The neuroimaging contribution to the Schizophrenia diagnosis and monitoring: A brief Systematic Review.

ABSTRACT

Introduction: Structural and functional neuroimaging in schizophrenia has gained strength in recent years, so this review aims to describe neuroimaging findings that contribute  to  the  physiopathological  understanding,  monitoring,  and  diagnosis of this pathology. Methods: A PubMed/Medline search was conducted for clinical studies  addressing  neuroimaging  in  schizophrenia.  Results:  The  search  yielded
2200 results, from which 13 studies were included, which provided findings, such as functional and structural alterations of the amygdala, which have shown to be associated with negative symptoms; morphometric reduction of the frontal region, alterations in the perfusion of the anterior cingulate gyrus and the lower-left parietal cortex, deregulation of the histone deacetylase enzyme, among others which translate clinically in neurocognitive deficits. Conclusions: This review provides an updated view on the findings of neuroimaging that can contribute to the understanding of the pathological mechanisms behind this psychotic disorder, its diagnostic usefulness, and its potential contribution to the prognosis and follow-up of this disease.

Keywords: Schizophrenia, neuroimaging, diagnostic, monitoring.

INTRODUCTION

Schizophrenia has a prevalence of slightly less than  1%  of  the  world  population,  although  it varies   according   to   geographical   location(1,2). It is characterized as a highly symptomatically heterogeneous psychotic disorder, characterized by positive, negative, and cognitive symptoms. The pathogenesis of this disease involves a complex interplay between genetic and environmental factors that alter early brain development and condition its biological adaptations throughout life(2,3). Although its underlying mechanisms are not yet fully elucidated, schizophrenia has been associated with altered brain connectivity.

Structural and functional neuroimaging studies in schizophrenia have been frequent in recent years. Magnetic Resonance Imaging (MRI) studies, both structural (sMRI) and functional (fMRI), and Positron Emission Tomography (PET), among others, have contributed to the study   of   neurobiological   alterations   in   the brains of these patients. Thus, abnormal brain volumes  have  been  found  in  subjects  from the first psychotic episode, as well as in long- standing conditions (called “residual or chronic schizophrenia”)(4).  As  the  disorder  progresses, the reduction in cortical volume becomes more pronounced and is associated with impaired cognitive function.(5)

Furthermore,  abnormal  information  processing has been linked to positive and negative symptoms characteristic of schizophrenia, with functional neuroimaging showing altered activation in cortical and subcortical structures in patients with schizophrenia.  In addition, it has been presumed that the connectivity of these regions also appears to be altered and associated with abnormal activation.(6)

This systematic review aims to describe neuroimaging findings in people with schizophrenia that contribute to the study of their neurodynamic mechanisms and to evaluate their usefulness in the diagnosis and follow-up of this pathology.

METHODS

This systematic review was constructed based on the PRISMA reporting guideline.(7)

Eligibility criteria

Randomized  observational  and  clinical  studies that studied patients with schizophrenia using neuroimaging, published in the last 10 years, were included in this review.

On the other hand, articles that had patients with other psychiatric illnesses and/or that did not use neuroimaging were excluded.

Database, search, and selection of studies

An  advanced  search  was  carried  out  using the    PubMed/Medline    platform    with    the terms “schizophrenia”, “neuroimaging” and “diagnostic biomarkers” present in the titles and/or abstracts of the publications. The initial search included all studies published between
15 July 2010 and 15 July 2020, the results of which were narrowed down after applying the filters “clinical study”, “observational study”, “randomized clinical trial”, “humans” and “10 years”. The titles and abstracts of each filtered article were then analyzed to pre-select those that were within the focus of this review. Following full-text analysis, the risk of bias of each study was assessed before final inclusion in this systematic review. A further 12 articles are cited to the extent that conceptual support was required for the introduction, methods, and discussion.

To describe neuroimaging findings that contribute to the study of neurodynamic mechanisms in schizophrenia, clinical studies were selected that presented a robust statistical analysis to verify that the neuroimaging findings were not due to chance and secondly had a clinical association with the patient’s symptomatology. On the other hand, to evaluate the usefulness of neuroimaging in the follow-up of the pathology, we selected clinical

studies   that   evaluated   neuroimaging   changes in schizophrenic patients undergoing therapies (pharmacological or non-pharmacological), verifying that these changes correlated in a statistically significant way with the images. Finally, a study was sought to evaluate the sensitivity and specificity of the technique in this diagnosis.

