Insight into ADHD diagnosis with deep learning on Actimetry: Quantitative interpretation of occlusion maps in age and gender subgroups.

Attention Deficit/Hyperactivity Disorder (ADHD) is a prevalent neurodevelopmental disorder in childhood that often persists into adulthood. Objectively diagnosing ADHD can be challenging due to the reliance on subjective questionnaires in clinical assessment. Fortunately, recent advancements in artificial intelligence (AI) have shown promise in providing objective diagnoses through the analysis of medical images or activity recordings. These AI-based techniques have demonstrated accurate ADHD diagnosis; however, the growing complexity of deep learning models has introduced a lack of interpretability. These models often function as black boxes, unable to offer meaningful insights into the data patterns that characterize ADHD. OBJECTIVE: This paper proposes a methodology to interpret the output of an AI-based diagnosis system for combined ADHD in age and gender-stratified populations. METHODS: Our system is based on the analysis of 24 hour-long activity records using Convolutional Neural Networks (CNNs) to classify spectrograms of activity windows. These windows are interpreted using occlusion maps to highlight the time-frequency patterns explaining ADHD activity. RESULTS: Significant differences in the frequency patterns between ADHD and controls both in diurnal and nocturnal activity were found for all the populations. Temporal dispersion also presented differences in the male population. CONCLUSION: The proposed interpretation techniques for CNNs highlighted gender- and age-related differences between ADHD patients and controls. Leveraging these differences could potentially lead to improved diagnostic accuracy, especially if a larger and more balanced dataset is utilized. SIGNIFICANCE: Our findings pave the way for the development of an AI-based diagnosis system for ADHD that offers interpretability, thereby providing valuable insights into the underlying etiology of the disease.

[Trend in medicines use for attention deficit hyperactivity disorder in children and adolescents (2010-2019).] / Tendencia del consumo de fárma-cos en el trastorno por déficit de atención e hiperactividad en niños y adolescentes (2010-2019).

OBJECTIVE: Attention deficit hyperactivity disorder (ADHD) is a common childhood neurodevelopmental disorder characterised primarily by three core symptoms: inattention, hyperactivity and impulsivity. It is one of the most commonly diagnosed childhood psychiatric disorders, with a worldwide prevalence of between 3% and 5%, and between 6% and 7% in the Spanish population. The aim of the study is to analyse the trend in the consumption of drugs used for the treatment of ADHD between 2010-2019 in Castilla y León. METHODS: Epidemiological registry study of all dispensing in pharmacies in Castilla y León between 2010 and 2019 to patients under 19 years of age, of active substance N06BA04 (methylphenidate), N06BA09 (atomoxetine), N06BA12 (lisdexamfetamine), N06BA07 (modafinil) and C02AC02 (guanfacine). Data on drug use were obtained from the information system for the pharmaceutical provision of Castilla y León, CONCYLIA. Frequencies in absolute values and the corresponding percentages were calculated. Student's t-test was used to estimate differences between continuous variables and Pearson's Chi-square test for categorical variables, while the trend in consumption was analysed using the Cochran-Armitage test. RESULTS: ADHD medication was dispensed annually to 1.77% of the population, with consumption being more than three times higher in boys than in girls (2.69% vs 0.81%; p=0.001). The age group with the highest peak use was 10-14 years with 3.42%. Methylphenidate was the drug used by the highest percentage of the population (2.44%) followed by lisdexamfetamine (0.37%). CONCLUSIONS: Approximately 2 out of every 100 people aged 0-19 years were treated with some ADHD medication, mainly methylphenidate, in Castilla y León between 2010 and 2019.

