Secondary Logo

Journal Logo

Research Reports

Anatomy Education and Classroom Versus Laparoscopic Dissection-Based Training

A Randomized Study at One Medical School

ten Brinke, Bart, MD, MSc; Klitsie, Pieter J., MD, MSc; Timman, Reinier, PhD; Busschbach, Jan J. V., PhD; Lange, Johan F., MD, PhD; Kleinrensink, Gert-Jan, PhD

Author Information
doi: 10.1097/ACM.0000000000000223


Anatomy education for undergraduate medical students and (surgical) residents mainly consists of traditional, two-dimensional anatomy education, most commonly given in the form of lectures, PowerPoint presentations, and videos. This two-dimensional education is the least expensive and easiest way to teach human anatomy. Anatomy, however, is three-dimensional. Therefore, in combination with two-dimensional anatomy education, education on embalmed specimens has been used for years.1 The use of embalmed specimens has always been subject to discussion.2–6 One of the main issues that has contributed to the intensity and duration of this discussion is the absence of indisputable evidence of the added educational value of dissection-based teaching. There is also little evidence that teaching without embalmed specimens leads to a reduction in the skills and knowledge of students.7 One exception is the study by Beermann and colleagues,8 which showed that three-dimensional presentation of anatomy improves the identification of complex anatomy.

With the introduction of laparoscopic surgery, good knowledge of anatomical orientation and recognition has become even more critical. Surgeons have to be able to identify anatomical structures from a screen while tactile feedback and spatial orientation are reduced. The additional challenges of laparoscopy result in more gradual learning curves,9–11 which warrant more effective teaching modes of (laparoscopic) anatomical knowledge. These teaching modes should consist of real, three-dimensional anatomy, so that students and surgical residents do not have to transfer the abstract and two-dimensional representation of the anatomical structures to real and three-dimensional human anatomy in practice. An enhanced understanding of spatial anatomy must be the aim of the training. In our opinion, laparoscopic dissection-based education can be an important part of this training.

At Erasmus Medical Center, The Netherlands, human specimens were embalmed using the new Rotterdam AnubiFiX technique that allowed the dissection models to remain supple and flexible. In contrast to traditionally embalmed specimens, these dissection models are insufflatable, thus enabling anatomy education using laparoscopy. We created six anatomy training models of the dorsal view anatomy of the groin area, which is known for its complex, three-dimensional form. Some of these models were fully prepared, important structures were separated from the surrounding tissue, and the relevant structures were colored with a newly developed formaldehyde proof paint to allow students to gradually make the transition between learning the anatomy from an atlas to learning the real anatomy. Other models were not prepared and colored, so they represented the real anatomy. (See Supplemental Digital Figure 1,, for one of the fully prepared and colored training models.)

Figure 1
Figure 1:
Estimated scores achieved by three groups of medical students (N = 45) who attended an anatomy training, Erasmus Medical Center, Rotterdam, 2011. Tests were scored out of 21 possible correct identifications. The tests were performed before the anatomy training (precourse), immediately after (postcourse), and two weeks after the training (follow-up). Group I (n = 15) attended lectures only, group II (n = 16) attended dissection-based training using three-dimensional laparoscopic dissection models only, and group III (n = 15) attended lectures as well as dissection-based laparoscopic training.

Education using these specimens offers obvious advantages. The anatomy as shown on these specimens is real anatomy, with the advantage that important anatomical structures that cannot always be seen during surgery can be shown and accentuated with colors. Disadvantages of this form of education are the labor-intensive nature and the additional associated costs. To justify these increased demands, the assumed superiority of this learning method should be evaluated critically. We therefore performed a randomized study to examine the added value of (laparoscopic) dissection models over traditional anatomy education, using supple embalmed specimen models.


Study design

We performed a randomized study to test whether laparoscopic dissection-based training improves understanding of anatomy. Therefore, we developed three different educational conditions. Groups in all three conditions were trained in exactly the same anatomical structures on an evening course in June 2011. Group I attended a 60-minute traditional classroom lecture, addressing the dorsal view anatomy of the inguinal region using a PowerPoint presentation. Group II attended a 60-minute dissection-based training using three-dimensional laparoscopic dissection models. Group III attended the lecture as well as the dissection-based training, and their education took 120 minutes. Because the students had no knowledge about this anatomical region at all, an experienced teacher explained the anatomy in the first 10 minutes of the dissection-based training, showing the structures in the three-dimensional models. During the remaining 50 minutes, students had the opportunity to see, touch, manipulate, and study the anatomy themselves, while the teacher was around to assist the students if necessary.