Risk of bias, data presentation, and analysis

The  Cochrane  Collaboration  tool  was  used  to assess the risk of bias in each pre-selected study.The   results   of   this   systematic   review   are presented in the form of summary tables with the imaging findings of each study, whose analysis and comparison with other systematic reviews and meta-analyses were purely qualitative.

RESULTS

The initial search yielded 2200 results, which
were reduced to 30 results after applying the filters mentioned above. Based on the analysis of  titles  and  abstracts,  20  publications  were pre-selected as being within the focus of this article. After full-text analysis, 5 articles were discarded  because  they  lacked  neuroimaging in their methodology and/or did not identify schizophrenic patients as the population of interest.  The risk of bias of each study was assessed  at  the  methodological  level,  where 2 articles were discarded due to selection, measurement, and loss bias, leaving 13 articles that were finally included due to their relevance and relevance to the objective of this systematic review. Of these 13 articles, 7 are observational studies and 6 are randomized clinical trials. The search strategy with subsequent exclusion and selection of studies is presented using a flow chart in figure 1.

Figure 1. Search and selection flowchart.

Contribution of neuroimaging to the study of neuropathological mechanisms underlying schizophrenia

Structural findings

The study of structural neuroimaging has shown a reduction in brain volume, initially localized (insula, cingulum, hippocampus, among others); however, over time, this volumetric reduction ends up being generalized. Structural alteration has been evidenced in both cortical and subcortical regions and white matter tracts, where studies such as that of Morgan et al. 2019 have shown significant morphometric differences in 18 cortical regions between schizophrenic patients and controls, including decreased volume in the superior frontal, caudal frontal, precentral, parstriangularis, and superior temporal regions, as well as greater volume in parietal and superior postcentral areas.(8)

On the other hand, negative symptoms of schizophrenia have been associated with structural and functional alterations of the amygdala. The study by Rahm et al. 2015 compared a control group with a group of patients with schizophrenia, where a negative correlation was found between stereotyped thinking and the volume of the right amygdala.(9)

Progressive white matter changes have also been described,  which  according  to  a  prospective case-control study conducted by Kraguljac et al.
2019 are microstructural changes, where using diffusion tensor imaging a reduction of fractional anisotropy was found in the left medial temporal white matter compared to controls (p < 0.05), while mean diffusivity increased in the fusiform area and white matter of the lingual gyrus compared to controls.(10)

Functional findings
Structural alterations of the amygdala associated with    negative    symptoms    of    schizophrenia have been described previously, but functional alterations of the amygdala have also been reported. The study by Rahm et al. 2015, mentioned above, found a negative correlation between emotional blunting and neural activation in  the  left  amygdala   during  the  processing of positive symptoms (features added to the individual’s behavior as a result of the disorder and  not  normally  observed  in  healthy  people, such as hallucinations, delusions, disorganized thinking, etc.) in patients with schizophrenia compared to a control group.(9)

Other studies such as that of Faget-Agius et al.
2017 showed that the use of brain Single-Photon Emission Computed Tomography (brain SPECT) shows a direct association between the levels of functionality of schizophrenic patients and the perfusion of the anterior cingulate gyrus and the left inferior parietal cortex, both areas related to emotional regulation and negative symptoms of this disease.(11)

It has been described that schizophrenia could be due to alterations in brain connectivity affecting neuronal networks at different levels, the  Default-Mode  Network  (DMN)  is  one  of the most studied networks. Recently, using Fractional Amplitude of Low-Frequency Fluctuation (fALFF), hyperactivity of the DMN has been shown in previously untreated patients with paranoid schizophrenia in the first episode, as a result of a compensatory effort modulated by an inflammatory process that takes place in the initial phases of the disease, which could translate into an alteration of self-referential and reflective activity, as well as attention to internal and external stimuli.(12)

