Interpretation of three-dimensional (3D) anatomical information has always been an integral part of both medical and veterinary health care and has become even more important with the rapid advancements in medical imaging (Marks 2000; Estevez et al., 2010). Classical anatomical education utilizes a combination of teaching modalities, with cadaveric dissection considered the gold standard for learning anatomical spatial relationships due to its engagement of multiple senses, 3D interaction and tactile manipulation of tissues (Sugand et al., 2010; DeHoff et al., 2011). However with the availability of cadavers in decline, as well as the moral and ethical issues surrounding their continued use, many institutions are introducing alternative methods for teaching anatomy (Parker, 2002; Salazar 2002; Nicholson et al., 2006; Kinnison et al., 2009; Sugand et al., 2010).
In circumstances where cadaveric dissections are limited, students often depend heavily on written and verbal explanations, as well as 2D visual representations. Such teaching has been strongly associated with a detrimental increase in cognitive load leading to a decrease in knowledge acquisition and retention (Khalil et al., 2005a, 2005b, 2010, 2005c, 2005d). Computer-aided learning and 3D computer models have been developed with the aim to decrease this cognitive load (Nicholson et al., 2006; Levinson et al., 2007; Nguyen and Wilson, 2009). Although generally considered more enjoyable and stimulating to the technologically-minded modern student, 3D computer models have been shown to rely heavily on the user's own innate spatial capabilities (Guillot et al., 2007; Estevez et al., 2010) and have actually been associated with increased cognitive load in students with poor spatial skills (Huk, 2006; Levinson et al., 2007; Khalil et al., 2010). Despite this mixed evidence for their efficacy in reducing cognitive load, 3D computer models can be useful for illustrating anatomy, where cadaveric dissection is considered an ineffective teaching tool due to the inaccessibility, inherent complexity and/or the size of the structure. Examples in human anatomy teaching include the ear (Nicholson et al., 2006), the cerebral ventricular system (Adams and Wilson, 2011), the brachial plexus (Brenton et al., 2007) and the female pelvic region (Sergovich et al., 2010). One example of such a region in veterinary science is the equine foot, which is of huge clinical importance with the majority of lameness attributable to foot pathology (Dyson and Marks, 2003; Turner, 2003). Dissection of the equine foot is extremely challenging and largely beyond the capabilities of undergraduate students due to the complex spatial relationships of its structures and presence of a rigid horn capsule. The use of prosections of this area is also somewhat limited due to the small size of some structures and the limited viewing access.
Tactile manipulation and the engagement of multiple senses is considered one of the greatest advantages of cadaver dissection, and is believed to promote better understanding and retention of spatial information and relationships (Rizzolo and Stewart, 2006; DeHoff et al., 2011). Computer models do not offer these benefits but physical models offer promise, with one study reporting an increase in student comprehension and understanding of the spatial relationships between atoms when using physical models (Wu, 2004). Several studies have attempted to encompass this principle within anatomy teaching, leading to the development of innovative, hands-on alternatives to dissection, including plastination (von Hagens et al., 1987; Dhingra et al., 2006), clay modeling (Krontiris-Litowitz, 2003; Oh et al., 2009, Estevez et al., 2010;) and body painting (McMenamin, 2008).
The benefits of clay modeling in anatomical education are well documented (Krontiris-Litowitz, 2003; Motoike et al., 2009; Oh et al., 2009; Estevez et al., 2010; DeHoff et al., 2011). Students utilizing clay modeling had better understanding of the complex 3D anatomy and principles of the periventricular structures compared with students using traditional teaching methods (Estevez et al,. 2010). Hand held “manipulatives” have also been shown to improve undergraduate student critical thinking with regards to spinal tract principles and functions (Krontiris-Litowitz, 2003) and to improve assessment scores of students learning human musculature (Waters et al., 2005; Motoike et al., 2009). Clay modeling has also been shown to enhance cross sectional anatomy knowledge among medical undergraduate students, assessed by using computed tomography images (Oh et al., 2009). Despite being an effective learning resource, clay modeling may prove less effective when applied to more intricate and complex tissues such as the equine foot. Nevertheless it illustrates that physical models have considerable potential as learning tools in anatomy.