We recruited medical students from Erasmus Medical Center through class announcements in 2011, which we made in four classes of 24 students (96 students were eligible to participate). Participation was voluntary, and students from all years of medical school (year 1–6) were represented. Participants were randomized in three groups, stratified by years of medical school. All years of medical school were randomized separately, so that students from all years of medical school were spread equally over the three groups. As the anatomy of the posterior (dorsal) inguinal region is complex and is not addressed during medical school, we were confident that our students were naive about the subject of the training. Hence, all groups were equal and had no pretest knowledge.


To assess the improvement of anatomical knowledge of each group, all students completed a test before and after anatomy training. The two tests were identical. Twenty-one structures of the inguinal region were marked inside anatomical models (List 1).

List 1 Anatomical Identification Test of Structures of the Inguinal Region, Marked for Identification Inside Dissection Models, and Administered to 45 Medical Students, Erasmus Medical Center, Rotterdam, 2011a Cited Here

  1. Genital branch of genitofemoral nerve
  2. Obturator vein
  3. Iliopsoas muscle
  4. Thompson ligament
  5. Gimbernat ligamentb
  6. External iliac vein
  7. Lateral femoral cutaneous nerve
  8. Testicular artery
  9. Cooper ligamentb
  10. Inferior epigastric artery
  11. Rectus abdominus muscle
  12. Inferior epigastric vein
  13. Triangle of “pain”
  14. Triangle of “doom”
  15. Pubic branch of inferior epigastric artery
  16. Localization of direct hernia (medial of inferior epigastric artery)
  17. Corona mortis
  18. Retzius space
  19. Bogros space
  20. Gimbernat ligamentb
  21. Cooper ligamentb

aThree groups of students had to write down the full scientific names of the marked structures to assess their anatomy knowledge before, immediately after, and two weeks after an anatomy training. Group I attended lectures only, group II attended dissection-based education only, and group III attended lectures as well as dissection-based training.

bSome structures were marked twice in different anatomy models.

To test anatomical structures inside the anatomical models and not by a two-dimensional (multiple-choice) test is the closest thing assessing the actual in vivo anatomy. Finally, medical students should be trained to recognize human anatomy in three-dimensional form as they will encounter in practice and not from two-dimensional images. Students had to write down the full scientific names of the marked structures (according to the official “nominal anatomica”). The test score was the same as the number of correct answered structures (maximum 21). To evaluate the long-term results, students were asked to take a third test two weeks after training. This test was identical to the first two tests.

Statistical analysis

Baseline differences among the three groups were tested with one-way ANOVA with Scheffé correction. We analyzed the longitudinal data using mixed or multilevel modeling in SPSS (PASW statistics 17, Chicago, Illinois). This method is able to handle incomplete time-series data with a minimal loss of information. In this multilevel model, the different time points were defined at level one and the participants were defined at level two. Group I, which only attended the lecture part of the course, was used as the reference group. Contrasts were formulated to analyze the significance of the estimated differences among the groups at posttest and at follow-up. Time linear and a spline (knot) at the end of the treatment were used to model time effects. Effects were the increase in knowledge as a result of the course and the decrease in knowledge as a result of loss of retention (forgetting). First, we postulated a saturated model with time-by-group interactions and an unstructured covariance matrix. Then the covariance matrix was simplified, as nonsignificant effects (P > .05) were deleted until a parsimonious final model was reached. We tested the differences between covariance matrices and saturated and parsimonious models with the deviance test.12 The residuals of the model were checked to be normally distributed, a necessary assumption for a correctly fitted mixed model. We calculated effect sizes by dividing the differences by the estimated standard deviations of the model at various time points:

For the interpretation of the effect sizes, we used Cohen13 d: an effect size of 0.20 is considered a small effect, 0.50 a medium effect, and 0.80 a large effect.

Given that there were no medical interventions related to this study, no patients were involved, and participation was voluntary, no medical ethical approval was requested.


Forty-six of 96 eligible students participated in this study (48%). Randomization resulted in 15 students in group I (only lectures), 16 students in group II (only dissection models), and 15 students in group III (both lectures and dissection models). Participants came from all six years of medical school. Nineteen (41%) were female, while 37 were male (59%). The three groups were comparable in years of medical school (mean 2.6 years, P = .980). All participants completed a test before the start of education. We excluded one outlier from group III with an exceptionally high pretest score who turned out to have prior knowledge about the anatomical region, resulting in 45 students total and 14 students in this group. No significant difference was found among the three groups before training (P I–II = .81; P II–III = .90; P I–III = .55). Twenty-two (48%) students participated in the follow-up after two weeks. No significant differences were found between participants and nonparticipants at baseline scores of the test (P = .79) and between the scores at the end of the course (P = .44).