OBJETIVO: El trastorno por déficit de atención e hiperactividad (TDAH) es un trastorno común del neurodesarrollo infantil que se caracteriza fundamentalmente por tres síntomas a nivel central: falta de atención, hiperactividad e impulsividad. Se trata de uno de los trastornos psiquiátricos infantiles más comúnmente diagnosticados, con una prevalencia a nivel mundial que oscila entre el 3% y el 5%, y entre el 6% y el 7% en la población española. El objetivo del estudio es analizar la tendencia del consumo de los fármacos utilizados para el tratamiento del TDAH entre los años 2010-2019 en Castilla y León. METODOS: Estudio epidemiológico de registro de todas las dispensaciones realizadas en las oficinas de farmacia de Castilla y León entre el año 2010 y 2019 a pacientes menores de 19 años, de los principios activos N06BA04 (metilfenidato), N06BA09 (atomoxetina), N06BA12 (lisdexanfetamina), N06BA07 (modafinilo) y C02AC02 (guanfacina). Los datos de utilización de medicamentos se obtuvieron a partir del sistema de información para la prestación farmacéutica de Castilla y León, CONCYLIA. Se calcularon frecuencias en valores absolutos y los porcentajes correspondientes. Se utilizó el test-t de Student para estimar diferencias entre variables continuas y el test Chi-cuadrado de Pearson para las categóricas, mientras que la tendencia de consumo se analizó mediante el test de Cochran-Armitage. RESULTADOS: Se dispensaron anualmente los medicamentos para el TDAH al 1,77% de la población, siendo el consumo más de tres veces mayor en niños que en niñas (2,69% vs 0,81%; p=0,001). La franja de edad donde se observó el mayor pico de consumo fue de los 10 a 14 años con un 3,42%. El metilfenidato fue el medicamento consumido por un mayor porcentaje de la población (2,44%) seguido de la lisdexanfetamina (0,37%). CONCLUSIONES: Aproximadamente 2 de cada 100 personas entre 0 y 19 años fueron tratadas con algún medicamento para la TDAH, fundamentalmente metilfenidato, en Castilla y León entre 2010 y 2019.

Trends in psychiatric hospitalization for alcohol and drugs in Castilla y León between 2005 and 2015. / Tendencias en la hospitalización psiquiátrica por alcohol y drogas en Castilla y León entre 2005 y 2015.

It has been estimated that alcohol, tobacco, and illicit drugs were responsible for more than 10 million deaths worldwide in 2016, and there are many opportunities for improvement. Regarding innovative data analysis, advances have been made in the extraction of information from administrative databases for analytics purposes. We studied trends in hospitalization rates for alcohol and drug abuse over eleven years with Joinpoint Trend Analysis software. This is a descriptive study of cross-associations in 3,758 hospital admissions of patients admitted with a main diagnosis of alcohol and drug abuse or dependence in psychiatry units of public health centres of Castilla y León (Spain) between 2005 and 2015. Hospitalization trends for alcohol and drug related conditions declined over the eleven-year period. Separately, there was a statistically significant decrease in alcohol and cocaine related conditions, but a strong upward trend in cannabis related conditions between 2013 and 2015. Alcohol was the main cause of admission to psychiatric units with a diagnosis of addiction. In the 11 years researched, there was a progressive and constant reduction in admissions for substance use except for cannabis. The innovative statistical methodology has already proven to be useful for identifying trends and changes in different pathologies over time.

A nivel mundial, se ha estimado que el alcohol, el tabaco y las drogas han sido responsables de más de 10 millones de muertes en 2016, y que existe mucho margen para reducir la mortalidad. Se han realizado avances en la extracción de información de bases de datos administrativas con el fin de analizar grandes volúmenes de datos sanitarios. Hemos estudiado las tendencias en las tasas de hospitalización con diagnóstico de adicción a alcohol y drogas durante once años con el software Joinpoint Trend Analysis. Se trata de un estudio descriptivo de asociación cruzada de 3.758 ingresos hospitalarios de pacientes con diagnóstico principal de abuso o dependencia de alcohol y drogas en unidades de Psiquiatría de centros públicos de Castilla y León entre 2005 y 2015. Las tendencias en la hospitalización por adicción al alcohol y/o drogas disminuyeron a lo largo de los once años. Además de una reducción estadísticamente significativa de los ingresos por alcohol y cocaína, se apreció una fuerte tendencia al alza en los ingresos por cannabis entre 2013 y 2015. El alcohol fue durante todo el periodo de estudio la principal causa de ingreso y el que más días de hospitalización ha generado. No obstante, en los 11 años se observó una reducción progresiva y constante en los ingresos por todas las sustancias a excepción del cannabis. La metodología utilizada ya ha demostrado ser muy útil para identificar cambios de tendencias en diferentes patologías.