With equine foot anatomy considered one of the most complex and closely compartmentalized anatomical structures in veterinary anatomy and at the same time one of the clinically most important structures, the aim of this study was to develop an anatomically accurate physical model of the equine foot and assess its value as a 3D learning resource.
We hypothesized that a physical model would significantly enhance students' visuospatial appreciation of the complex anatomy of the equine foot, and that students utilizing the physical model would have superior ability to apply their 3D anatomical knowledge compared with students using anatomy textbooks or a 3D computer model. We also hypothesized that the use of a physical model would be more enjoyable, stimulating and enhance student confidence levels compared to textbooks or 3D computer model use.
MATERIALS AND METHODS
Development of the Equine Foot Model
The basis of the equine foot model used in this study were a high-resolution magnetic resonance images (MRI) (T1 weighted, 3D Fast Spin Echo) of a fresh cadaver horse foot obtained with a high-field scanner (1.5T, Intera, Philips, Reigate, UK). Virtual 3D reconstruction of the equine foot and its structures using “Mimics” (Materialise Corp, Leuven, Belgium), a dedicated medical imaging software program were produced. Twenty-two structures including bones, hoof cartilages, tendons, ligaments, digital cushion and hoof capsule were segmented. Automated segmentation of MRI images is challenging, therefore in this study we utilized a semiautomated technique using gray-scale thresholding followed by manual editing. The segmented structures were rendered in 3D using polynomial surface wrapping. The virtual reconstructions were rapid-prototyped to produce a physical model with the different anatomical structures printed in plastic material of various properties (Industrial Plastic Fabrications, Nazeing, UK). Bones and firmer structures were reproduced in a rigid material, while tendons and ligaments were printed in a more flexible material. Small neodymium magnets (e-Magnets UK, Sheffield, UK) were used to attach the different structures, enabling the user to repeatedly re-build the model, effectively creating a 3D “jigsaw puzzle” (Fig. 1). A series of brief instructional videos were created to accompany the model. These videos were designed as a technical aid only, replicating figure legends and labels found in textbooks and computer-based programs.
Students and Educational Context
Subjects were recruited from the third year of the Bachelor of Veterinary Medicine course at the authors' institution. At this stage of their studies, students have completed basic anatomical teaching including a one hour lecture on the horse's foot, and 7 hr of equine-specific limb dissections and have not yet started their extramural clinical training. While the curriculum is highly integrated, meaning some clinical material is introduced early in the course, students are not formally exposed to MRI of the equine foot during years 1 and 2. Given that all students entering the third year of the BVetMed course have undertaken the same rigorous anatomy assessment, it was assumed that students sampled for this study held a similar anatomical knowledge.
Validation of the Equine Foot Model
The efficacy of the equine foot model as a learning resource was evaluated by comparing it with two commonly used teaching methods: textbook and 3D computer model, with students' visuospatial anatomical knowledge assessed through identifying anatomical structures on MRI of the equine foot. Student confidence was evaluated at the beginning and at the end of the teaching session with two questionnaires. Student enjoyment of the teaching session was evaluated as part of the second questionnaire. The study was approved by the institution's ethics and welfare committee.
All 234 third year students were asked to take part on a voluntary basis via mass e-mail and intranet message. A total of 64 students were recruited to the study, representing 27.4% of the 3rd year class, and were randomly assigned to one of three teaching aid groups. Gender distribution was 89% females and 11% males. The Model group (n = 21, males = 3, females = 18) had access to the physical equine foot model. The textbook group (n = 19, males = 3, females = 16) had access to the four most commonly withdrawn veterinary anatomy textbooks from the authors' institution: König and Liebich (2007); Dyce et al., (2002); Ashdown and Done (2011); Pasquini et al., (1989). The Glasshorse group (n = 24, males = 2, females = 22) had full access to the The Glasshorse Distal Limb 3D computer model (The Glasshorse Project, Athens, GA). This is an interactive model with informative narrations and animations that can be manipulated in 3D by the user.
