The concept of individual semantic maps(2012-Q&Q).pdf

The concept of individual semantic maps in clinical psychology: a feasibility study on a new paradigm Enzo Grossi, Angelo Compare & Massimo Buscema Quality & Quantity International Journal of Methodology ISSN 0033-5177 Qual Quant DOI 10.1007/s11135-012-9746-8 1 23 Your article is protected by copyright and all rights are held exclusively by Springer Science+Business Media B.V.. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your work, please use the accepted author’s version for posting to your own website or your institution’s repository. You may further deposit the accepted author’s version on a funder’s repository at a funder’s request, provided it is not made publicly available until 12 months after publication. 1 23 Author's personal copy Qual Quant DOI 10.1007/s11135-012-9746-8 The concept of individual semantic maps in clinical psychology: a feasibility study on a new paradigm Enzo Grossi · Angelo Compare · Massimo Buscema © Springer Science+Business Media B.V. 2012 Abstract In this paper we propose a new technology able to map the underlying connection scheme among several psychological variables in a single individual. Nine patients with chronic heart failure underwent at regular intervals, two electronic questionnaires to evaluate depression (STAI—short form) and anxiety (STAY-6). Individual semantic maps were developed by Auto Contractive Map, a new data mining tool based on an artificial neural networks acting on the small data set formed by questionnaires items applied serially along time. The clinical psychologist involved in the clinical evaluation of the cases was asked to score the consistency between the information emerging from the graph depicting the structure of the main relationships among items and the clinical picture resulting from the psychological colloquium. All cases reported overall judgments of a good consistency suggesting that the mathematical architecture of the system is able to capture in the dynamics of items value variations through time the underlying construct of the patient psychological status. This technology is promising in remote monitoring of patients’ psychological condition in different settings with the possibility to implement personalized psychological interventions. Keywords Semantic map · Single individual · Artificial neural networks · Anxiety · Depression E. Grossi (B) Medical Department, Bracco SpA, Via E. Folli 50, 20136 Milan, Italy e-mail: [email protected] A. Compare Bergamo University, Bergamo, Italy M. Buscema Research Centre of Sciences of Communication, Rome, Italy 123 Author's personal copy E. Grossi et al. 1 Introduction The statistics of single individual is an unfulfilled dream. A set of variables collected to study a specific medical problem constitutes a geometric spatial structure heavily characterized not properly by the actual properties of the variables per se, but instead by the network of associations of the variables according to their strength. In other words, the emerging relations among variables more than the variables per se might constitute the real conceptual structure of the medical condition for which that set of variables has been collected. The pattern of this relations structure is very difficult to discern with classical statistical techniques. Complex networks mathematical is helping us in establishing the hierarchy of variables within a specific set, generally looking to the presence of “hubs”, i.e. variables with the highest number of connections in an undirected graph in which a minimum spanning tree is built up starting from the complete matrix of distances of variables. In recent year some of us have developed new data mining systems based on a special kind of artificial neural networks (ANN) able to develop a semantic connectivity map. More specifically we have introduced a new methodology based on an ANN architecture, the Auto Contractive Map (AutoCM), which allows for basic improvements in both robustness of use in badly specified and/or computationally demanding problems, and output usability and intelligibility. In particular, AutoCM ‘spatialize’ the correlation among variables by constructing a suitable embedding space where a visually transparent and cognitively natural notion such as ‘closeness’ among variables reflects accurately their associations. Through suitable optimization techniques that will be introduced and discussed in detail in what follows, ‘closeness’ can be converted into a compelling graph-theoretic representation that picks all and only the relevant correlations and organizes them into a coherent picture. Such representation is not actually constructed through some form of cumbersome aggregation of two-by-two associations between couples of variables, but rather by building a complex global picture of the whole pattern of variation. Moreover, it fully exploits the topological meaningfulness of graph-theoretic representations in that actual paths connecting nodes (variables) in the representation carry a definite meaning in terms of logical interdependence in explaining the data set variability. We are aware of the fact that these techniques are novel and therefore not entirely understood so far in all of their properties and implications, and that further research is called for to explore them. But at the same time