After exclusion of the outlier, the residuals were normally distributed. We considered a heterogeneous autoregressive covariance matrix (Singer and Willett12) an adequate covariance structure (deviance χ2(2) = 1.22, P = .54). The main effect of the group could be removed from the saturated mixed model, which could be anticipated as there were no initial differences among the groups. The difference between the saturated and final parsimonious model was not significant (χ2(2) = 2.71; P = .26). The estimated values from this model are provided in Table 1, and the model is depicted in Figure 1.

Table 1
Table 1:
Estimations of Test Scores and Effect Sizes for Change in Time, by Group (I, II, or III) of Medical Students Who Completed a Test Before (Precourse), Immediately After (Postcourse), and Two Weeks After (Follow-up) an Anatomy Training, Erasmus Medical Center, Rotterdam, 2011a

Immediately after the course, groups II and III scored higher than group I (P < .001; P < .001). This difference continued at follow-up (P = .003; P = .002). Group II scored higher than group III in the test that was taken immediately after the course (P = .009; d = .86, a medium effect). All groups scored lower at follow-up than immediately after the course. The superiority of group II over group III disappeared at the follow-up test.


This study demonstrates that anatomy education using dissection models enhances medical students’ ability to learn anatomy. Students who received the three-dimensional teaching (groups II and III) scored far better in the short- and long term when compared with students who received only conventional, two-dimensional teaching (group I).

There might be a number of reasons why anatomy education using embalmed specimen results in better retention of knowledge compared with two-dimensional anatomy education using lectures.

The first reason is that the students from group II and III were taught on realistic three-dimensional models. The students did not have to transfer the abstract and two-dimensional representation of the anatomical structures to the real and three-dimensional anatomy. In other words, using embalmed specimens is the more realistic learning environment and has therefore better learning results. We believe it is fair to say that students in groups II and III, who trained within that realistic environment, were best prepared for recognition of human anatomy.

A second reason might be that the students from group II and III are actively involved, instead of being brought into a more passive mode, as in a classical lecture. An active mode has an important contribution to better learning results, as the students discover the material themselves.14 Besides that, students are given the opportunity to learn interactively, which results in peer tutoring, whereby those with more knowledge assist others with less knowledge.15

A third supportive hypothesis is the possibility for manual interaction with the material in the dissection models. Piotr Galperin assumes that learning follows the sequence of manual interaction, verbalization, and memorization.16 Manual interaction allows the students to study the anatomy from multiple views, which improves the spatial understanding.17 Therefore, the manual interaction is an essential step to understand the three-dimensional, complex form of anatomy. However, in classical teaching using lectures, this important step is not present, and structures are mostly presented from two viewpoints.

The three hypotheses above fit well in the learning theory of “constructivism” as defined by Shuell18: “Learning is a constructive, active, cumulative and goal directed process.” Knowledge has to be constructed actively through the adjustment of new information into already-acquired knowledge (assimilation) and through the adjustment of existing mental structures to new experiences (accommodation). Because students do not have to transfer the abstract and two-dimensional anatomy to the real three-dimensional anatomy and have the possibility to touch, see, and discover the anatomy themselves, it will be easier to process and to assimilate the new information. Moreover, dissection-based anatomy education in the dissection room activates and motivates the students instead of placing the students in a passive role like in a lecture. According to the learning theory of constructivism, new knowledge is added to prior knowledge, which means that learning is a cumulative process. A short theoretical introduction in the dissection room for the students of group II and III provides a basis on which new knowledge can be built. This allows for an active integration of the new knowledge. Anatomy education using embalmed specimens is, as well as education using lectures, goal directed, as students aim to understand the three-dimensional anatomy. All above-mentioned aspects are present in our three-dimensional anatomy training method and may contribute to better results of three-dimensional, dissection-based training.

One could argue that the number of students is low, particularly the number of students that took the test after two weeks. This low attendance after two weeks occurred because the academic year had ended when the follow-up test was taken, and many students were on holidays. The course could not be organized earlier in the academic year because the dissection room was occupied for regular education and other courses. However, our analyses show clear statistically significant results. To confirm these results, we suggest further research with better-planned logistics to prevent drop-out of students.

Some of the reported effect sizes are exceptionally high. Our explanation is that because students hardly had any knowledge of this subject prior to the course, all precourse scores were near zero, resulting in a very small standard deviation. This, in turn, led to the high effect sizes.