Anatomical Knowledge and Understanding
A testing exercise using MRI images of the equine foot was designed to objectively assess the effectiveness of the three teaching aids in facilitating correct identification of anatomical structures. MRI is the “gold standard” for the diagnosis of foot pathology in the horse; hence a thorough understanding of MRI foot anatomy is essential for equine clinicians. The testing exercise consisted of five MRI images depicting the cross sectional anatomy of the equine foot in various orientations; sagittal, frontal and three transverse planes at the level of the middle phalanx (transverse-MP), the navicular bone (transverse-NB), and distal phalanx (transverse-DP). Pilot images were made available for each of the MRI slices. A pilot image is one that depicts the same anatomy but in a different plane, giving the viewer a reference point from which the original MRI image has been taken. Students had access to their respective teaching aids throughout the duration of the assessment and were required to shade various structures (bones, joints, tendons, ligaments, cartilages, digital cushion) in specified colors on the MRI. Four bones, two tendons, five ligaments, and various other miscellaneous integuments were part of the assessment with some structures shaded on multiple MRI depicting the structures in varying orientations. The maximum score achievable across the five MRI was 28. The assessment took place under examination conditions with no access to information outside their respective teaching aids. Images were reviewed for correct identification of structures by one of the authors who had no knowledge which group the participants were allocated to using a template agreed upon by experts in the field of equine diagnostic imaging. After completion of the experiment, verbal feedback was provided for interested students by one of the authors and the students were informed that they are welcome to contact any of the authors for more feedback at any time.
Student Confidence and Enjoyment
Before the testing session student confidence was evaluated with a questionnaire comprising three positive statements; ‘I am confident in my ability to identify the anatomical structures within the equine foot’, ‘I am confident in my ability to mentally visualize the anatomy of the equine foot in 3D’, ‘I am confident in my ability to identify the anatomy of the equine foot on MRI’. Student response was recorded for each statement using a five-point Likert scale (strongly disagree, disagree, indifferent, agree, and strongly agree). Students were also asked their preferences with regards to learning anatomy and stated which resource (cadaver dissection, histology specimens, textbooks, lectures, 3D computer models) they considered most useful for learning anatomy in general.
A second questionnaire was designed to assess changes in confidence levels following the teaching exercise and to gather feedback regarding enjoyment of the three teaching aids. Students responded to the same three positive statements given prior to the teaching exercise, as well as statements relating to how helpful and enjoyable they considered their respective teaching aids, and if they would consider using the teaching aid in the future. All participants completed the questionnaires and the testing exercise within 45–60 min.
The data distribution was assessed using a Kolmogorov-Smirnov test. Data were found to be not normally distributed. Differences between the three teaching aid groups were examined using a Kruskal-Wallis test and between pairs of groups a Mann-Whitney U test was used. Examination of the difference between confidence levels before and after the testing exercise was conducted using a Wilcoxon Signed Rank test. Quantitative analysis was conducted using SPSS, version 19 (IBM, Chicago, IL), and P-value was set at 0.05.
Development of the Equine Foot Model
Development took ∼280 hr with the majority of this time spent segmenting and volume rendering the individual structures. Twenty-two structures were segmented in total, three bones, two tendons, the hoof capsule, two hoof cartilages, the digital cushion, and the rest ligaments. Rapid-prototyping took five days at a total cost of £1,145, the cost for magnets was £25, the cost for the MRI scan £250, which brings the total development cost to £1,420.
Validation of the Equine Foot Model
Anatomical knowledge and understanding
The percentage of correct answers overall was significantly higher in the Model group than that of both the Textbook group (P < 0.001) and the Glasshorse group (P < 0.001); 86.4% compared with 62.6% and 63.7%, respectively (Fig. 2). There was no significant difference in the overall test scores between the Textbook and the Glasshorse group (P = 0.685). When the questions were divided into those requiring the identification of bones, tendons, and ligaments (Fig. 2), the Model group scored significantly higher than the Textbook and Glasshorse groups for both tendons and ligaments (P < 0.001) with no significant difference between groups for identification of bones (P = 0.351). Tendon (P = 0.748) and ligament (P = 0.218) scores for the Textbook and Glasshorse groups did not differ significantly.