Construct Validity of the Routine Assessment of Patient Index Data 3 (RAPID3) in the Evaluation of Axial Spondyloarthritis.

OBJECTIVE: Although there are different tools to evaluate axial spondyloarthritis (axSpA), they are hardly used in routine clinical practice due to time constraints. The Routine Assessment of Patient Index Data 3 (RAPID3) is a composite measure feasible for use as a sole metric in busy clinics. We aimed to test its measurement properties in patients with axial SpA in a real-world clinical setting. METHODS: This cross-sectional study included 131 consecutive patients with axial SpA. The convergent (Spearman ρ) and discriminant (receiver-operating characteristic [ROC] curve analysis) validity of RAPID3 were tested against several axSpA-specific measures (Bath Ankylosing Spondylitis Disease Activity Index [BASDAI], Ankylosing Spondylitis Disease Activity Score [ASDAS], Bath Ankylosing Spondylitis Functional Index [BASFI], and modified Stoke Ankylosing Spondylitis Spinal Score [mSASSS]). A multivariate model was built to detect disease factors associated with RAPID3 remission (values ≤ 3). RESULTS: The study included 82 men and 49 women, with a median age of 55 (IQR 46-61) years, and a median disease duration of 11 (IQR 6-24) years. Mean RAPID3 was 9.45 ± 6.7. The BASDAI showed moderate correlation with ASDAS (ρ 0.66, P < 0.0001), but higher correlations with BASFI (ρ 0.78, P < 0.0001) and RAPID3 (ρ 0.75, P < 0.0001). The ASDAS had moderate correlations with BASFI, BASDAI, and RAPID3 (ranges 0.66-0.68, P < 0.0001). Higher correlations were found between BASFI and BASDAI (ρ 0.78, P < 0.0001), and BASFI and RAPID3 (ρ 0.73, P < 0.0001). The mSASSS did not show any correlation with any of the above composite measures. κ agreement between RAPID3 remission and other SpA remission criteria was moderate (κ 0.46-0.56). The RAPID3 thresholds to define remission ranged from values ≤ 2 to ≤ 6 with areas under the ROC curve between 0.86-0.91. Female sex (OR 0.34, 95% CI 0.12-0.90, P = 0.03) and nonsteroidal antiinflammatory drug intake (OR 0.26, 95% CI 0.10-0.66, P = 0.005) were independently associated with lower odds of achieving RAPID3 remission. CONCLUSION: RAPID3 demonstrated construct validity in this cross-sectional study. This index can be useful for a more comprehensive assessment of axSpA in busy clinical settings.

[Impact of COVID-19 on mortality in the autonomous community of Castilla y León (Spain)]. / Impacto de la COVID-19 en la mortalidad de la comunidad autónoma de Castilla y León.

OBJECTIVE: To estimate the increase in mortality associated with the SARS-CoV-2 coronavirus pandemic in the autonomous community of Castilla y León (Spain). METHOD: Ecological study based on population and death data for the months of March 2016 to 2020 in Castilla y León. The general and provincial standardized rates, the relative risks of the year 2020 with respect to previous years and the risks adjusted by sex, periods and province, using Poisson regression, were calculated. Trend analysis was performed using joinpoint linear regression. RESULTS: An increase in mortality was observed in March 2020 with respect to previous years, with an increase of 39% for men (relative risk [RR]: 1.39; 95% confidence interval [95%CI]: 1.32-1.47) and 28% for women (RR: 1.28; 95%CI: 1.21-1.35). The model predicts excess mortality of 775 deaths. In the trend analysis there is a significant turning point in 2019 in men, globally and for almost all provinces. The increase in mortality is general, although heterogeneous by sex, age group and province. CONCLUSIONS: Although the observed increase in mortality cannot be totally attributed to the disease, it is the best estimate we have of the real impact on deaths directly or indirectly related to it. The number of declared deaths only reaches two thirds of the increase in mortality observed.

