Tools and resources for neuroanatomy education: a systematic review

The aim of this review was to identify studies exploring neuroanatomy teaching tools and their impact in learning, as a basis towards the implementation of a neuroanatomy program in the context of a curricular reform in medical education. Computer-assisted searches were conducted through March 2017 in the PubMed, Web of Science, Medline, Current Contents Connect, KCI and Scielo Citation Index databases. Four sets of keywords were used, combining “neuroanatomy” with “education”, “teaching”, “learning” and “student*”. Studies were reviewed independently by two readers, and data collected were confirmed by a third reader. Of the 214 studies identified, 29 studies reported data on the impact of using specific neuroanatomy teaching tools. Most of them (83%) were published in the last 8 years and were conducted in the United States of America (65.52%). Regarding the participants, medical students were the most studied sample (37.93%) and the majority of the studies (65.52%) had less than 100 participants. Approximately half of the studies included in this review used digital teaching tools (e.g., 3D computer neuroanatomy models), whereas the remaining used non-digital learning tools (e.g., 3D physical models). Our work highlight the progressive interest in the study of neuroanatomy teaching tools over the last years, as evidenced from the number of publications and highlight the need to consider new tools, coping with technological development in medical education.


Background
Among the basic sciences providing relevant medical awareness, human anatomy, which includes gross and neuroanatomy, has historically been considered a key science educational area in medical education [1,2].
The first descriptions of human anatomy teaching in Europe dates back to Greece, in third century BC, with the introduction of systemic human cadaveric dissection. Although the practice of human dissection was prohibited during the Middle Ages due to religious and popular beliefs, it revival at the beginning of fourteenth century and becomes the core basis in medical education and anatomy teaching until the twentieth century [3,4]. By that time, significant changes have occurred in undergraduate medical education, on one hand because of the introduction of new subjects into curricular programmes as medical scientific knowledge increases and on the other hand because of the move towards skillsbased teaching to face clinical practice [5][6][7][8]. Within this new reality, many preclinical medical curricula started to integrate systems-based units, abandoning the traditional, isolated, discipline-based curricular approaches [9][10][11][12][13][14].
These changing concepts greatly influenced the modern teaching of medical anatomy, with many schools now delivering anatomy using integrated, clinicallyoriented modules, with considerably less time allocated to anatomy [15][16][17]. For example, within the USA contact hours for gross anatomy has fallen from an average of 170 h in 2002 to~150 h in 2012 and in neuroanatomy contact hours decreased from 95 to 83 h from 2002 to 2012 [18]. This general reduction in time dedicated to anatomy teaching at medical schools, associated with the increased demand for clinical importance of the topics covered in anatomy curricula, have led to a redefinition of program content and students' learning objectives, accompanied by the introduction of innovative teaching and learning approaches. Despite the long history, the role of cadaveric dissection, as the primary tool for anatomical teaching, has been reduced or replaced in most medical schools by prosection, use of plastic models and/or multimedia-based learning packages [19].
Although initially integrated with the teaching of gross anatomy, neuroanatomy can now be found as a stand-alone course or, most frequently, as an integrated part of the systems-based approach, taught alongside other neurosciences. Teaching of neuroanatomy to students is known to be particularly challenging, due to the sheer complexity and interconnectedness of the central nervous system [20]. Students are required to learn not only anatomical structures, but also be able to understand their topography, spatial relationships and clinical significance. In 1994, Jozefowicz [21] introduced the term "neurophobia" as "a fear of the neural sciences and clinical neurology that is due to the students' inability to apply their knowledge of basic sciences to clinical situations". In fact, poor teaching and the challenging nature of aspects of neuroanatomy were identified, in one study, as reasons for considering neurosciences/ neurology so difficult. To face changes in medical education curricula and to help reduce neurophobia, some anatomists have developed and implemented innovative teaching techniques and strategies. In this context, Moxham et al. [22] also proposed a core syllabus for teaching neuroanatomy to medical students, to provide guidelines concerning neuroanatomical knowledge. However, the debate over how best to teach neuroanatomy in undergraduate medical education continues, with each institution using its own method.
The major aim of the present work is to review the most common methods for teaching neuroanatomy, and their effectiveness. More specifically, we intend to: a) identify the studies that explore neuroanatomy teaching tools; and b) to assess their impact on learning.