Comparison of scores for the different MRI planes within teaching aid groups showed all three groups to have significantly higher scores in the sagittal plane compared with all other planes (Fig. 3) (P < 0.05). There was no significant difference between the three teaching aid groups for the sagittal (P = 0.159) and transverse-MP (P = 0.143) planes (Fig. 3). The Model group scored significantly higher than the Textbook and Glasshorse groups for both the transverse-NB plane (P < 0.001) and the transverse-DP plane (P < 0.001), with no significant difference between the Textbook and Glasshorse groups for either plane (P = 0.320 and P = 0.259). Both the Model and Textbook groups scored significantly higher compared with the Glasshorse group for the frontal plane (P < 0.001 and P = 0.007 respectively) with no significant difference between the Model and Textbook groups (P = 0.205).
The spread of the data was much narrower for the Model group compared to the other groups for all structures and views (Figs. 2 and 3).
Student confidence and enjoyment
Student responses to three positive statements relating to their confidence levels prior to the testing session showed no statistically significant differences between the three groups (P = 0.828, P = 0.612, and P = 0.081 respectively). 23.4% of the students (strongly) agreed that they were confident in “identifying the anatomical structures within the equine foot,” 7.2% (strongly) agreed that they “felt confident in visualizing the anatomy of the foot in 3D” and 3.2% of students (strongly) agreed that they felt confident in “identifying the anatomical structures of the foot on MRI images.”
Following the MRI teaching exercise confidence levels with regards to “identifying the anatomical structures within the equine foot” significantly increased in the Model (P = 0.001) and Glasshorse (P = 0.026) groups with 80.9% and 58.3% of students respectively, agreeing or strongly agreeing with the statement. Both Model and Glasshorse group confidence levels were significantly higher than those of the Textbook group (P = 0.001 and P = 0.033, respectively), with the Textbook group showing no significant change in confidence levels following the MRI teaching exercise with only 36.8% of students agreeing or strongly agreeing with the statement (P = 0.971). The confidence levels were also significantly higher in the Model group compared with the Glasshorse group (P = 0.018); however, the degree of increase in confidence did not statistically differ between the two groups (P = 0.194).
Confidence levels with regards to “mentally visualizing the anatomy of the equine foot in 3D” significantly increased in all three groups with a higher percentage than prior to testing agreeing that they (strongly) agreed with the statement: Model (95.3%, P < 0.001), and Glasshorse (62.5%, P< 0.001) Textbook (31.6%, P = 0.029). Confidence levels were significantly higher in the Model group compared with both the Textbook (P < 0.001) and the Glasshorse (P = 0.008) groups, and significantly higher in the Glasshorse group compared with the Textbook group (P = 0.003). The degree of increase was significantly greater in the Model and Glasshorse groups (P < 0.001) compared with the Textbook group (P = 0.037), respectively. The increase observed in the Model and Glasshorse groups did not statistically differ (P = 0.107).
There was a significant increase in confidence levels with regards to “identifying the anatomical structures of the equine foot on MRI images” following the MRI teaching exercise in the Model (P < 0.001), Textbook (P = 0.013) and Glasshorse (P = 0.001) groups with 71.5%, 36.8%, and 37.5% of students respectively, (strongly) agreeing with the statement. Confidence levels following the MRI teaching exercise were significantly higher in the Model group compared with both the Textbook (P < 0.004) and the Glasshorse (P = 0.034) groups with no significant difference between the Textbook and Glasshorse groups (P = 0.092). Compared with the Textbook and Glasshorse groups the Model group exhibited the greatest increase in confidence (P = 0.004 and P = 0.006 respectively). The increase in confidence did not statistically differ between the Textbook and Glasshorse groups (P = 0.578).
The Model and Glasshorse group responses were significantly more positive than the Textbook group for all seven statements, and the Model group responses were significantly more positive than the Glasshorse group for all but one of the statements (p < 0.001), with both teaching aids considered equal with regards to ease of use (P = 0.058). Student responses relating to how helpful they considered their respective teaching aid in facilitating ‘mental visualization of the anatomical structures of the foot in 3D’, ‘orientation with regards to the MRI planes and slices' and ‘mental visualization of the MRI images in 3D’ were significantly more positive in the Model group compared with both Textbook and Glasshorse groups, with 95% of students either agreeing or strongly agreeing with all three statements. The Glasshorse received good student feedback with regards to aiding ‘mental visualization of the anatomical structures of the foot in 3D’ with 95.8% of students agreeing or strongly agreeing with the statement. Glasshorse student responses relating to ‘orientation with regards to the MRI planes and slices' and ‘mental visualization of the MRI images in 3D’ were slightly less positive with 66.7% of students (strongly) agreeing with both statements. Student feedback in the Textbook group was poor with 26.3%, 15.8% and 15.8% of students agreeing or strongly agreeing with the three statements, respectively.