In a case-control study by Gilbert et al. 2019, Positron Emission Tomography (PET) showed lower histone deacetylase enzyme expression in the dorsomedial prefrontal cortex and orbitofrontal gyrus, and higher histone deacetylase enzyme expression in the cerebral white matter, pons and cerebellum compared to controls.(13)

In another case-control study, Takeshi et al. 2010 used near-infrared spectroscopy (NIRS), which measured  the  concentration  of  oxyhemoglobin in different brain areas of schizophrenic patients during idea fluency tests (to assess disintegrated thinking) and letter fluency tests, showed a significant increase in the concentration of oxyhemoglobin in different brain areas of schizophrenic patients, showing a significant increase in concentration in the ventral area of the frontopolar cortex (p = 0.002 to 0.036) and the  dorsal  area  of  the  frontopolar  cortex  and the dorsolateral prefrontal cortex (p = 0.000 to 0.016)  compared  to  controls(14).  The  increase in oxyhemoglobin concentration in the ventral part  of  the  frontopolar  region  during  the  test of  the  fluency  of  ideas  correlated  positively and  significantly  with  the  Global  Assessment of Functioning (GAF) score of schizophrenic patients.(14)

The information presented above is summarized in
Table 1.

Contribution of neuroimaging to the diagnosis of Schizophrenia

In a prospective observational study, Karageorgiou et al. 2011 developed early diagnostic models for schizophrenia using structural Magnetic Resonance Imaging (MRI) and Neuropsychological Testing (NP)(15). These models are based on a Linear Discriminant  Analysis  (LDA),  which  provides, on the one hand, a Stepwise Approach (STP), and on the other hand, an approach given by variables created through Principal Component Analysis (PCA)(15). Based on this, the study concluded that using only sMRI and the STP-LDA model has a sensitivity of 64.3% and a specificity of 76.6% for early diagnosis of schizophrenia while using PCA-LDA showed a sensitivity of 67.9% and a specificity of 72.3%(15). On the other hand, using NP alone with the STP-LDA model has a sensitivity of
71.4% and a specificity of 80.9%, while PCA-LDA showed a sensitivity of 78.5% and a specificity of 91.5%(15). Finally, when combining the study by sMRI and NP associated with the STP-LDA model, the sensitivity was 64.3% and specificity
83.0%, but when associating only PCA-LDA there
was a sensitivity of 89.3% and a specificity of
93.6%.45(15)

Contribution of neuroimaging to the monitoring of Schizophrenia

Currently, the management of schizophrenic patients  is  based  on  antipsychotic  drugs,  but the use of other therapies such as cognitive training(16)  or cognitive remediation therapy(17,18), and minocycline(19), among others, have also been described. However, the aim of this systematic review is not to provide information about available treatments, but to assess how neuroimaging can contribute to the follow-up of schizophrenic patients treated with such therapies.

Recently Rhindress et al. 2017 in a randomized clinical trial involving 29 patients with the first episode  of  psychosis  treated  with  Aripiprazole or Risperidone (widely used Antipsychotics) demonstrated at the hippocampal level by MRI an increase in subiculum volume with an average difference of 32 mm3 (p = 0, 024) compared to the control group, while there was a volumetric decrease in the dentate gyrus/cornu ammonis subregion (DG/CA4) with an average difference of -68 mm3 (p = 0.016) compared to the control group.(20)

On the other hand, Morimoto et al. based their work on cognitive remediation, which is an evidence-based non-pharmacological treatment aimed at improving functions related to daily tasks,  including  those  related  to  school,  work, social interactions, and independent living. A randomized clinical trial involving a cognitive remediation therapy intervention group and a control group treated with a classical antipsychotic (Chlorpromazine) showed an increase in right hippocampal volume in the intervention group compared to the control (p<0.001), which was also associated with improvements in verbal fluency (p=0.012) and global cognitive scale scores (p=0.049)(17). In addition, a randomized clinical trial conducted by Penadés et al. that also considered cognitive remediation therapy as an intervention showed elevated fractional anisotropy at the level of the corpus and knee of the corpus callosum, and in the right posterior thalamic radiation of the treated group, which was significantly correlated with cognitive and functional improvement.(18)