Databases searched and search terms
The electronic databases searched in this review included those identified as the most relevant to the topic. More specifically, computer-assisted searches were conducted in six online databases: PubMed, Web of Science, Medline, Current Contents Connect, KCI and Scielo Citation Index. As keywords, four sets were used, combining "neuroanatomy" with "education", "teaching", "learning" and "student*".

Inclusion and exclusion criteria
The search was restricted to English-language studies that focus on the teaching of neuroanatomy. A comprehensive search was performed for papers available for search from each database's inception through March 2017. Papers available online ahead of the print version were also analyzed. Manuscripts were included if they were original research studies assessing the impact of using a specific method on student's learning of neuroanatomy.
The exclusion criteria were as follows: i) descriptive studies on the use of a teaching method without assessing the impact on learning; ii) studies describing the development of a teaching method; iii) studies not focused on the teaching of human neuroanatomy; iv) studies in languages other than English; v) reviews, editorial material, proceeding papers, notes, letters to the Editors and meeting abstracts; and vi) duplicate papers.

Selection of papers
All databased were reviewed independently by two readers (M.A. and J.A.) using the above stated criteria. More specifically, each manuscript identified was placed on an Excel spreadsheet, and the readers applied the exclusion criteria independently. Disagreements were discussed in a meeting and resolved by consensus. After removal of duplicate manuscripts, all potentially eligible manuscripts were screened by both readers. Then, the full text of all screened manuscripts was carefully read. All data collected was confirmed by a third reader (M.F.), and discussions occurred until a final consensus was reached.
Charting collating and summarizing the data Spreadsheets were used to register the most important features of each study, namely the title of the papers, authors, year of publication, university and country where the study was conducted, type and number of participants, teaching tool, aim, methodology, number of participants, and main results/conclusions. Data were summarized, and were then grouped according to these features.

Studies included in this review
The search of PubMed, Web of Science, Medline, Current Contents Connect, KCI and Scielo Citation Index databased yielded 214 manuscripts. After removal of duplicate studies (n = 92), a total of 122 manuscripts were identified. On applying the inclusion and exclusion criteria by the two independent readers, 53 manuscripts were excluded because they were written in languages other than English, were abstracts, letters to the Editors, editorial material, proceeding papers, or notes. Therefore, a total of 69 manuscripts were then assessed for eligibility. After these manuscripts were read in their entirety, 40 studies were excluded because they were descriptive reports of a teaching method without assessing its effectiveness, they described the development of a new teaching tool, did not focused on the learning of neuroanatomy (e.g., neuroanatomy of schizophrenic patients) or were review manuscripts. Thus, a total of 29 manuscripts meet the criteria to be included in this review (see Fig. 1). Table 1 presents some of the features of the papers included in the present work. Of the 29 studies, the first study [36] was published 50 years ago, in 1966. However, the majority of the studies (n = 15; 52%) were published in 2012-2016, and 24 (83%) of the studies found were published in the last 8 years. Only 4 (14%) were published before 2005. Most studies were conducted in the United States of America (n = 19; 65.52%), followed by the United Kingdom (n = 4; 13.79%) and Australia (n = 2; 6.90%). The remaining four studies were from Canada, India, Poland and Spain.
The teaching methods used in the studies included in this review can be classified into digital tools (n = 13; 46. 43%) and non-digital learning tools (n = 15; 53.57%). The digital tools include 3D computer neuroanatomy models, computer-based tools (i.e., computer-aided instruction/ learning), and apps installed in tablets. The non-digital tools include the use of case studies, equivalence-based instruction (EBI), 3D physical models, face-to-face teaching, flipped classroom, inquiry-based laboratory instruction, intensive mode of delivery, interpolation of questions, near-peer teaching, Renaissance artists' depictions, selfinstructional stations, and truncated lectures, conceptual exercises and manipulatives.

