After this pre-processing, 21 input variables were selected and different machine learning systems have been used to develop a predictive model based on a training testing crossover procedure. RESULTS Eight out of twelve subjects have shown an improved global functioning at the end of the follow-up. The best machine learning system (three-layers feed- forward neural network with 8 hidden nodes) obtained a global accuracy of 83.3% ( 91.7 % sensitivity and 75% specificity ) with a ROC of 0.89. The variables selected for the predictive model included previous pharmacological and non-pharmacological treatments, actual treatment plan, and baseline scores of different subscales of Vineland, SDQ, HoNOSCA and DC-GAS. CONCLUSION Machine learning systems shows a promising potential in predicting the outcome of personalized multidimensional interventions in low-functioning ASD subjects. Accurate data collection, considering multidimensional aspects, and the use of a complex and complete statistical analysis, as the machine learning systems, could be useful in order to highlight predictable positive treatment factors.

BACKGROUND According to the literature, about one fourth of individuals with ASD present with toe walking. In a previous study, we found that this behavior is present not only during walking but also while standing and running, and described three mutually exclusive clinical functional classes with a different degree of severity: TTB during running (TTB1), TTB during walking and running (TTB2) and TTB during standing, walking and running (TTB3). In another study we also found a positive relationship between the presence and severity of TTB and the Achilles’s tendon shortening in ASD subjects. In this perspective, assessing and monitoring TTB in ASD subjects could become critical to identify ASD subjects at risk of developing muscle shortening. Moreover, to our knowledge in the literature systematic observations of the natural history of Tip-toe Behavior (TTB) in ASD children using standardized assessment are lacking. OBJECTIVES The aim of this prospective cohort study is to describe the natural history of TTB and NO-TTB ASD subjects at short-medium term. METHODS The prospective study included 72 consecutive subjects (62 males; mean age: 15,68 years – SD 3,85) present in our Institute. The inclusion criteria were: an ASD diagnosis according to the DSM V criteria, a diagnosis confirmation based on the ADOS–2. The exclusion criteria were: presence of co-morbid diagnoses that would have an impact on gait. The assessment of presence/absence of TTB during standing, walking and running was done through direct observation and a structured interview of the main caregiver living with the child. We repeated the same evaluation 12 to 44 During months (mean: 29,88 months – SD 8,86; median: 34,5 months) after the first observation. RESULTS During the first TTB assessment 49 subjects resulted Non-TTB (68,1%) and 23 resulted TTB. In TTB group, 4 subjects were TTB1, 9 subjects were TTB2 and 10 subjects were TTB3. At follow up 7 out of 23 TTB subjects (30%) changed the severity class, 5 decreasing and 2 increasing severity (table 1). No one shifted to NO-TTB group. In NO-TTB group 8 subjects (16%) shifted to TTB group (5 in TTB 1 class, 2 in TTB 2 class and 1 in TTB class 3. The ADOS calibrated severity score of the converters was not different from non-converters (7,5 – 1,4 SD versus 7,56 – 1,74 SD). CONCLUSION TTB behavior can change over time in ASD subjects. In an average time of two years about one third of TTB subjects change their severity class and 16% of Non-TTB subjects become TTB. This finding underlies the importance of close monitoring of TTB with standardized protocols.