Objective ADHD Diagnosis Using Convolutional Neural Networks Over Daily-Life Activity Records.

Attention Deficit/Hyperactivity Disorder (ADHD) is the most common neurobehavioral disorder in children and adolescents. However, its etiology is still unknown, and this hinders the existence of reliable, fast and inexpensive standard diagnostic methods. OBJECTIVE: This paper proposes an end-to-end methodology for automatic diagnosis of the combined type of ADHD. METHODS: Diagnosis is based on the analysis of 24 hour-long activity records using Convolutional Neural Networks to classify spectrograms of activity windows. RESULTS: We achieve up to [Formula: see text] average sensitivity, [Formula: see text] specificity and AUC values over [Formula: see text]. Overall, our figures overcome those obtained by actigraphy-based methods reported in the literature as well as others based on more expensive (and not so convenient) acquisition methods. CONCLUSION: These results reinforce the idea that combining deep learning techniques together with actimetry can lead to a robust and efficient system for objective ADHD diagnosis. SIGNIFICANCE: Reliance on simple activity measurements leads to an inexpensive and non-invasive objective diagn-ostic method, which can be easily implemented with daily devices.

[Multiple birth trends in the region of Castilla y León (Spain) in a 13 year period]. / Evolución temporal de los partos múltiples en Castilla y León durante 13 años.

INTRODUCTION: Multiple pregnancy has increased in prevalence in the last few years, which could lead to more foetal and maternal morbidity issues. The aim of this study is to describe the trend of multiple pregnancy deliveries in Castilla y León during the last 13years and the subsequent impact on foetal and maternal health. MATERIAL AND METHODS: Data was collected from the hospital discharge reports registered in the Regional Health-care database (SACYL: Health care in Castilla y León) between 2001 and 2013. A cross sectional descriptive study was conducted, including trend analysis with log-linear joint point model, a rhythm metric study, as well as a risk assessment with multivariate analysis. RESULTS: A pronounced upward trend was observed in the proportion of multiple deliveries in this time period, compared to single ones, with an annual percentage change of 3.4% (95% CI: 2.5-4.4). Multiple pregnancy was significantly correlated with advanced maternal age, abnormal glucose tolerance, dystocia and caesarean section delivery, premature birth, foetal malposition, foetal macrosomia, stillbirth, in vitro fertilisation, and hypertensive episodes of pregnancy. In vitro fertilization showed a 9.3 fold increased risk in multiple pregnancy (95% CI: 7.4-11.5), with maternal age increasing the risk up to 5% per year of age (OR: 1.05: 95%CI: 1.04-1.05). No seasonal rhythm was observed in multiple deliveries compared with single ones. CONCLUSION: Multiple pregnancy has experienced a continuous increase, with no seasonal trend, and is associated with the increase in assisted reproductive technology and advanced maternal age. This involves more problems regarding foetal and maternal morbidity and mortality.

Prevalence of oppositional defiant disorder in Spain.

INTRODUCTION: Oppositional defiant disorder (ODD) is characterized by a pattern of negative, defiant, disobedient and hostile behavior toward authority figures. ODD is one of the most frequent reasons for clinical consultation on mental health during childhood and adolescence. ODD has a high morbidity and dysfunction, and has important implications for the future if not treated early. OBJECTIVE: To determine the prevalence of ODD in schoolchildren aged 6-16 years in Castile and Leon (Spain). MATERIAL AND METHODS: Population study with a stratified multistage sample, and a proportional cluster design. Sample analyzed: 1,049. Cases were defined according to DSM-IV criteria. RESULTS: An overall prevalence rate of 5.6% was found (95% CI: 4.2%-7%). Male gender prevalence=6.8%; female=4.3%. Prevalence in secondary education=6.2%; primary education=5.3%. No significant differences by gender, age, grade, type of school, or demographic area were found. ODD prevalence without considering functional impairment, such as is performed in some research, would increase the prevalence to 7.4%. ODD cases have significantly worse academic outcomes (overall academic performance, reading, maths and writing), and worse classroom behavior (relationship with peers, respect for rules, organizational skills, academic tasks, and disruption of the class). CONCLUSIONS: Castile and Leon has a prevalence rate of ODD slightly higher to that observed in international publications. Depending on the distribution by age, morbidity and clinical dysfunctional impact, an early diagnosis and a preventive intervention are required for health planning.