Digital tools
Computer-based neuroanatomy tools Table 2 summarizes each study included in this review, including the teaching tool used, aims, methodology employed, and main results/conclusion. Some researchers [23][24][25][26][27][28] focused their studies on the impact of using computer-based tools for teaching neuroanatomy. More specifically, McKeough et al. [23,24] investigated the effect of a computer-based tool on students' performance and their attitudes. Before and after test questions revealed that scores improved significantly after working with the learning model. In addition, students reported the computer-aided neuroanatomy learning modules as a valuable and enjoyable learning tool, and perceived their clinical-self efficacy as higher as a result of working with them.
Foreman et al. [25] conducted a prospective evaluation by asking questions to the students regarding tool     Students were expected to collaborate with their colleges in small groups. In each case study, the entire class was asked to participate in some aspect of the case.
Students report enjoying learning brain structure using this method, and commented positively on the class activities associated with learning brain anatomy.
Using equivalence-based instruction to increase efficiency in teaching neuroanatomy XV    To investigate the effect of interpolation of questions into a lecture presentation.
Participants were divided into 2 matched groups.
Each group was shown lectures on the visual system which differed only in regard to taxomic level of 10 questions posed lecture. Performance was then measured (betweensubject design).
Emphasis on recall questions aids performance on subsequent recall questions more than emphasis on problemsolving questions helps in the solution of subsequent problem-solving questions.
navigation and benefits, clarity of the images, and by requesting them to compare these tools to traditional ones. Results showed that most of the students agreed that the computer-based tool was easy to navigate and overall beneficial, educational in structure identification and had clear images, and somewhat better than traditional learning tools. Only one study [26] assessed the perception of both students and faculty members with a computer-based tool. Analysis of the results supported the research hypotheses that the prototype was well-designed for different types of users in various educational contexts, and that would be useful as a neuroanatomy review tool for health-professions students.
The studies mentioned above suggested that computerbased tools seem to be effective in teaching neuroanatomy. Lamperti and Sodicoff [27] compared students' performance between those classes that previously had the traditional anatomy laboratory with two succeeding classes that used the computer-based laboratory. When assessing the total performances, results showed no statistically significant differences in the average grades for classes during the 2 years prior to and 2 years following the introduction of computers in the course. However, Svirko and Mellanby [28] compared the students' approach to learning for a computer-based course in Neuroanatomy with that for their studies in general. Students reported lower deep approach scores (seeking the meaning of the information being taught) and higher surface approach scores (rote-learning motivated by fear of failure and without integrating current and previous knowledge) for the computer-aided course than for their studies in general. Also, only approximately one quarter of the students agreed or strongly agreed they enjoyed this course.

3D computer neuroanatomy tools
More recently, studies have used three-dimensional (3D) computer graphical models of human brain as a teaching tool in neuroanatomy classes [29][30][31][32][33][34]. For example, Drapkin et al. [29] compared students' performance when learning through a new 3D program or through traditional methods. They divided the students into two groups: an experimental (3D program) and a control group (traditional). Results showed that scores extracted from questions involving C-shaped internal brain structures were higher for the experimental group, and that these students reported higher confidence levels. Allen et al. [30] also divided the students into two groups, but each group was exposed to two types of teaching resources, presented in a contrabalanced order: 3D new learning module and cadaveric laboratory session. After accessing each teaching resource, participants completed a test. Findings showed that participants who initially learned using the 3D learning module scored significantly better than students who learned using the gross anatomy resources. In addition, scores significantly improved for students who accessed the 3D learning module following exposure to the cadaveric resources. Palomera et al. [31] assessed if students' evaluation of a new 3D computer-based tool depended on their visuospatial skills to establish handle spatial relationships. Findings revealed that students with both high visuospatial ability and low visuospatial ability assigned similar high educational value to this tool.
Naaz et al. [32], Pani et al. [33] and Chariker et al. [34] focused on the learning of whole and sectional neuroanatomy using neuroanatomical 3D computer models. Their findings suggested that: i) explicit graphical demonstration of the spatial relations between 3D whole anatomy and 2D sectional anatomy leads to high longterm retention of sectional neuroanatomy [32]; ii) an integrative learning method, that presents whole and sectional neuroanatomy in alternating trials, increases the students' performance [33]; and iii) instruction of neuroanatomy designed on the basis of substantial transfer of learning from whole to sectional anatomy is an effective method for teaching neuroanatomical structures [34].

Other digital tools
Only one study used tablet devices (Apple iPads) in neuroanatomy practical sessions to investigate the effectiveness of specific apps in students' perceptions and performance [35]. Results showed that the students considered the apps to be beneficial for learning. In addition, their performance in neuroanatomy-related questions increased after the introduction of the tablet devices.