we are convinced that their actual performance in the context of well-defined, well understood problems provides an encouraging test to proceed in this direction. In recent years this technique has been applied in a number of medical settings like Alzheimer disease (Licastro et al. 2010a,b), Down syndrome (Coppedè et al. 2010), gastrooesophageal reflux disease (Buscema and Grossi 2008), uterine growth retardation (Buscema and Grossi 2008) and myocardial infarction (Street et al. 2008) showing the added value of this approach in comparison with traditional statistical techniques. As a further step we have now envisaged the way to apply this technique at single subject level. It is well known that the prerequisite to apply any statistical technique at single individual level is to handle a data set rather than a simple input vector. From a statistical point of view the value of a single measurement carried out just once in a single individual is extremely poor. Every kind of statistical inference looses its meaning 123 Author's personal copy The concept of individual semantic maps in clinical psychology (and confidence interval does not escape from this rule) in absence of a sample which by definition requests a value on N > 1. This trivial observation implies a series of very important consequences often overlooked. The most relevant consequence is that one cannot utilize inferentially in single individuals tools, like multi-items scales, born to define properties of groups of people. Psychological tests based on multi-items questionnaire applied serially along time allow to build small data set relative to a single individual experience. As shown in Table 1 each row represents a specific timeline of the application of the scale to the subject, while the column represents the single items values. The AutoCM system when applied to samples can work efficiently also with a relative small number or records (e.g. 10) due to the fact that adaptive learning algorithms of inference, based on the principle of a functional approximation as ANN, overcome the problem of dimensionality. Therefore this approach can be applied to single cases in which a specific questionnaire has been applied several times along a certain period. In this paper we show the feasibility of individual semantic maps in a group of subjects undergoing a rehabilitation program after cardiovascular acute events followed for 3–4 months with two questionnaires exploring anxiety and depression through a computerized electronic diary. The results obtained seem to delineate a new methodological paradigm in this specific context of applied clinical psychology. 2 Methods 2.1 Patients The study sample consists of nine patients with chronic heart failure with an ejection fraction lower than 40 %, in III–IV NYHA class at hospitalization. These patients took part in a the study called ICAROS carried out at three Italian research centers and monitored for one year in order to compare the evolution of clinical and functional conditions between the Integrated Management (IntM). Patients in the IntM group were monitored daily through a PDA. 2.2 Overall system architecture A new technology providing a continuous interaction between patient and doctor, has been developed and employed. The front-end designed for the patient provides a portable solution for the day-to-day management of the patient’s healthcare needs, while the clinical front-end provides both medical and psychological web-based software designed to facilitate carerelated decisions. The data repository and the core applications are hosted in a remote server. The patient front-end consists in a health organizer resident on the patient’s PDA or Smartphone. This system is designed to help individuals better manage their lifestyle, medical care and drug intake, to track treatment progress, and to enter information about their therapy by inputting actions performed (e.g., drug intake, exercise, questionnaires). All information collected is automatically reported through synchronization/downloading to their healthcare professional. Questionnaires are submitted periodically, by a fixed time frequency (see Table 1) while the time within the scheduled day is chosen randomly in order to monitor and evaluate psychological conditions more objectively. 123 Author's personal copy E. Grossi et al. Table 1 The translation of serial application of two hypothetical scales from a single patient in an individual data set Scale A Scale B Item 1 Item 2 Item 3 Item 4 Item 5 Item 1 Item 2 Item 3 Item 4 Item 5 Item 6 Item 7 Week 1 2 3 1 3 2 4 3 4 4 3 4 4 Week 2 1 4 2 2 1 3 4 3 3 4 3 3 Week 3 2 3 2 3 1 5 4 5 5 5 5 5 Week 4 1 2 2 2 1 6 6 6 3 5 3 6 Week 5 2 3 3 3 2 5 4 5 5 4 5 5 Week 6 2 2 1 2 3 4 3 4 4 5 4 4 Week 7 3 2 2 2 2 6 6 6 5 7 6 6 Week 8 2 2 3 1 3 4 5 4 4 5 4 4 Week 9 1 1 2 2 2 5 6 5 5 6 5 5 Week 10 2 1 1 3 1 3 5 3 3 5 3 3 Table 2 Psychological constructs and relative assessment instruments STAI—short form PHQ-9 Psychological construct State anxiety Depression Assessment frequency Weekly Bi-weekly Type of answer Multiple choice Multiple choice 2.3 Handheld device application The application serves as a diary for the collection of clinical and vital data, and completed psychological questionnaires. Collected data can be either locally stored (in the memory of the PDA) or sent to the central repository. 