Both the Model and the Glasshorse scored highly relating to ease of use and enjoyment with 92.9% and 75.0% of students respectively, (strongly) agreeing that the teaching aids were “easy to use” and “enjoyable and stimulating.” Textbook group responses were significantly less positive with 36.8% and 15.8% of students (strongly) agreeing that the textbooks were “easy to use” and “enjoyable and stimulating”. All students from the Model group (strongly) agreed that they would consider using the physical model again in the future both “when looking at MRI images” and “as an anatomy revision tool.” 79.1% and 83.3% of students in the Glasshorse group, and 31.6% and 57.9% of students in the Textbook group (strongly) agreed with the two statements.
Across all three teaching aid groups students considered “cadaver dissection” as the “most useful teaching modality in learning veterinary anatomy” (54.7%), followed by “3D computer models” (35.9%), “textbooks” (4.7%) and “lectures” (4.7%).
The area of medical and veterinary pedagogical research has made significant advances in recent years with institutions constantly striving for novel educational innovations to enhance learning experiences. These advances have influenced the anatomical sciences in particular, leading to the development of new and exciting learning resources such as 3D computer models (Nicholson et al., 2006; Sierra and Enderle, 2006; Nguyen and Wilson, 2009; Sergovich et al., 2010; Venail et al., 2010; Adams and Wilson, 2011; Codd and Choudhury, 2011), body painting (Op Den Akker et al., 2002; McMenamin, 2008), clay modeling (Waters et al. 2005; Motoike et al., 2009; Oh et al., 2009; Estevez et al., 2010), hand-held manipulatives (Krontiris-Litowitz 2003), plastinated specimens (Jones 2002; Dhingra et al., 2006; Latorre et al., 2007), and haptic simulators (Kinnison et al., 2009).
Rapid prototyping is a relatively new technique to the medical and veterinary industries, primarily being used to produce bespoke surgical implants (Ciocca et al., 2012; Frame and Huntley 2012; Galamatucci et al., 2006; Kuipers von Lande et al., 2012). In the first step a stack of cross-section images is converted into a virtual 3D, which is then used to cut the corresponding physical model from a block of material using laser cutting technology. It has the major advantage to be able to produce highly detailed models using different materials. The use of rapid prototyping techniques in the medical and anatomical sciences is well documented (Webb, 2000; Suzuki et al., 2004; Mäkitie et al., 2010; Esses et al., 2011; Torres et al., 2011; Li et al., 2012). Its application varies widely from bespoke surgical implants, advanced surgical approach training, to basic anatomy teaching models. Despite its widespread use in the medical sciences, use by the veterinary profession is largely limited to bespoke surgical implants (Kuipers von Lande et al., 2012), and has not been widely used to create veterinary anatomical models.
While there is no shortage of development and implementation of novel learning resources, empirical evidence relating to their efficacy, and perhaps more importantly their comparative efficacy, is scarce (Lewis, 2003). Many of these resources are incorporated into anatomy curricula with validation based purely on student perception, attitude and enjoyment. While the development of innovative learning resources should be actively encouraged, their incorporation into scientific education should not be solely based on student feedback, and should include quantitative evidence supporting their efficacy at enhancing student acquisition and understanding of knowledge. To this end this study was designed not only to develop a physical equine foot model using rapid prototype technology, but also to empirically validate its efficacy as an anatomical and imaging teaching aid. The efficacy of two commonly used alternative resources were also examined in this study, both for comparative purposes and in an attempt to provide some quantitative evidence regarding their efficacy, which seems largely absent from the literature.