Non-digital tools
Regarding the use of non-digital tools, there is a variety of resources that can be used in neuroanatomy classes. In their pioneer work, published in 1966, Geeartsma and Matzke [36] investigated the effect of interpolation of questions into a lecture presentation. Findings revealed that the emphasis on recall questions led to an increase in students' performance on subsequent recall test questions.
Krontiris-Litowitz [37] studied a revised curriculum using truncated lectures, conceptual exercises, manipulatives, and that was shorter. She found that students' learning was more effective under this new curriculum. Whillier & Lystad [38] also compared units of neuroanatomy for undergraduate students: an intensive mode and a traditional mode. Even though students showed similar levels of satisfaction, grades were lower in the new intense mode. In addition, Whillier & Lystad [39] compared two other units of neuroanatomyan old and a restructured unit. Results showed that the increase in total face-to-face teaching hours in the restructured unit led to an increase in the students' satisfaction. However, it does not improve their grades.
Sheldon [40], Kennedy [41] and Greenwald and Quitadamo [42] included several case studies during classes and students were expected to collaborate and participate in their discussions. Results showed that students evaluated this method as enjoyable, helpful for remembering or learning the material, commented positively on the class activities and gained more national percentile ranks than students in a conventional neuroanatomy course. Watson [43] used interactive classroom exercises using well-known Renaissance artists' depictions of the brain, and found that these exercises increase the interest of the students in neuroanatomy. Veeramani et al. [44] investigated the impact of another method that also aims to increase students' collaboration and participation in class: a flipped classroom approach. In this method, students are expected to attend the class with basic understanding of the subject to be able to participate and engage in discussions. Findings revealed that most of the participants felt that the work-sheet with questions provided before class allowed them to adquired a deeper understanding of the subject and believed that the resources provided increased their interest to read.
Fisher et al. [45] used neuroanatomy self-instruction laboratory stations to present neuroanatomical laboratory material making the students' more active in their learning process. Test results indicated a mastery of station material and positive students' attitudes. Gardner et al. [46] exposed students to novel research projets into their laboratory experience. Findings showed that working within the context of a research question of a member of the faculty increase students' motivation and excitement, and encouraged good scientific practice. Hall et al. [47] developed and delivered a near-peer programme of study, in which two medical students delivered the teaching to their colleges, aiming them to grow through their similar knowledge base and shared experiences. After a series of seven sessions, students perceived their level of knowledge as being higher.
Pytte and Fienup [48] and Greville et al. [49] used equivalence-based instruction (EBI) as a tool for teaching neuroanatomy to undergraduate studentsteaching how physically disparate stimuli are functionally equivalent, or interchangeable. Findings suggested that: i) selection of associations by the teacher can led to the spontaneous emergences of novel associations within a concept or category; and ii) EBI is a useful tool to teach students to read an MRI of the brain and speciflly useful for teaching C-shaped internal brain structures.
Finaly, even though the use of 3D computer models to teach neuroanatomy has been increasing since 2011, 1 year before Estevez et al. [50] developed and assessed a 3D physical tool. Whereas the control group was exposed to 2D brain cross-sections, the experimental group constructed 3D color-coded physical models. Test results showed that the overall quiz scores for the experimental group were significantly higher than the control group. However, only the scores for questions requiring 3D visualization were significantly higher in the experimental group.

Cross-cultural comparisons
Zurada et al. [51] investigated similarities and differences in study methods among American, Asian, and European medical students. For that, the authors asked participants to fill a questionnaire reporting the study methods they use to study, and which methods they believed were the most effective in terms of comprehension, memorization, and review. Results revealed differences in the study techniques among students from the different countries. For example, Polish and American tended to prefer the use of dissection and prosected specimens compared to the Taiwanese students.