BACKGROUND In addition to the core symptoms, both fine and gross motor delays/disorders have been reported in children with Autism Spectrum Disorders (ASD). However, it is still unclear whether motor impairments are uniformly distributed across the entire ASD spectrum and whether they are related to DSM-5 specifiers (i.e. intelligence, language, comorbidity and associated conditions). In the light of the high heterogeneity in ASD it could be possible that “a single symptom approach analysis” do not provide comprehensive information. The strong inherent non-linearity of the relationships between clinical variables may account for the inability to grasp the core problem by the traditional analysis. Artificial neural networks (ANNs) are computational adaptive systems particularly adapting to solving non-linear problems. The goal of this data mining model is to discover hidden trends and associations among variables. Recently this approach has been successfully applied to the ASD field. OBJECTIVES To investigate associations between motor skills and clinical/developmental features in preschoolers with ASD. We hypothesized that ANNs will be able to find hidden trends among the variables revealing the clinical profiles related to motor functioning in ASD. METHODS This study was carried out according to the standards for good ethical practice of the IRCCS Stella Maris Foundation and in accordance with the guidelines of the Declaration of Helsinki. 32 male with ASD (age range: 30-60 months; nonverbal IQ ≥ 70) were recruited at the IRCCS Stella Maris Foundation, a tertiary care university hospital. Multidisciplinary comprehensive diagnostic evaluation was associated with a standardized assessment battery for motor skills, the Peabody Developmental Motor Scale- Second Edition (PDMS-2). The PDMS-2 consists of six motor subscales (Reflexes, Stationary, Locomotion, Object Manipulation, Grasping and Visual-Motor Integration) and three motor quotients (MQ) (Gross MQ, Fine MQ, Total MQ). According to PDMS-2, motor skills were classified into 7 categories: very superior, superior, above average, average, below average, poor and very poor. Analyses were performed through the Auto Contractive Map system which is a fourth-generation unsupervised ANNs. Auto Contractive Map system ‘spatializes’ the correlation among variables (‘closeness’) providing a graph that identifies the relevant associations and organizes them into a coherent picture. RESULTS Preliminary linear correlation analysis revealed that motor impairment was associated with both cognitive skills and repetitive behaviors in children with ASD (Table 1). The ANNs analysis (Figure 1) shows that motor disorders appeared to be strongly related to low level of expressive language and high level of repetitive behaviors in preschoolers with ASD. In addition, the ANNs approach consider the entire spectrum of relationship among clinical variables, revealing hidden trends among motor, cognitive and social skills. CONCLUSIONS The ANNs approach revealed motor skills as moderators of ASD core symptoms. This appears to be consistent with the growing literature suggesting that the systematic observation of motor development in ASD may improve the knowledge about clinical and neurobiological involvement as well as guide development of treatments. ACKNOWLEDGMENTS This work has been partially supported by the European Community's Horizon 2020 Program under grant agreement n. 642996 (BRAINVIEW) and by the Italian Minister of Health Network Project 'Italian Autism Spectrum Disorders Network: filling the gaps in the National Health System Care' (NET 2013-02355263)

BACKGROUND In a previous study the authors have shown the ability of a novel kind of Machine Learning System(MLS) named MS-ROM/I-FAST developed by The Semeion Research Institute in Rome to extract interesting features in computerized EEG that allow an almost perfectly distinction of ASD children from typically developing ones. The proof of concept study, published in 2017 in Computer Method and Programs in Biomedicine showed accuracy values near to 100% using training-testing protocol and to 84% - 92.8% using Leave One Out protocol. The similarities among the MLS weight matrixes measured with apposite algorithms were not affected by the age of the subjects suggesting that the MLS do not read age-related EEG patterns, but rather invariant features related to the brain’s underlying abnormalities. AIM OF THE STUDY The aim of the study is to assess how effectively this methodology distinguishes ADS subjects from children affected with other neuro-psychiatric disorders . METHODS Twenty definite ASD subjects and twenty subjects with neuropsychiatric disorders matched for age and gender distribution observed at Tarnow Center for Self-Management, Huston (US) were included in the study. The two groups had the same age range ( 4-14 yrs) and male/female ratio (14/6). ASD patients received independent Autism diagnoses according to DSM-V criteria, subsequently confirmed by a qualified psychiatrist using the ADOS scale. No autistic child was affected by genetic conditions and/or cerebral malformations documented by neuroimaging and epilepsy. In the comparison group the range of primary diagnoses was the following: Attention-Deficit Disorder ( N= 13), Disorder of social functioning( N=3), Anxiety disorders( N= 2), Major depressive disorder(N= 1), Specific developmental disorders of scholastic skills(N=1) A continuous segment of artefact-free EEG data lasting 10 minutes in ASCCI format was used to compute multi-scale entropy values and for subsequent analyses. A Multi-scale ranked organizing map (MS-ROM), based on the self-organizing map (SOM) neural network, coupled with the TWIST system (an evolutionary system able to select predictive features) created an invariant features vector input of EEG on which supervised machine learning systems acted as blind classifiers. RESULTS After MS-ROM/I-FAST preprocessing, twelve features were extracted representing the EEG signature. Acting on these features the overall predictive capability of different machine learning systems in deciphering autistic cases from other NP disorders ranged between 93% and 97.5% (Table 1). These results were obtained at different times in separate experiments performed on the same training and testing subsets. The similarities among the ANN weight matrixes measured with apposite algorithms were not affected by the age of the subjects. CONCLUSION This study demonstrates the value of EEG processing with advanced MLS in the differential diagnosis between ASD and other NP disorders confirming therefore the existence of a specific EEG signature in ASD.