2.4 Measures Patient’s front-end presents, at regular intervals, by PDA questionnaires aimed to evaluate depression, anxiety and perception of quality of life. Data synchronization occurs in the same way as it does for medical data. Table 2 lists administered questionnaires, including timing, psychological constructs investigated, and response modality. Anxiety is measured weekly using a short form of Spielberger’s State Trait Anxiety Inventory (STAI-6). Six items were selected by the authors from the state scale, which measures anxiety in terms of feelings of insecurity and powerlessness in front of a perceived threat (Spielberger et al. 1999; Spielberger and Sydeman 1994). Depression is evaluated using the Patient Health Questionnaire (PHQ-9). The PHQ is frequently used to diagnose depressive conditions in medical settings. It consists of nine items for evaluating DSM-IV criteria for major depression. Two studies, one involving 3,000 patients with a number of medical problems, and the other, also conducted on 3,000 patients at gynecological-obstetric clinics, show the instrument’s diagnostic value in relation to measures comprising a larger number of items. Moreover, PHQ-9 not only helps to establish the presence of a depressive condition, but also provides an index of severity of depressive symptomatology (Kroenke et al. 2001; Lowe et al. 2004). In this paper we present the results obtained by combining together depression and anxiety items from the two corresponding scales. The items are abbreviated as following: 123 Author's personal copy The concept of individual semantic maps in clinical psychology Items PHQ-9 DEP-No-Interest DEP-Depressed DEP-No-Sleep DEP-Tired DEP-Too/No-Appetite DEP-Guilt DEP-No-Concentration DEP-Slow/Exited DEP-dead STAY-6 ANX_Calm ANX_Tens ANX_Troub ANX_No Relax ANX_No Happy ANX_Worry Description Little interest or pleasure in doing things Feeling down, depressed, or hopeless Trouble falling or staying asleep, or sleeping too much Feeling tired or having little energy Poor appetite or overeating Feeling bad about yourself, or that you are a failure, or have let yourself or your family down Trouble concentrating on things, such as reading the newspaper or watching television Moving or speaking so slowly that other people could have noticed. Or the opposite—being so fidgety or restless that you have been moving around a lot more than usual Thoughts that you would be better off dead, or of hurting yourself in some way Calm condition state Tension condition state Trouble condition state No relaxed condition state No happy condition state Worry condition state In view of the fact that anxiety and depression scales had different frequency of application, we skipped for the depression one of the bi-weekly assessment employing for both one assessment per week performed within one calendar day variation. AutoCM system This goal to define the inherent structure of items relationship in a single subject has been achieved through a new data mining method able to point out the relative relevance of each variable in representing a major biological hub. This method is based on an artificial adaptive system able to define the association strength of each variable with all the others in any dataset, named the AutoCM. The architecture and mathematics of AutoCM is described elsewhere (Licastro et al. 2010a; Buscema and Grossi 2008). The AutoCM is a new data mining tool based on an ANN developed at Semeion Research Centre. The AutoCM is a special kind of ANN able to find, by a specific data mining learning algorithm, the consistent patterns and/or systematic relationships and hidden trends and associations among variables. After the training phase the weights developed by AutoCM are proportional to the strength of associations of all variables each-other. The weights are then transformed in physical distances. Variables couples whose connection weights are higher become nearer and vice versa. A simple mathematical filter represented by minimum spanning tree is applied to the distances matrix and a graph is generated. This allows seeing connection schemes among variables and detecting variables acting ad “hubs”, being highly connected. This matrix of connections preserves non linear associations among variables and captures connection schemes among clusters. 123 Author's personal copy E. Grossi et al. This matrix of connections presents many suitable features: (a) Non linear associations among variables are preserved; (b) Connections schemes among clusters of variables is captured, and (c) Complex similarities among variables became evident. The AutoCM connections matrix filtered by MST generates an interesting graph, whose biological evidence has been already tested in medical field defined connectivity map as detailed by Buscema and Grossi (2008). The equations inherent to AutoCM system are shown below: AutoCM: the equations Signal transfer and learning:   ν [n] (1) h i = xi · 1 − Ci ;  (2) νi = (xi − h i ) · 1 − νi[n] C  · xi ; (3) νi[n+1] = νi[n] + νi ;   N  w[n] h j · 1 − Ci, j ; (4) Neti = j=1   (5) yi = h i · 1 − NCeti ;   wi,[n]j (6) wi, j = (h i − yi ) · 1 − C · h j ; (7) wi,[n+1] = wi,[n]j + wi, j . j Convergence condition: (8) lim wi, j = 0, ∀vi = C. n→∞ where x is the input vector, h the hidden units, y the output vector, ν the input–hidden weights, √ w the hidden–output weights, N