Discussion
From searching PubMed, Web of Science, Medline, Current Contents Connect, KCI and Scielo Citation Index databases, 29 studies were identified. Even though the first study was published 50 years ago, more than four-fifths of the studies were published in the last 8 years, evidencing a growing awareness of this thematic over the most recent years. In fact, several modifications in anatomy and, in particular, in neuroanatomy education have been made over the last few decades, and numerous strategies have been recognized to increase the performance of students [52]. The studies emerged from different countries, including United States of America, United Kingdom, Australia, Canada, India, Poland and Spain. This indicates that the interest for the teaching of neuroanatomy is cross-cultural. However, the paper published by Zurada et al. [51] showed that there are some differences in the study methods adopted by medical students from different countries, such as American, Asian, and European.
Concerning the type of participants, medical students are the most studied sample, which is not surprising taking in consideration that, although several changes have occurred in medical curricula worldwide, the anatomic background is still considered a keystone for approaching clinical medicine [53]. Regarding the number of participants, approximately two-thirds used a sample of 100 participants or less, leading to a low average statistical power. This is a concern regarding published literature in this area, as many of the studies found in this review have limitations imposed by sample size. For example, Gardner et al. [46] investigated only 13 students, Krontiris-Litowitz [37] 19 students, Watson [43] 27, and Sheldon [40] 28 students. A small statistical power is known to reduce chance of detecting a true effect, as well as to reduce the likelihood that a statistically significant result reflects a true effect [54].
In terms of teaching methods used in the studies included in this review, almost half of them used digital tools, such as computer-based tools, 3D computer neuroanatomy models and apps installed in tablets. The majority of the six studies that focused on computerbased neuroanatomy tools showed that: i) it is welldesigned for both students and faculty members; ii) the performance of the students increased after working with the learning model; and iii) the students had positive attitudes towards these tools. However, the two remaining studies found that there were no statistically significant differences in the average grades of the students after the introduction of computers in the course, and that students reported lower deep approach scores for the computer-aided course than for their studies in general. These results suggest that even though the computer-based tools seem to be effective in teaching neuroanatomy in certain contexts, this assumption cannot be generalized without further research.
Since 2014, half of the studies focused on 3D computer-based tools, highlighting a growing interest in exploring 3D models on learning of neuroanatomy. Overall, these studies revealed that this digital tool is an effective method for teaching neuroanatomical structures. Findings also showed that students assigned a high educational value to this tool. These results are somewhat inconsistent with those from Azer and Azer [55] who concluded in their review that there is no evidence that the use of 3D models is superior to traditional tools for teaching anatomy. It is possible that the structure of the brain have some particularities that require more the use of the students' visual-spatial abilities than other anatomical structures of the body. Therefore, the use of 3D tools, by facilitating the mental rotation and manipulation of the brain structures, may facilitate the learning of neuroanatomy.
The non-digital tools include a variety of resources used in neuroanatomy classes.
Findings revealed that the following strategies led to an increase in students' performance and positive attitudes: i) emphasis on recall questions; ii) use of case studies; iii) inclusion during class of truncated lectures, conceptual exercises, and manipulatives; iv) practice of exercises using well-known Renaissance artists' depictions of the brain; v) adoption of the flipped classroom approach; vi) use of neuroanatomy self-instruction laboratory stations; vii) inclusion of novel research projets into the laboratory experience; viii) near-peer programmes; ix) EBI; and x) 3D physical models. The increase in total face-to-face teaching opportunities was shown to increase students' satisfaction but not their grades, and teaching neuroanatomy in an intense mode was shown to lower students' grades compared to a traditional mode.

Limitations
Even though a rigorous approach was adopted to undergo this systematic literature review, our study presents some limitations. First, we restricted our search to six databases: PubMed, Web of Science, Medline, Current Contents Connect, KCI and Scielo Citation Index databases. Thus, it is possible that some studies addressing our aim could be found if searches in other databases were conducted. Second, in our search, four sets of keywords were used, combining "neuroanatomy" with "education", "teaching", "learning" and "student*". It is also possible that some studies may focus on neuroanatomy teaching tools but use other terminology to describe them. Third, this review included only papers written in English, and therefore 12 out of 117 studies were eliminated. Some of those papers written in languages other than English may address the aim of our study, and we did not consider them. Fourth, even though all studies were carefully reviewed independently by two readers, and all data collected was confirmed by a third reader, data may been biased by the subjectivity of the readers.

Conclusions
Our work highlights the progressive interest in the study of neuroanatomy teaching tools over the last 8 years, as evidenced from the number of publications. The view of the different strategies to teach neuroanatomy, may provide guidelines for curricular improvements in this complex area of medical education. Ethics approval and consent to participate Not applicable.

Competing interests
The authors declare that they have no competing interests.