In a different line of study, the tricyclic antibiotic minocycline has been reported to have a neuroprotective effect and could eventually be useful in the treatment of schizophrenia. Chaves et al. in a randomized clinical trial of schizophrenic patients, where the intervention was minocycline, concluded through the use of MRI and SPECT that, after 12 months of follow-up, the placebo group showed a significant reduction in the grey matter in the posterior medial cingulate and precentral cingulate cortex compared to the intervention group (neuroprotective effect). These changes were associated with a significant reduction in positive and negative symptoms in the treated group, and a reduction in the uptake of the radioisotope Tc-99 in frontotemporal areas, recorded by SPECT.(19)

Haut  et  al.  2010  using  a  randomized  clinical trial  were  able  to  demonstrate  that  in  patients with schizophrenia, cognitive training improves symptomatology, as well as increases the functionality of the prefrontal cortex measured by fMRI, where an increase in baseline functionality was found at the level of the left frontopolar cortex (p = 0.03 for words and p = 0.03 for words). 03 for words and p = 0.02 for images) and left dorsolateral prefrontal cortex (p = 0.04 for words and p =
0.05 for images) associated with both verbal and image working memory training. There was also increased activation of the right frontopolar cortex associated with imagery working memory training, but not with verbal working memory (p = 0.08 for words and p = 0.03 for imagery).(6)

The information presented above is summarized in
table 2.

DISCUSSION AND CONCLUSIONS

This systematic review highlights findings of neurocognitive relevance, among which the following  stand  out:  functional  and  structural alterations in the amygdala associated with negative symptoms, as well as alterations in the perfusion of the anterior cingulate gyrus and left inferior parietal cortex, together with an increase in oxyhemoglobin concentration in the ventral part of the frontopolar region during the test of fluency of ideas correlated in  a  positively  significant  way  with  the  GAF score  of  schizophrenic  patients,  hyperactivity of the DMN in patients with schizophrenia, deregulation of the histone deacetylase enzyme involved in cognitive structuring processes, etc. However, structural neuroimaging also provides neuroimaging findings of interest in the study of neuropathological  mechanisms  associated  with this psychotic disorder, mainly the morphometric reduction of the frontal region, as well as structural alterations of the amygdala associated with negative symptoms.

On   the   other   hand,   the   evidence   collected on the diagnostic utility of neuroimaging in schizophrenia suggests that, according to the study by Karageorgiou et al. 2011, maximum diagnostic accuracy is achieved by combining neuroimaging methods and clinical tests such as NP, highlighting that the latter still appears to be more accurate in the early diagnosis of schizophrenia than neuroimaging alone. However, there is evidence that in combination they can further increase the accuracy of early intervention strategies.

The study by Kraguljac et al. 2019 revealed, using Diffusion Tensor Imaging, an increased mean diffusivity in the fusiform area and the white matter of the lingual gyrus compared to controls, which would imply axonal involvement of these areas in schizophrenic patients. This could be related to the social disconnection of these patients (autism) with the consequent alteration of their social-cognitive functions.

The  usefulness  of  neuroimaging  in  the  follow- up and evaluation of the response to treatment of schizophrenia is fundamentally provided by MRI with its functional subtype. The main structural imaging findings associated with clinical changes are increased overall hippocampal volume and at the subiculum level with a volumetric reduction at the level of the DG/CA4 subregion, and preservation of the grey matter in the posterior medial cingulate and precentral cortex associated with a significant reduction in positive and negative symptoms in these patients. The main functional neuroimaging findings involve elevated fractional anisotropy at the level of the corpus and knee of the corpus callosum, as well as in the right posterior thalamic radiation which correlated significantly with a cognitive and functional improvement of the patient. All these neuroimaging findings could eventually be markers for disease follow-up.