This study has confirmed the visuospatial advantages of physical models over textbooks and 3D computer models, both objectively and subjectively. Students in the Model group scored significantly higher than both the Textbook and Glasshorse groups in the MRI assessment supporting the hypothesis that physical models would significantly enhance student anatomical visuospatial capabilities. The high scores in the identification of bones in all groups and the lack of statistical difference between groups were not surprising, since the identification of the bones requires only basic anatomical knowledge and visuospatial skills. A more comprehensive test of visuospatial capabilities was the identification of ligaments and tendons with the Model group scoring significantly higher than the other groups. This suggests that using physical models that can be manipulated in a 3D space can significantly benefit visuospatial understanding of structures with complex spatial relationships. These results are supported by studies in which plastinated specimens (Latorre et al., 2007), hand- held manipulatives (Krontiris-Litowitz, 2003; Jittivadhna et al., 2009, 2010), and clay models in particular (Waters et al., 2005; Motoike et al., 2009; Oh et al., 2009; DeHoff et al., 2011) significantly enhanced anatomical knowledge and visuospatial understanding.
The lack of significant difference between assessment scores for the Textbook and Glasshorse groups was unexpected. While textbooks provide detailed anatomical information, they lack the ability to illustrate spatial relationships in 3D and invariably contain more information than necessary for specific tasks, increasing the cognitive load on the user. Therefore the textbooks were expected to prove less effective than The Glasshorse at demonstrating complex spatial relationships. These results add to the already conflicting body of evidence surrounding the efficacy of 3D computer models (Lewis 2003; Estevez et al., 2010), confirming the need for objective analysis of their efficacy as anatomy learning resources.
Three-dimensional anatomical visualization and visuospatial skills are needed to interpret 2D cross-sectional images of 3D structures in different planes. This is fundamental to image interpretation in clinical practice. Students performed best in sagittal plane assessment images, which may be because this is the view most commonly depicted in textbook illustrations. The frontal plane appeared to be the most challenging. The students in the Glasshorse group struggled with this particular plane scoring significantly lower than both the Model and the Textbook group. While the mean scores were higher in the Model group there was no statistical difference compared with the Textbook group for the frontal plane. This was a surprising finding as textbooks are notoriously poor at conferring 3D spatial information. A likely explanation for this finding is that some textbooks contained photographs and illustrations in a similar orientation to the MRI Frontal plane, allowing students to match the structures directly from the textbook pictures to the MRI image, bypassing the need for mental visualization and orientation of the anatomical structures, and therefore not a true representation of visuospatial skills.
Some students, mostly in the Textbook and Glasshorse groups, scored very poorly in some aspects of the assessment, for example in the identification of the ligaments of the foot. This supports the growing body of evidence suggesting that the standards of anatomy knowledge among medical graduates are less than desirable (Cottam, 1999; Prince et al., 2005; Waterston and Stewart 2005; Estevez et al., 2010). While no such study has investigated this trend in the veterinary profession, anecdotal evidence together with the results from this study suggests that the concerns surrounding medical anatomical education are also applicable to veterinary education.
The participants in this study had just completed the preclinical part of their course and hence their anatomy training. Theoretically they should be confident in their ability to identify the anatomical structures of the equine foot. However, there was a distinct lack of confidence with regards to both identifying the anatomical structures of the equine foot and mentally visualizing the spatial relationships. Interestingly, when asked to state the anatomy learning resource they found most useful for learning general anatomy, the majority of students selected cadaver dissection. The extremely limited nature of equine foot cadaver dissection may offer an explanation for students' apparent lack of confidence with regards to this subject. This fact together with the significantly greater increase in confidence levels exhibited by the Model group following the teaching exercise suggests a correlation between hands-on, physical learning resources, and confidence levels and this will be investigated in further studies.
Although objective confirmation of efficacy should form the basis of learning resource validation, student confidence levels, and feedback are also important factors. Using the physical model significantly increased student confidence levels. This increase in confidence corresponded with good test scores. The same was not true of the Glasshorse group. While confidence levels significantly increased following the MRI teaching exercise, overall performance in the MRI assessment was poor, highlighting the worrying possibility that 3D computer models promote false confidence in 3D visuospatial capabilities. While students in the Glasshorse group scored relatively poorly in the test, student feedback was good. Other studies have illustrated similarly positive student feedback regarding 3D computer models (Nicholson et al., 2006; Venail et al., 2010; Codd and Choudhury 2011). With student feedback regarding the textbooks significantly less positive than the Glasshorse, 3D computer models could be used to augment and perhaps replace content traditionally taught through textbooks, with increased stimulation and engagement likely to lead to increased effort and ultimately a higher likelihood of private study. However with seemingly no correlation between the positive student feedback and objective learning benefits of The Glasshorse, this brings into question the validity of other teaching aids and learning resources developed and implemented into curricula on the basis of student feedback alone. A study utilizing a 3D computer model to enhance student learning of the human forearm musculature reported a similar finding with regards to a lack of correlation between good student feedback and objective efficacy (Codd and Choudhury, 2011). This seems to illustrate the need to evaluate new teaching tools individually and objectively before introducing them into a course.