the input number, C the constant (typically: C = N ), i, j ∈ [1, 2, ..., N ], and n the number of iteraction cycles. The specificity of AutoCM algorithm is to minimize a complex cost function respect to the traditional ones: Traditional energy minimization equation ⎫ ⎧ ⎬ ⎨ N N M q q xi · x j · σi, j ; σi, j > 0. E = Min ⎭ ⎩ q i j AutoCM energy minimization equation ⎧ N ⎨ N N E = Min ⎩ i q ⎫ ⎬ M j=i k = j =i q q q q xi · x j · xk · Ai, j · Ai,k ; ⎭ w w where xi is the value of ith input unit at qth pattern, Ai, j = 1 − Ci, j , Ai,k = 1 − Ci,k , wi, j < C, wi,k < C, i, j, k is the indices for the variables—columns, q the index for the patterns—rows, N the number of variables (columns), and M the number of patterns (rows). From mathematical point of view is evident how the traditional minimization includes only second order effects, while the AutoCM considers also a third order. Practically, this 123 Author's personal copy The concept of individual semantic maps in clinical psychology means that AutoCM algorithm is able to discover variables similarities completely embedded into the dataset and invisibles to the other classic tools. This approach highlights affinities among variables as related to their dynamical interaction rather than to their simple contingent spatial position. This approach describes a context typical of living systems where a continuous time dependent complex change in the variable value is present. After the training phase, the matrix of the AutoCM represents the warped landscape of the dataset. We apply a simple filter (minimum spanning tree) to the matrix of AutoCM system to show the map of main connections between and among variables and the principal hubs of the system. These hubs can also be defined as variables with the maximum amount of connections in the map. The AutoCM algorithms used for all the elaborations presented in this paper are implemented only in a Semeion proprietary research software, available only for academic purposes. 3 Results The Figures 1, 2, 3, 4, 5, 6, 7, 8 and 9 represent the semantic connection maps related to the nine patients. The graphs depict the structure of the main relationships among items according to the minimum spanning tree build from the weights matrix developed by AutoCM system. The circles represent the hubs of the graphs, i.e. the items having the highest relevance among the others. The detailed description of the psychological profile of each patient registered by a clinical psychologist not aware of the graph is reported in the Appendix 2. The clinical psychologist involved in the clinical evaluation of the cases was asked to score the consistency between the information emerging from the graph and the clinical picture resulting from the psychological colloquium according a scale ranging from 0 = absolute absence of consistency to 10 = perfect consistency. Scores ranging from 0 to 2 were considered expression of a overall judgment of a scarce consistency, scores from 3 to 5 of a mild consistency, scores from 6 to 8 of a good consistency and from 9 to 10 of a very good consistency (Table 3). All cases reported overall judgments of a good consistency irrespectively to the number of hubs and number of links per hub. Therefore rather than quantitative features of the graphs were the qualitative features (matching of clinically important variables with hubs and vice versa) to influence the judgment. The matching is described in the Appendix 2 4 Discussion This study for the first time shows the feasibility of a semantic map analysis applied to the single individual. The obtained results seem to show that thanks to AutoCM algorithm a statistics of single individual is feasible, provided that a minimum number of test replications is available. The mathematical architecture of the system is able to capture in the dynamics of items value variations through time the underlying construct of the patient psychological status. This could be of great importance for the remote monitoring of patients affected by different form of psychological disturbances during a treatment plan, improving also the compliance and confidence of the patient in the treatment program. Similar systems for data analysis have been applied in clinical psychology to groups of patients. 123 Author's personal copy E. Grossi et al. Fig. 1 Semantic map of Subject 1343 Self-organizing map has been applied to derive semantic representations of words (Zhao et al. 2011), in analyses of cognitive test data to identify potential hidden structures in cognitive performance in a sample of schizophrenic patients and healthy individuals (Silver and Shmoish 2008) and in study of disease descriptors in a sample Alzheimer disease patients (Grossi et al. 2005). Other have used experience sampling method was used to examine the relationship between mood states and thought content in a sample of hospitalized sex offender (Hillbrand and Waite 1994). Although some studies have used similar models for data analysis, none of these models has been applied to single subjects. In clinical practice when there is a clear need to tailor a treatment this type of analysis might help to highlight the salient aspects of the individual case, both during the diagnostic and treatment phases. The maintenance of long-term outcomes after treatment is a very important goal in treating many chronic diseases (Morris et al. 2012; Renner et al. 2011; Van Buren and Sinton 2009). The psychological symptoms, when associated with a medical condition (heart disease, obesity, cancer) frequently influence the course over time. According to future DSM-5 (Rosmalen et al. 2010) it is therefore very important to monitor the progress of the psychological symptoms of patients with medical illness, in order to highlight conditions of severity and to act quickly to avoid the negative effects of psychological symptoms on physiological and behavioral healthy lifestyles aspects (Compare et al. 2011a,b; Manzoni et al. 2011). The information communication technology (ICT) applied to health has allowed the development of wireless remote monitoring tools able to detect psychological and medical parameters along time. The automatic transformation of serial psychological features 123 Author's personal copy The concept of individual semantic maps in clinical psychology Fig. 2 Semantic map of Subject 1358 Fig. 3 Sematic map of Subject 1403 collected along time by wireless monitoring systems in a semantic map could help to detect the evolution of patients’ psychological condition and automatically highlight risk and therefore allow to implement timely clinician interventions. 123 Author's personal copy E. Grossi et al. Fig. 4 Semantic map of Subject 1440 Fig. 5 Semantic map of Subject 1581 123 Author's personal copy The concept of individual semantic maps in clinical psychology Fig. 6 Semantic map of subject 1743 Fig. 7 Semantic map of subject 1760 Another important contribution of semantic maps applied to individual cases in clinical psychology concerns the dynamic screening of the risk profile associated with somatization disorders. Recent studies show (Cairney et al. 2009; McKeith and Cummings 2005) that psychological risk profiles of patients are subject to high variability over time due to several factors, both environmental and genetic. The application of semantic maps of risk profiles could help to highlight the variability of the risk profile of individual patients over time. This study has several limitations. First of all the sample size is limited to nine patients. Therefore we cannot rule out the possibility that outside this pilot feasibility study in a 123 Author's personal copy E. Grossi et al. Fig. 8 Semantic map of subject 1840 Fig. 9 Semantic map of subject 1881 123 Author's personal copy The concept of individual semantic maps in clinical psychology Table 3 Overall summary of map features and consistency score between graph information and psychological profile emerging from visit Case N. Number of hubs Number of hubs links Mean number of links per hub Consistency score 1358 3 10 3.33 7 1343 3 10 3.33 7 1403 4 13 3.25 8 1440 3 12 4.00 8 1581 3 12 4.00 7 1743 4 14 3.50 8 1760 3 9 3.00 7 1840 4 13 3.25 7 1881 3 8 2.67 8 larger setting the consistency between the map construct and the psychological visit could appear less remarkable. Secondly we have used only two psychological scales exploring anxiety and depression. It has to be shown that the integration with other kind of scales for which is feasible a serial application during short timeframe maintains the information power shown in this particular experimental design. Third, the development of semantic connectivity maps require data download from the instrument, data set preprocessing and software application; therefore for the moment the map is not immediately available to the psychologist. This problem has to be addressed with appropriate engineering and programming efforts in order to have an applet working in real time on the remote server. The multifactor analysis of the health condition and the need to personalize as much as possible the diagnosis and treatment of diseases is one of the future challenges in health sector. The analysis with the semantic maps associated to technological developments in ICT could help to tackle this challenge allowing the develop of psychological treatments not only disease-specific but also person-specific. Appendix 1: software related to AutoCM system 1. 2. 3. 4. 5. 6. Buscema M. Contractive Maps. Software for programming Auto Contractive Maps, Semeion Software #15, Rome, Ver 2.0 (2007). Buscema M. Constraints Satisfaction Networks. Software for programming Non Linear Auto-Associative Networks, Semeion Software #14, Rome, Ver 12.0 (1999–2009). Buscema M. MST. Software for programming Trees from artificial networks, Semeion Software #38, Rome, Ver 6.5 (2008–2010). Massini G. Trees Visualizer, Ver 7.0, Semeion Software #40, Rome, 2007–2010. Massini G. Semantic Connection Map, Ver 3.0, Semeion Software #45, Rome, 2007– 2010. Buscema M. Modular AutoAssociative ANN, Ver 10.0, Semeion Software #51, Rome, 2009–2010. 