Compared to other systematic reviews and meta- analyses, the results suggest that neuroimaging reports  structural  and  functional  differences during  treatment,  which  could  function  as markers for follow-up. Torres et al. 2013, found that schizophrenic patients on antipsychotic medication showed a volumetric reduction in MRI at the level of the left lateral temporal cortex, left inferior frontal gyrus, and right rectus gyrus, but there was an increase in volume at the level of the left dorsal anterior cingulate cortex, left ventral anterior  cingulate  cortex  and  right  putamen(21). In addition, Molent et al. 2019 evidenced the presence of alterations in activation and functional connectivity in the frontotemporal network, corticostriatal, and relevance networks, which were associated with resistance in the treatment of schizophrenia.(22)

On the other hand, other studies report concrete neuroimaging findings that contribute to the diagnosis of schizophrenia, such as Kambeitz et al. 2015, which determined that there are strong neuroimaging differences between schizophrenic subjects and healthy controls, with a sensitivity and specificity of approximately 80%. Furthermore, this group adds that there is significantly higher sensitivity in functional MRI compared to structural MRI, with higher sensitivity found in older people, and higher sensitivity and specificity in patients in the chronic stage of schizophrenia compared to patients with a first episode.(23)

Regarding   the   contribution   of   neuroimaging to the study of neuropathological mechanisms associated with schizophrenia, Haukvik et al. 2018 found significant reductions in the volume of all hippocampal subfields in schizophrenia compared to healthy controls(24). However, Kuo et al. 2019 found no significant differences in the variability of cortical and subcortical volumes in schizophrenic patients compared to controls in MRI studies, but significant differences in intracranial volumes, especially lateral and third ventricle volumes, compared to controls.(25)

Because of the above, it can be summarized that sMRI appears to be the neuroimaging test that provides the most information in the therapeutic follow-up of this pathology, and whose imaging findings correlate with changes in the patients’ symptomatology.   On   the   other   hand,   sMRI showed encouraging results for the early diagnosis of schizophrenia when combined with the clinical (mental examination). Finally, the test that offers the most imaging findings that contribute to the understanding of the underlying neuropathology of this disease seems to be fMRI.

Among the limitations of this systematic review is that, firstly, many clinical studies lacked a correlation    between    neuroimaging    findings and the symptomatology presented by patients, because some studies did not assess this parameter, especially in studies presented as evidence of usefulness  at  follow-up.  Secondly,  few  studies address   the   diagnostic   utility   and   properties that neuroimaging brings to the follow-up of schizophrenia, so the aim of this review was often a secondary outcome in the studies analyzed.

It is important to emphasize that the diagnosis of schizophrenia  continues  to  be  clinical,  however, this review provides an updated view on the findings of  neuroimaging  that  can  contribute  to the understanding of the pathological mechanisms behind this psychotic disorder and the diagnostic utility it can offer, with specific imaging findings, even to detect early alterations in at-risk populations. On the other hand, neuroimaging findings that may eventually contribute to the follow-up of this disease are evidenced. However, the major problem that currently arises with the use of these imaging tools is their high cost. Notwithstanding the above, more studies are needed to provide evidence in this regard.

REFERENCES

1. McGrath   J,   Saha   S,   Chant   D,   Welham   J.
   Schizophrenia: a concise overview of incidence, prevalence, and mortality. Epidemiol Rev [Internet]. 2008;30:67–76. Available from: http:// www.ncbi.nlm.nih.gov/pubmed/18480098
   
2. Kahn  RS,  Sommer  IE,  Murray  RM,  Meyer- Lindenberg A, Weinberger DR, Cannon TD, et al. Schizophrenia. Nat Rev Dis Prim [Internet]. 2015
   Dec 17;1(1):15067. Available from: http://www. nature.com/articles/nrdp201567
   
3. Howes   OD,   Murray   RM.   Schizophrenia:   an integrated sociodevelopmental-cognitive model. Lancet [Internet]. 2014 May;383(9929):1677–87. Available from: https://linkinghub.elsevier.com/ retrieve/pii/S014067361362036X
   
4. Haijma S V., Van Haren N, Cahn W, Koolschijn PCMP, Hulshoff Pol HE, Kahn RS. Brain Volumes in  Schizophrenia: A Meta-Analysis  in  Over  18 000  Subjects.  Schizophr  Bull  [Internet].  2013 Sep;39(5):1129–38. Available from: https:// academic.oup.com/schizophreniabulletin/article- lookup/doi/10.1093/schbul/sbs118
5. Kubota M, van Haren NEM, Haijma S V., Schnack HG, Cahn W, Hulshoff Pol HE, et al. Association of IQ Changes and Progressive Brain Changes in Patients With Schizophrenia. JAMA Psychiatry [Internet].  2015  Aug  1;72(8):803.  Available from:     http://archpsyc.jamanetwork.com/article. aspx?doi=10.1001/jamapsychiatry.2015.0712
   