Limitations associated with the study include that all subjects were volunteers, and by definition may not have been completely representative of the class as a whole. However veterinary students are generally considered to be high achievers (Zennerg et al., 2005) therefore the volunteer nature of the subject recruitment is likely to be less significant when extrapolating these results to the rest of the class. In this study no analysis of student learning styles was carried out. This could have elucidated associations between student learning styles, objective assessment scores and subjective feedback for the three teaching aids. We would have expected those students with kinesthetic learning preferences to benefit more from the physical model, while the textbooks and computer based program may have appealed more to visual learners. However this analysis was beyond the scope of this study. Similar studies have shown little to no correlation with regards to learning style (Estevez et al., 2010) and in our study the spread of data was much narrower in the Model group compared to the other groups, which may indicate that a higher proportion of students benefited from the physical model regardless of learning style. There are also known gender related differences in spatial abilities (Coluccia and Louse, 2004) however due to the considerable female bias in the study group, gender analysis was statistically impossible.
Another limitation was that the study concentrated on short-term memory and general understanding and appreciation of special relationships, with students' long-term retention of knowledge not assessed. However, with regards to the model developed in this study, long-term retention of knowledge was not the primary aim, with enhancement of visuospatial and 3D visualization capabilities, and the appreciation of complex spatial relationships considered more important. The development of the model was both labor intensive and expensive. There are other commercially available equine foot models; however, these invariably lack the manipulability and detachability of all the intricate structures within the equine foot, most notably the numerous ligaments within the foot. The detachability of each structure allowing complete visualization of each structure's origin, path, and insertion is unique to the model developed in this study.
This study has demonstrated the significant advantages associated with using a physical model in enhancing students' visuospatial appreciation and understanding of the complex anatomical architecture of the equine foot. While evaluating the potential of physical models as anatomical learning resources, this study has also exposed limitations associated with 3D computer models in illustrating complex 3D spatial information. The area of medical and veterinary pedagogical research is expanding rapidly and this study has highlighted the potential of physical models in providing the 3D teaching necessary for successful understanding, appreciation, and application of 3D information.
NOTES ON CONTRIBUTORS
DANIEL PREECE, B.Sc. (Hons), B.Vet.Med. (Hons), M.R.C.V.S., graduated from the Royal Veterinary College, University of London, North Mymms, Hertfordshire, United Kingdom in 2012 and is currently undertaking a small animal junior clinical training scholarship at the Royal Veterinary College, University of London.
SARAH B. WILLIAMS, B.Sc. (Hons), Ph.D. FHEA. is a lecturer in veterinary anatomy in the Department of Comparative Biomedical Sciences at the Royal Veterinary College, University of London, London UK. She teaches anatomy to students on multiple undergraduate programs and undertakes anatomical, biomechanics and educational research.
RICHARD LAMB, V.Sc. (Hons I), M.A.C.V.Sc. (Diagnostic Imaging), M.R.C.V.S, graduated from the University of Sydney, Australia in 2004 and is a resident in Diagnostic Imaging at the Royal Veterinary College, University of London, North Mymms, Hatfield, Hertfordshire, United Kingdom.
RENATE WELLER, D.V.M., Ph.D., M.Sc.Vet.Ed., Assoc. E.C.V.D.I., M.R.C.V.S., F.H.E.A., is a senior lecturer in diagnostic imaging in the Department of Veterinary Clinical Sciences at the Royal Veterinary College, University of London, North Mymms, Hatfield, Hertfordshire, United Kingdom. She splits her time between clinics, research and teaching in multiple undergraduate and postgraduate programs.