The education of group III was twice as long compared with the other two groups. This might explain the superior results of this group. On the other hand, group II had comparable results, while the classes of this group were of the same duration as group I. Therefore, we do not believe that the variability of the duration interferes with the conclusions of this experiment. Indeed, one could even argue that our results may indicate that more education is not necessarily better education: Group II, which received only three-dimensional education, performed better than group III, which received both two- and three-dimensional education. The two-dimensional information may have interfered with the three-dimensional representation on the short-term results, or fatigue may have reduced the effectiveness of the three-dimensional education of group III.

This study shows that identification and recognition of anatomical structures can be better trained in a three-dimensional environment because the actual use of this kind of information (the actual test) will always be in a three-dimensional setting, the operation theater. It is reasonable to assume that students and surgical residents have benefits from three-dimensional, dissection-based anatomy education because this training prepares them best for the “real thing.” Further clinical studies should be performed on the issue of knowledge transition. The question, whether a resident is actually operating more efficiently (faster and with less complications) after the proposed three-dimensional training, has yet to be answered.

Although the costs and effort in this form of teaching are higher, the results of this study are clear. Educators and medical schools in general should make an effort to maintain the anatomy education in the dissection room. We suggest that medical schools should equip an accommodation in the dissection room where surgical residents could study the anatomy using (laparoscopic) dissection models. In this way especially residents can use the dissection room in a more personalized way to acquire the three-dimensional anatomical knowledge when they need it, and in a way they need it (e.g., during laparoscopic views).

Three-dimensional anatomy training with dissection models resulted in an increase of spatial and three-dimensional understanding of the anatomy of the inguinal region for the participants in our study. Although education in the dissecting room is more expensive, anatomy training with dissection models has a clear added value in the education of medical students.


1. Dyer GS, Thorndike ME. Quidne mortui vivos docent? The evolving purpose of human dissection in medical education. Acad Med. 2000;75:969–979
2. Aziz MA, McKenzie JC, Wilson JS, Cowie RJ, Ayeni SA, Dunn BK. The human cadaver in the age of biomedical informatics. Anat Rec. 2002;269:20–32
3. Crisp AH. The relevance of anatomy and morbid anatomy for medical practice and hence for postgraduate and continuing medical education of doctors. Postgrad Med J. 1989;65:221–223
4. Ellis H. Teaching in the dissecting room. Clin Anat. 2001;14:149–151
5. Gregory SR, Cole TR. MSJAMA. The changing role of dissection in medical education. JAMA. 2002;287:1180–1181
6. Jones DG. Reassessing the importance of dissection: A critique and elaboration. Clin Anat. 1997;10:123–127
7. McLachlan JC, Bligh J, Bradley P, Searle J. Teaching anatomy without cadavers. Med Educ. 2004;38:418–424
8. Beermann J, Tetzlaff R, Bruckner T, et al. Three-dimensional visualisation improves understanding of surgical liver anatomy. Med Educ. 2010;44:936–940
9. Eklund AS, Montgomery AK, Rasmussen IC, Sandbue RP, Bergkvist LA, Rudberg CR. Low recurrence rate after laparoscopic (TEP) and open (Lichtenstein) inguinal hernia repair: A randomized, multicenter trial with 5-year follow-up. Ann Surg. 2009;249:33–38
10. Neumayer L, Giobbie-Hurder A, Jonasson O, et al.Veterans Affairs Cooperative Studies Program 456 Investigators. Open mesh versus laparoscopic mesh repair of inguinal hernia. N Engl J Med. 2004;350:1819–1827
11. Simons MP, Aufenacker T, Bay-Nielsen M, et al. European Hernia Society guidelines on the treatment of inguinal hernia in adult patients. Hernia. 2009;13:343–403
12. Singer JD, Willett JB Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. 2003 Oxford, UK Oxford University Press
13. Cohen J. A power primer. Psychol Bull. 1992;112:155–159
14. Bruner JS Actual Minds, Possible Worlds. 1986 Cambridge, Mass Harvard University Press
15. Granger NA. Dissection laboratory is vital to medical gross anatomy education. Anat Rec B New Anat. 2004;281:6–8
16. Haenen J Piotr Galperin: Psychologist in Vygotsky’s Footsteps. 1996 New York, NY Nova Science Publishers, Inc.
17. Garg AX, Norman G, Sperotable L. How medical students learn spatial anatomy. Lancet. 2001;357:363–364
18. Shuell TJ. The role of the student in learning from instruction. Contemp Educ Psychol. 1988;13:277–278

Supplemental Digital Content

© 2014 by the Association of American Medical Colleges