Automatic diagnosis of ADHD based on multichannel nonlinear analysis of actimetry registries.

Attention-Deficit Hyperactivity Disorder (ADHD) is the most common mental health problem in childhood and adolescence. It is commonly diagnosed by means of subjective methods which tend to overestimate the severity of the pathology. A number of objective methods also exist, but they are either expensive or time-consuming. Some recent proposals based on nonlinear processing of activity registries have deserved special attention. Since they rely on actigraphy measurements, they are both inexpensive and non-invasive. Among these methods, those shown to have higher reliability are based on single-channel complexity assessment of the activity patterns. This way, potentially useful information related to the interaction between the different channels is discarded. In this paper we propose a new methodology for ADHD diagnosis based on joint complexity assessment of multichannel activity registries. Results on real data show that the proposed method constitute a useful diagnostic aid tool reaching 87:10% sensitivity and 84.38% specificity. The combination of ADHD indicators extracted with the proposed method with single-channel complexity-based indices previously proposed lead to sensitivity and specifity values above 90%.

[Relation between myocardial infarction and circadian rhythm in patients attended in a prehospital emergency service]. / Relación entre infarto de miocardio y ritmo circadiano en pacientes atendidos por un servicio de emergencias prehospitalario.

BACKGROUND AND OBJECTIVES: The aim of this study is to analyze the presence of circadian rhythm in the time of onset of symptoms of acute myocardial infarction treated by a prehospital emergency system and the influence of modifiable cardiovascular risk factors and non-modifiable as modulators of that circadian rhythm. PATIENTS AND METHODS: Retrospective analysis of 709 patients clinically diagnosed with acute myocardial infarction on-site in the prehospital setting. The variables were time to onset of symptoms, age, sex, previous ischemic heart disease, hypertension, diabetes mellitus, hyperlipidemia and smoking. We analyzed the rhythm with cosinor multiple sinusoid method, with 3 harmonics (24, 12 and 8h) for the adjustment. RESULTS: The time of onset of pain showed circadian rhythm (P <,001), peaking at 10.39 and a valley at 4.28, showing a sinusoidal curve fitting bimodal aspect with a predominant morning peak and another evening one of lower amplitude. All subgroups categorized by the study variables showed circadian rhythm, with a cosine curve similar to the global infarction. Smokers had a predominantly evening peak. CONCLUSIONS: Acute myocardial infarction shows a circadian rhythm. Smoking and diabetes mellitus can modify the standard incidence rate of occurrence of myocardial infarction.

Characterization of activity epochs in actimetric registries for infantile colic diagnosis: Identification and feature extraction based on wavelets and symbolic dynamics.

The diagnosis and therapy planning of high prevalence pathologies such as infantile colic can be substantially improved by statistical signal processing of activity/rest registries. Assuming that colic episodes are associated to activity episodes, diagnosis aid systems should be based on preprocessing techniques able to separate real activity from rest epochs, and feature extraction methods to identify meaningful indices with diagnostic capabilities. In this paper, we propose a two step diagnosis aid methodology for infantile colic in children below 3 months old. Identification of activity periods is performed by means of a wavelet based activity filter which does not depend on the acquisition device (as so far proposed methods do). In addition, symbolic dynamic analysis is used for extraction of discriminative indices from the activity time series. Results on real data yielded 100% sensitivity and 80% specificity in a study group composed of 46 cases and 10 control subjects.

[Usefulness of the Stroop test in attention deficit hyperactivity disorder]. / Utilidad del test de Stroop en el trastorno por déficit de atencion/hiperactividad.