BACKGROUND Stereotypies, despite their high frequency and strong diagnostic significance within autism, have not yet been fully elucidated due to their broad spectrum of presentation and pattern complexity. Standardized video-recordings can help to depict the complex pattern of stereotypies commonly observed in autism, thus allowing for a better definition of major phenotypes. AIM The aim of this study is to analyze stereotypies patterns observed in a sample of children and adolescents residing at our Institute and subsequently classify them by means of video-recordings. METHODS 20 expert caregiver wearing a body cam recorded specific stereotypic behavior in a natural context during the everyday activities of 67 autistic subjects for 3 months of close follow-up. After a few minutes of recording, the possibility to interrupt their behavior by intervening physically to divert attention was registered. A team consisting of a senior child neuro-psychiatrist and a senior psychologist reviewed all the video recordings (1868) selecting 780 of them as the most meaningful to summarize the whole spectrum in each individual in the given time window. Each video was classified according to components (motor, sensorial, vocal, intellective), complexity (2 classes, simple and complex), body parts involved (n=18) and sensory channels involved (hearing, sight, proprioception, taste, pain, smell). RESULTS The vast majority ( 87%) of the780 patterns occurred several times generally (73%) in a state of tranquility. In more than half of cases (53.4%), the interruption required intervening physically, but a poor correlation was found between the pattern type and possibility to interrupt the behavior by way of verbal or physical intervention ( r= -0.08/+0.08). The individual stereotypies spectrum ranged from 1 to 33 different patterns (average= 11.6; S.D= 6.82). The most frequent pattern was represented by the combination of simple motor and sensorial components ( accounting for 23% of the total number) followed by simple motor and simple sensorial (9% and 8% respectively). The other 47 patterns with combinations from 1 to 4 components accounted for the remaining 60%. In the 569 patterns containing motor components, whole body and arms movements constituted the most frequent body parts involved (41% e 38% respectively) followed by mouth and hands (10% and 9.8% respectively). In the 531 patterns containing sensorial components, the most frequent channel involved was tactile (50%) followed by proprioceptive( 34%) and acoustic (19.5%). Most of the 127 stereotypies with vocal components were constituted by simple vocalizations, 85.8% and only 14.2% by phonemes or words. CONCLUSIONS This study represents a first attempt to systematically document the patterns of expression of stereotyped behavior in a cohort of autistic subjects closely followed by professional educators. Open access to this video bank and to the clinical data will be allowed to interested researchers , with the aim of improving the comprehension of this complex phenomenon and its correlation with clinical and demographic features.

BACKGROUND Kindergarten is a privileged opportunity to observe children’s behavior in a natural context. This possibility allows for detecting signs of derangement in motor and neuropsychological development not otherwise previously observed. Our group has developed a standardized protocol with a checklist containing over 284 items for carefully detecting delay in the acquisition of a given capacity according to age related international developmental nomograms. OBJECTIVES The aim of this study is to assess the feasibility and effectiveness in the screening of ASD by simply observing child behavior while at Kindergarten. METHODS The protocol included two different tests: - The Denver Developmental Screening Test (Denver test)is a 41 item test for screening development of infants and preschool-aged children; items cover four general functions: personal social (e.g. smiling), fine motor adaptive (e.g. grasping and drawing), language (e.g. combining words), and gross motor (e.g. walking). Test age range 0-6. - The Adaptive Behavior Assessment System – Second edition (ABAS – II) containing 216 items is a global and normative assessment scale of behavior that measures daily life skills. The project protocol provides for teacher/caregivers questionnaires regarding subjects aged 2-5. It investigates 10 adaptive areas: Communication, Preschool/School skills, self-Control, Playing/leisure, Socializing, Self-Care, Home Care/School, Environmental Use, Health and Safety, Work. Expert psychomotor and education therapists with the supervision of a senior neuro-pediatrician (MN), observed and interacted with 62 children, aged 36 - 65 months, attending the “Istituto Comprensivo Don Milani” in Tavernerio (Como, Italy) during two sessions and after obtaining written informed consent. Seven out of 62 children were found to be affected by neuro-psychiatric disorders (ASD; Down syndrome; delay of psychomotor development; Cerebral Palsy; developmental disorder; hyperactivity) Our staff was pleasantly accepted in the classroom and was able to collect all the information required. The school teachers learned basic skills to heighten their observation capabilities in child behavior as a result of interacting with our professional staff. RESULTS In the 55 children without previous diagnoses of neuro-psychiatric disorders, the application of the Denver test and ABAS scales pointed out the presence of 7 and 12 cases respectively with at least one (range 1 - 6) item not corresponding to chronological age. The integration of this information with the experience and skills of our staff resulted in the recognition of possible undetected developmental disorders and the subsequent invitation to parents to bring the children in question to a neurodevelopmental diagnostic unit. In one of these cases, an Autism Spectrum Disorder (ASD) diagnosis was made. In one of two other cases, parents did not continue with the diagnosis, in the other, a pediatric psychiatrist did not confirm the presence of language developmental delay (table 1). CONCLUSIONS This experience shows that structured external observation in a kindergarten is a feasible and promising approach for the screening and the early detection of neurodevelopmental disorders. ABAS II seems to provide increased sensitivity in detecting suspicious cases.