6. de  Weijer  AD,  Neggers  SFW,  Diederen  KMS, Mandl RCW, Kahn RS, Hulshoff Pol HE, et al. Aberrations in the arcuate fasciculus are associated with auditory verbal hallucinations in psychotic and in non-psychotic individuals. Hum Brain Mapp [Internet]. 2011;n/a-n/a. Available from: http://doi. wiley.com/10.1002/hbm.21463
   
7. Moher  D,  Liberati  A,  Tetzlaff  J,  Altman  DG.
   Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ [Internet]. 2009 Jul 21;339(jul21 1):b2535–b2535. Available from: https://www.bmj.com/lookup/ doi/10.1136/bmj.b2535
   
8. Morgan  SE,  Seidlitz  J,  Whitaker  KJ,  Romero- Garcia R, Clifton NE, Scarpazza C, et al. Cortical patterning of abnormal morphometric similarity in psychosis is associated with brain expression of schizophrenia-related genes. Proc Natl Acad Sci [Internet]. 2019 May 7;116(19):9604–9. Available from: http://www.pnas.org/lookup/doi/10.1073/ pnas.1820754116
   
9. Rahm C, Liberg B, Reckless G, Ousdal O, Melle I, Andreassen OA, et al. Negative symptoms in schizophrenia show association with amygdala volumes and neural activation during affective processing. Acta Neuropsychiatr [Internet]. 2015
   Aug 17;27(4):213–20. Available from: https:// www.cambridge.org/core/product/identifier/ S0924270815000113/type/journal_article
   
10. Kraguljac  NV,  Anthony  T,  Skidmore  FM, Marstrander  J,  Morgan  CJ,  Reid  MA,  et  al. Micro-           and    Macrostructural    White    Matter Integrity               in     Never-Treated     and     Currently Unmedicated Patients With Schizophrenia and Effects of Short-Term Antipsychotic Treatment. Biol Psychiatry Cogn Neurosci Neuroimaging [Internet].       2019    May;4(5):462–71.    Available from: https://linkinghub.elsevier.com/retrieve/pii/ S2451902219300163
    
11. Catherine F-A, Aurélie V, Eric G, Pierre M, Raphaëlle R, Marine A, et al. Defining functioning levels      in    patients    with    schizophrenia:    A combination of a novel clustering method and brain SPECT analysis. Psychiatry Res Neuroimaging [Internet]. 2017 Dec;270:32–8. Available from: https:// linkinghub.elsevier.com/ ret ri eve/ pi i/ S092549271730094X
    
12. Guo W, Liu F, Chen J, Wu R, Li L, Zhang Z, et al.
    Hyperactivity of the default-mode network in first- episode, drug-naive schizophrenia at rest revealed by family-based case–control and traditional case– control designs. Medicine (Baltimore) [Internet].
    2017  Mar;96(13):e6223.  Available  from:  http://
    journals.lww.com/00005792-201703310-00005
    
13. Gilbert TM, Zürcher NR, Wu CJ, Bhanot A, Hightower BG, Kim M, et al. PET neuroimaging reveals histone deacetylase dysregulation in schizophrenia. J Clin Invest [Internet]. 2018 Dec
    10;129(1):364–72. Available from: https://www. jci.org/articles/view/123743
    
14. Takeshi K, Nemoto T, Fumoto M, Arita H, Mizuno M. Reduced prefrontal cortex activation during divergent thinking in schizophrenia:Amulti-channel NIRS   study.   Prog   Neuro-Psychopharmacology Biol Psychiatry [Internet]. 2010 Oct;34(7):1327– 32. Available from: https://linkinghub.elsevier. com/retrieve/pii/S0278584610002873
    
15. Karageorgiou E, Schulz SC, Gollub RL, Andreasen NC, Ho B-C, Lauriello J, et al. Neuropsychological         Testing     and     Structural Magnetic Resonance Imaging as Diagnostic Biomarkers Early in the Course of Schizophrenia and        Related     Psychoses.     Neuroinformatics [Internet]. 2011 Dec 19;9(4):321–33. Available from:                    http://link.springer.com/10.1007/s12021-
    010-9094-6
    