AIMS: The aim of this study is to analyse the differences in the Stroop effect between cases with attention deficit hyper-activity disorder (ADHD) and controls. It also seeks to find the best model based on the third task of the colours and words test (Stroop-CW) for predicting ADHD and to analyse the validity of the Stroop-CW test for diagnosing the disorder. SUBJECTS AND METHODS: The sample studied consisted of 100 cases of ADHD -according to Diagnostic and statistical manual of mental disorders (DSM-IV) criteria- and 100 controls, between 7 and 11 years of age, who were evaluated using the Stroop test. The controls were recruited at random and paired by age, sex and sociodemographic area with the cases. RESULTS: The cases present a mean cognitive style that is significantly less flexible (d = -1.06) and they also display a lower capacity to inhibit or control automatic responses than the controls at all ages (7 years: d = 1.67; 8 years: d = 1.02; 9 years: d = 1.32; 10 years: d = 2.04; 11 years: d = 0.89). The model of logistic regression analysis that best predicts ADHD is made up of age and Stroop-CW. The formulation derived from the model offers a sensitivity of 81% and a specificity of 72%, taking the criteria of the DSM-IV for ADHD as the reference test. CONCLUSIONS: The Stroop-CW test presents usefulness and complementary criteria validity for the diagnosis of ADHD.

Decline and loss of birth seasonality in Spain: analysis of 33,421,731 births over 60 years.

BACKGROUND AND AIM: Several seasonal variations have been found in birth rates in different countries at different periods. The characteristics of the rhythmic patterns vary according to geographical location and chronological changes. This study presents data on spanish birth seasonality over six decades. METHODS: A time series composed of 33,421,731 births in Spain in the period 1941-2000 was analysed. The series comes from the National Institute of Statistics and was processed according to the following norms: (1) normalisation of the duration of months and years; (2) clinical analysis of temporal series (isolation of seasonal component); (3) Fourier's spectral analysis; and (4) cosinor analysis (adjustment to the cosine curve of two harmonics). RESULTS: Significant seasonal rhythm was found in the set of births, both for a 12-month period and a 6-month period. The rhythm shows bimodal morphology, with a pronounced birth peak in April and a smaller one in September. These peaks correspond to July and December conceptions, respectively. The major birth peak shifted to March-May between the 1940s and the 1980s. Birth rhythm changed after the 1960s, with a decrease in amplitude and later loss of seasonality in the 1990s. CONCLUSIONS: In Spain, seasonal birth rhythm shows a decline from 1970, and, finally, lack of birth seasonality in 1991-2000. This trend is similar to other European countries, although Spain shows a more intense loss of seasonality.

[Differential circadian rhythms in myocardial infarction according to its extent by electrocardiogram]. / Diferencias en el ritmo circadiano del infarto de miocardio según su extensión electrocardiográfica.

BACKGROUND AND OBJECTIVE: To determine the existence of circadian rhythm in the time of onset of acute myocardial infarction (AMI) according to their extension type (Q-wave vs. non-Q-wave). PATIENTS AND METHOD: We studied a retrospective cohort of patients from a multicentre study of myocardial infarction (ARIAM study group). We collected information about 54,249 infarctions from the data base of the ARIAM (Analysis of Delay in AMI) Spanish multicentre study. The following variables were analysed: general variables --age, gender, previous ischemic heart disease, outcome at coronary care unit, infarction electrocardiograph type (Q wave or non-Q wave) and location of AMI--, cardiovascular risk factors, and previous drug treatment of the patients. To verify the presence of circadian rhythm we developed a simple test of equality of time series based on the multiple-sinusoid cosinor analysis. Three sinusoids (24-12-8 h periods) were used. RESULTS: The time of pain onset shows circadian rhythm (p < 0.0000), which also is observed in both infarction electrocardiograph characteristics subgroups (Q-wave infarction and non-Q-wave infarction) (p < 0.0000). Q-wave infarction shows sinusoid curve with one maximum morning peak and non-Q-wave shows bimodal curve, with two peaks. Comparison between their curves shows statistical significance (p < 0.0000). CONCLUSIONS: AMI onset follows a circadian rhythm pattern, which is also observed in analysed subgroups. Differences in the circadian rhythm according to the Q/non-Q wave infarction characteristics, could be determined by different physiopathologic mechanism. The cosinor model fit with three components (24, 12 and 8-hour-periods) show good sensitivity to determine circadian rhythm.