123 Author's personal copy E. Grossi et al. Appendix 2: detailed cases description and individual maps Case 1358 Map description Variables Dep-agitation/motor slowness Number of links 4 Dep-loss of appetite/overeating 3 Anxiety-agitation (state) 3 Variables linked Fatigue, difficulty concentrating, depression, anxiety Difficulty sleeping, suicidal ideation Unhappiness, difficulty relaxing Psychological visit Anamnestic information Patient 56 years old woman married with a son. There are no traumatic events, physical or psychological, in the past. BMI: 28.8, overweight Heart failure and metabolic syndrome Visit observations Main symptoms: During the visit, the patient showed an objective psychological symptoms of moderate depression. The clinical condition presenting symptoms be framed as a dysthymic disorder (DSMIV). Secondary symptoms: The depressive symptoms was associated with a psychological tendency to somatization by two ways: - Motor: agitation body - Eating behavior: overeating associated with emotional eating Psychometric tests: - MMSE: suitable Conclusions The graph shows the variables that characterize the salient psychological profile revealed by physical examination of the patient. In particular, the non-centrality suicidal ideation within the graph corresponds to a level of mild symptoms that coincides with the findings during the visit. The behavior of somatization, finds its counterpart in the centrality of the components of motor agitation, and overeating, without emerging from the graph. 1 2 3 4 5 6 7 8 9 10 Case 1343 Map description Variables Dep-difficulty concentrating Number of links 4 Dep-agitation/motor slowness 3 Anxiety-difficulty relaxing (state) 3 123 Variables linked Fatigue, loss of appetite/overeating/guilt, difficulty staying calm Suicidal ideation, agitation, anxiety, difficulty relaxing Slow/motor agitation, tension, anxiety, unhappiness Author's personal copy The concept of individual semantic maps in clinical psychology Psychological visit Anamnestic information Patient of 61 years, man, married with two children. There are no traumatic events, physical or psychological, in the past. BMI: 24.8 Heart failure and metabolic syndrome Visit observation Main symptoms: During the visit, the patient showed an objective psychological symptoms of anxiety can be framed as a moderate OCD (DSM-IV) associated with the compulsion to wash their hands frequently. Secondary symptoms: The symptoms was associated with a psychological profile characterized by poor concentration and hyperactive behavior has not been finalized, which appeared to be a consequence of excessive activation due to cognitive anxiety symptoms. Psychometric tests: - MMSE: suitable Conclusions The graph shows, in line with what has made the visit, the centrality of anxiety symptoms, as a main aspect, and the presence of secondary problems related to difficulties in attention and agitation. 1 2 3 4 5 6 7 8 9 10 C as e 1403 Map description Variables Dep-fatigue Number of links 4 Anxiety-unhappiness 3 Anxiety-agitation (state) 3 Anxiety-hard to stay calm (state) 3 Variables linked Agitated, anxious, depression, difficulty sleeping, agitation/slow motor Difficulty concentrating, difficulty relaxing, difficulty staying calm Concern, loss of appetite/overeating, fatigue Guilt, suicidal ideation, unhappiness Psychological visit Anamnestic information Patient 59 years old woman with a separated child. There are no traumatic events, physical or psychological, in the past. BMI: 23.6. Heart failure and metabolic syndrome Visit observation Main symptoms: During the visit, the patient showed an objective psychological symptoms of depression can be framed as a moderate dysthymic disorder (DSM-IV). Secondary symptoms: The depressive symptoms was associated with a psychological profile characterized by a high sense of fatigue and physical exhaustion. Psychometric tests: - MMSE: below the level normally 123 Author's personal copy E. Grossi et al. Conclusions The graph shows, such as visits, focused primarily on the component symptoms of depression. Similarly the visit, the graph highlights the behavioral effects (delay, difficulty concentrating, fatigue) depressive symptoms. 1 2 3 4 5 6 7 8 9 10 Case 1440 Map description Variables Anxiety-concern Number of links 5 Dep-loss of appetite/overeating 4 Anxiety-hard to stay calm 3 Variables linked Depression, fatigue, restlessness, stress, overeating/loss of appetite Worry, guilt, suicidal ideation, difficulty relaxing Unhappiness, inability to relax, sleep disorders Psychological visit Anamnestic information Patient of 62 years, woman, separated, no children. Abused in childhood. BMI: 38.8, obese Heart failure and metabolic syndrome Visit observation Main symptoms: During the visit, the patient showed an objective psychological symptoms of anxiety can be framed as a moderate generalized anxiety disorder (DSM-IV). Secondary symptoms: Be framed as an eating disorder BED (DSM-IV) and characterized by intense craving with self-injurious behavior. Psychometric tests: - MMSE: suitable Conclusions The graph shows, such as a visit, symptoms mainly focused on the anxiety component. Similarly the visit, the graph highlights the effects of secondary symptoms characterized by an eating disorder associated with guilt and suicidal thoughts (self-injury highlighted during the visit). 1 2 3 4 5 6 7 8 9 10 Case 1581 Map description Variables Dep-difficulty concentrating Number of links 5 Dep-depression 4 Anxious-unhappy 3 123 Variables linked Slow/motor hyperactivity, guilt, sleep disorders, suicidal thoughts, difficulty staying calm Loss of appetite/overeating, tension, agitation, trouble staying calm fatigue, difficulty relaxing, restlessness Author's personal copy The concept of individual semantic maps in clinical psychology Psychological visit Anamnestic information Patient of 58 years, male, married with three children. There are no traumatic events, physical or psychological, in the past. BMI: 23.7, Heart failure and metabolic syndrome Visit observation Main symptoms: During the visit, the patient showed an objective psychological symptoms of depression can be framed as a high major depressive disorder (DSM-IV). Secondary symptoms: The depressive symptoms was associated with a psychological profile characterized by cognitive deficits have been linked to memory, and attention. Psychometric tests: - MMSE: below the level normally Conclusions The graph shows, like the visit, the centrality of depressive symptoms by emphasizing also the secondary symptoms that affect the cognitive component. 