16. Haut KM, Lim KO, MacDonald A. Prefrontal Cortical  Changes  Following  Cognitive  Training in  Patients  with  Chronic  Schizophrenia:  Effects of Practice, Generalization, and Specificity. Neuropsychopharmacology  [Internet].  2010 Aug
    28;35(9):1850–9. Available from: http://www. nature.com/articles/npp201052
    
17. Morimoto  T,  Matsuda  Y,  Matsuoka  K,  Yasuno F,  Ikebuchi  E,  Kameda  H,  et  al.  Computer- assisted cognitive remediation therapy increases hippocampal volume in patients with schizophrenia: a randomized controlled trial. BMC Psychiatry [Internet].  2018  Dec  27;18(1):83.  Available from:           https://bmcpsychiatry.biomedcentral.com/ articles/10.1186/s12888-018-1667-1
    
18. Penadés  R,  Pujol  N,  Catalán  R,  Massana  G, Rametti G, García-Rizo C, et al. Brain Effects of Cognitive Remediation Therapy in Schizophrenia: A Structural and Functional Neuroimaging Study. Biol Psychiatry [Internet]. 2013 May;73(10):1015–
    23. Available from: https://linkinghub.elsevier. com/retrieve/pii/S0006322313000620
    
19. Chaves C, Marque CR, Maia-de-Oliveira JP, Wichert-Ana L, Ferrari TB, Santos AC, et al. Effects of minocycline add-on treatment on brain morphometry and cerebral perfusion in recent- onset  schizophrenia.  Schizophr  Res  [Internet].
    2015 Feb;161(2–3):439–45. Available from: https:// linkinghub.elsevier.com/ ret ri eve/ pi i/ S0920996414007087
    
20. Rhindress K, Robinson DG, Gallego JA, Wellington R, Malhotra AK, Szeszko PR. Hippocampal subregion volume changes associated with antipsychotic treatment in first-episode psychosis. Psychol Med [Internet].  2017  Jul  14;47(10):1706–18. Available from:         https://www.cambridge.org/core/product/
    identifier/S0033291717000137/type/journal_article
    
21. Torres   US,   Portela-Oliveira   E,   Borgwardt   S, Busatto GF. Structural brain changes associated with   antipsychotic   treatment   in   schizophrenia as      revealed    by    voxel-based    morphometric MRI: an activation likelihood estimation meta- analysis.  BMC  Psychiatry  [Internet].  2013  Dec 20;13(1):342. Available from: http://bmcpsychiatry. biomedcentral.com/articles/10.1186/1471-
    244X-13-342
    
22. Molent C, Olivo D, Wolf RC, Balestrieri M, Sambataro        F.    Functional    neuroimaging    in treatment resistant schizophrenia: A systematic review.    Neurosci    Biobehav    Rev    [Internet].
    2019       Sep;104:178–90.       Available       from: https:// linkinghub.elsevier.com/ ret ri eve/ pi i/ S0149763418309485
    
23. Kambeitz  J,  Kambeitz-Ilankovic  L,  Leucht  S,Wood S, Davatzikos C, Malchow B, et al. Detecting Neuroimaging Biomarkers for Schizophrenia: A Meta-Analysis of Multivariate Pattern Recognition Studies.    Neuropsychopharmacology    [Internet].
    2015 Jun 20;40(7):1742–51. Available from: http://
    www.nature.com/articles/npp201522
    
24. Haukvik UK, Tamnes CK, Söderman E, Agartz I.    Neuroimaging    hippocampal    subfields    in schizophrenia and bipolar disorder: A systematic review and meta-analysis. J Psychiatr Res [Internet].     2018    Sep;104:217–26.    Available from: https://linkinghub.elsevier.com/retrieve/pii/ S0022395618303820
    
25. Kuo SS, Pogue-Geile MF. Variation in fourteen brain structure volumes in schizophrenia: A comprehensive meta-analysis of 246 studies. Neurosci      Biobehav     Rev     [Internet].     2019 Mar;98:85–94. Available from: https://linkinghub. elsevier.com/retrieve/pii/S0149763418302963