1 2 3 4 5 6 7 8 9 10 Case 1743 Map description Variables Dep-fatigue Number of links 4 Dep-depression Dep-guilt 3 3 Dep-suicidal ideation 4 Variables linked Sleep disturbances, lack of interest, anxiety, depression, Agitation, guilt, fatigue Depression, suicidal ideation, loss of appetite/overeating Unhappiness, lack of relaxation, difficulty concentrating, feelings of guilt Psychological visit Anamnestic information Patient 54-year-old male with a separated child. There are no traumatic events, physical or psychological, in the past. BMI: 26.8, Heart failure and metabolic syndrome Visit observation Main symptoms: During the visit, the patient showed an objective psychological symptoms of depression can be framed as a high major depressive disorder (DSM-IV). Secondary symptoms: During the interview disclosed any factor related to the responsibility and the sense of failure for the separation. The patient reported suicidal ideation, Psychometric tests: - MMSE: suitable 123 Author's personal copy E. Grossi et al. Conclusions The graph shows, like the visit, the absolute centrality of the depressive component. The anxiety component is marginal. In addition, the severity of the symptoms is underlined also in the graph, as well as in the visit, highlighting the role of the variable related to suicidal ideation. 1 2 3 4 5 6 7 8 9 10 Case 1760 Map description Variables Anxiety-hard to stay calm Number of links 3 Variables linked Guilt, lack of relaxation, hyperactivation/motor slowing Anxiety-difficulty relaxing 3 Dep-lack of interest 3 Sleep disturbance, unhappiness, trouble staying calm Voltage, loss of appetite/overeating, preoccupation Psychological visit Anamnestic information Patient of 59 years, woman, married with two children. There are no traumatic events, physical or psychological, in the past. BMI: 24.8, Heart failure and metabolic syndrome Visit observation Main symptoms: During the visit, the patient showed an objective psychological symptoms of anxiety and depressive moderate framed as cyclothymic disorder (DSM-IV). Secondary symptoms: during the interview, the patient has been particularly agitated (speech kidnapped, rapid movements of the body) Psychometric tests: - MMSE: suitable Conclusions The graph shows, like the visit, the presence of symptoms does not "core" but the copresence of depressive symptoms and anxiety symptoms, that the visit has shown to be temporally and alternately connected to each other. 1 2 3 4 5 6 7 8 9 10 Case 1840 Map description Variables Anxiety-tension Anxiety-agitation Number of l inks 3 3 Dep-guilt 4 Dep-lack of interest 3 123 Variables linked Lack of interest, guilt, agitation Loss of appetite/overeating, diff. to relax, tension Suicidal ideation, poor concentration, lack of calm, tension Fatigue, depression, tension Author's personal copy The concept of individual semantic maps in clinical psychology Psychological visit Anamnestic information Patient of 64 years, woman, separated. There are no traumatic events, physical or psychological, in the past. BMI: 24.3. Heart failure and metabolic syndrome Visit observation Main symptoms: During the visit, the patient showed an objective psychological symptoms of anxiety can be framed as a moderate generalized anxiety disorder (DSM-IV). Secondary symptoms: The depressive symptoms was associated with a psychological tendency to somatization by overeating associated with emotional eating, and tendency to brooding cognitive connected to repetitive thoughts of guilt Psychometric tests: - MMSE: suitable Conclusions The graph shows the centrality of anxiety symptoms, compared to the depressive appears to be mild and secondary. Similarly, as shown in the cards, reveal anything related to depressive guilt as salient aspects of the clinical picture of the patient. 1 2 3 4 5 6 7 8 9 10 Case 1881 Map description Variables Dep-depression Number of links 2 Anxiety-tension 2 Anxious-unhappy 4 Variables linked Sleep disorders, suicidal ideation Difficulty in concentration, unhappiness Slow/motor hyperactivity, lack of relaxation, restlessness, tension Psychological visit Anamnestic information Patient of 54 years, man, married with two children. There are no traumatic events, physical or psychological, in the past. BMI: 25.1, Heart failure and metabolic syndrome Visit observation Main symptoms: During the visit, the patient showed an objective psychological symptoms polarized along two dimensions: anxiety and depression. The picture is framed as depressive dysthymic disorder (DSM-IV) and anxiety symptoms are being viewed as a health-related anxiety (DSM-IV) which was severe. Secondary symptoms: High difficulty relaxing during the interview Psychometric tests: - MMSE: suitable Conclusions The graph shows, such as visiting a polarized symptoms. It is also highlighted as a major difficulty in relaxing central anxiety symptoms. The theme of happiness is not found during the visit but it could be envisaged as a factor underlying the anxiety related to health/disease. 1 2 3 4 5 6 7 8 9 10 123 Author's personal copy E. Grossi et al. References Buscema, M., Grossi, E.: The semantic connectivity map: an adapting self-organising knowledge discovery method in data bases. Experience in gastro-oesophageal reflux disease. Int. J. Data. Min. Bioinform. 2(4), 362–404 (2008). PubMed PMID:19216342 Cairney, J., Faulkner, G., Veldhuizen, S., Wade, T.: Changes over time in physical activity and psychological distress among older adults. Can. J. Psychiatry 54(3), 160–169 (2009) Compare, A., Germani, E., Proietti, R., Janeway, D.: Clinical psychology and cardiovascular disease: an up-to-date clinical practice review for assessment and treatment of anxiety and depression. Clin. Pract. Epidemiol. Ment. Health 7, 148–156 (2011a) Compare, A., Proietti, R., Grossi, E., Del Forno, A., Giallauria, A., Vitell, A., Grieco, C., Vigorito, C.: Vulnerable personality and Takotsubo cardiomyopathy consequent to emotional stressful events: a clinical case report. Monaldi Arch. Chest Dis. 76(2), 99–103 (2011b) Coppedè, F., Grossi, E., Migheli, F., Migliore, L.: Polymorphisms in folate-metabolizing genes, chromosome damage, and risk of Down syndrome in Italian women: identification of key factors using artificial neural networks. BMC Med. Genomics 3, 42 (2010). PubMed PMID:20868477; PubMed Central PMCID:PMC2949778 Grossi, E., Massini, G., Buscema, M., Savarè, R., Maurelli, G.: Two different Alzheimer diseases in men and women: clues from advanced neural networks and artificial intelligence. Gend. Med. 2(2), 106–117 (2005) Hillbrand, M., Waite, B.M.: The everyday experience of an institutionalized sex offender: an idiographic application of the experience sampling method. Arch. Sex. Behav. 23(4), 453–463 (1994) Kroenke, K., Spitzer, R.L., Williams, J.B.: The PHQ-9: validity of a brief depression severity measure. J. Gen. Intern. Med. 16, 606–613 (2001) Licastro, F., Porcellini, E., Chiappelli, M., Forti, P., Buscema, M., Ravaglia, G., Grossi, E.: Multivariable network associated with cognitive decline and dementia. Neurobiol. Aging 31(2), 257–269 (2010a). PubMed PMID:18485535 Licastro, F., Porcellini, E., Forti, P., Buscema, M., Carbone, I., Ravaglia, G., Grossi, E.: Multi factorial interactions in the pathogenesis pathway of Alzheimer’s disease: a new risk charts for prevention of dementia. Immun. Ageing 16(7 Suppl 1), S4 (2010b). PubMed PMID:21172063; PubMed Central PMCID:PMC3024878 Lowe, B., Kroenke, K., Herzog, W., Grafe, K.: Measuring depression outcome with a brief self-report instrument: sensitivity to change of the Patient Health Questionnaire (PHQ-9). J. Affect. Disord. 81, 61–66 (2004) Manzoni, G.M., Villa, V., Compare, A.: Short-term effects of a multi-disciplinary cardiac rehabilitation programme on psychological well-being, exercise capacity and weight in a sample of obese in-patients with coronary heart disease: a practice-level study. Psychol. Health Med. 16(2), 178–189 (2011) McKeith, I., Cummings, J.: Behavioural changes and psychological symptoms in dementia disorders. Lancet Neurol. 4(11), 735–742 (2005) Morris, J., Oliver, T., Kroll, T., MacGillivray, S.: The importance of psychological and social factors in influencing the uptake and maintenance of physical activity after stroke: a structured review of the empirical literature. Stroke Res. Treat. 2012, 1–20 (2012) Renner, B., Hankonen, N., Ghisletta, P., Absetz, P.: Dynamic psychological and behavioral changes in the adoption and maintenance of exercise. Health Psychol. 31, 306–315 (2011) Rosmalenh, J.G., Tak, L.M., de Jonge, P.: Empirical foundations for the diagnosis of somatization: implications for DSM-5. Psychol Med. 16, 1–10 (2010) Silver, H., Shmoish, M.: Analysis of cognitive performance in schizophrenia patients and healthy individuals with unsupervised clustering models. Psychiatry Res. 159(1–2), 167–179 (2008) Spielberger, C.D., Sydeman, S. : State-Trait Anxiety Inventory and State-Trait Anger Expression Inventory. In: Maruish, M.E. (ed.) The Use of Psychological Testing for Treatment Planning and Outcome Assessment, pp. 292–321. Lawrence Erlbaum, Hillsdale (1994) Spielberger, C.D., Sydeman, S., Owen, A.E., Marsh, B.J. : Measuring anxiety and anger with the StateTrait Anxiety Inventory (STAI) and the State-Trait Anger Expression Inventory (STAXI). In: Maruish, M.E. (ed.) The Use of Psychological Testing for Treatment Planning and Outcomes Assessment, 2nd edn, pp. 993–1021. Lawrence Erlbaum, Mahwah (1999) Street, M.E., Grossi, E., Volta, C., Faleschini, E., Bernasconi, S.: Placental determinants of fetal growth: identification of key factors in the insulin-like growth factor and cytokine systems using artificial neural networks. BMC Pediatr. 8, 24 (2008). PubMed PMID:18559101; PubMed Central PMCID: PMC2438355 123 Author's personal copy The concept of individual semantic maps in clinical psychology Van Buren, D.J., Sinton, M.M.: Psychological aspects of weight loss and weight maintenance. J. Am. Diet. Assoc. 109(12), 1994–1996 (2009) Zhao, X., Li, P., Kohonen, T.: Contextual self-organizing map: software for constructing semantic representations. Behav. Res. Methods 43(1), 77–88 (2011) 123