Comparative Analysis of Multidimensional Learning Tools in Anatomy: A Randomized Control Trial

Abstract

Introduction:

Anatomy teaching has traditionally been based on dissection. However, reduced hours in total and laboratory hours in gross anatomy along with a dearth of cadavers have ensued the search for a less time-consuming tool.

Materials and Methods:

The study was conducted in the Department of Anatomy in Sheikh Bhikhari Medical College, Hazaribag. A total of 282 medical students were taught gross anatomy, using three different learning modalities: dissection (n = 95), plastic models (n = 94), and three-dimensional (3D) anatomy software (n = 93). The knowledge of the students was examined by 100 multiple-choice question (MCQ) and tag questions followed by an evaluation questionnaire.

Results:

When performance is considered, the dissection and 3D group performed better than the plastic models group in total, MCQs, and tag questions. In the evaluation questionnaire, dissection performed better than the other two modalities. Moreover, dissection and 3D software emerged as superior to the plastic models group.

Statistical Analysis:

All data were analyzed using the one-way ANOVA and t-test. Group-based analysis by ANOVA and gender-based analysis were done by Student’s t-test. A comparison of students’ perceptions was done by Kruskal–Wallis H-test.

Conclusion:

Dissection remains a favorite with students and accomplishes a significantly higher attainment of knowledge. Plastic models are less effective but are a valuable tool in preparation for cadaveric laboratories.

INTRODUCTION

Anatomy was illustrious in the bygone days. It laid a foundation for all of biomedical science. The subject is considered the bedrock for medical studies. However, with emerging newer techniques in medical educational, teachings in anatomy have also evolved. Teaching in anatomy has traditionally been dissection based. Nevertheless, a decrease in time devoted to the laboratory coupled with cadaver shortage has initiated a search for less time-consuming tools. Alternative teaching modalities in anatomy constantly appear.

Aim of the study

1. To compare the efficiency of the most common teaching methods in anatomy

2. To scrutinize student’s perceptions on each teaching method.

Objectives of the study

1. Recognize any incongruity in efficacy between dissection groups, plastic model groups, and three-dimensional (3D) software groups – anatomy software as evaluated by exam performance

2. Establish a difference in perception of students for each teaching modality, based on their evaluation questionnaire answers.

Inclusion criteria

First-year, undergraduate medical students.

Exclusion criteria

Students having previous knowledge of human anatomy.

PICO model

The PICO model for the study was defined as: Population: medical students, Intervention applied: Various Tools (Dissection, Plastic models, 3D software), control: Inter group comparison, Outcome: Effectiveness and perception by students.

MATERIALS AND METHODS

This study took place in Sheikh Bhikhari Medical College, Hazaribag, between 2019 and 2022. Table 1 shows the distribution of students in 3 years.

Table 1.

Total sample size and pattern of distribution in each group

Modality	2019	2021	2022	Total
Lectures with plastic models	33	30	31	94
Lectures with 3D software	30	31	32	93
Lectures with dissection	32	33	30	95
Total				282

3D=Three-dimensional

This study was done 15 days before the start of formal classes of 1^st year students. Informed consent was taken from all participants in informed consent form. A preliminary introductory presentation was made to sensitise students about the goal of the study as well as its advantages. All probable participants were well briefed about the duration of the study and its venue. Details about the teaching modality and the teaching subjects had not been revealed to avoid a possible bias. Identity of the participants was blinded to the investigators.

Students were preprimed that both the examination sheet and the questionnaire for evaluation would be made incognito to secure unbiased answers and reduce performance pressure.

The study was arranged to compare student’s performance and perception following their training in gross anatomy, using 3 different teaching modalities presented in Table 1.

Three randomized student groups were created using the auto-draw computer program random picker. The total sample (n = 282) of the study had been the sum of three groups.

Dissection was performed on fresh frozen cadavers available in the department of anatomy. Plastic models were available already in our department. The software used was the Biodigital Human (version 1.0.4, www.biodigital.com), which was freeware and available in Windows Android and iOS.

Educational process

Two lectures and two laboratory sessions were conducted for each group, and each session lasted for 2 h. The first educational session (1^st lecture - >1^st laboratory) was about bone and ligament and 2^nd one was about muscle anatomy. Lectures were similar for all groups. Detailed handouts of both lectures were provided to students to aid them during laboratory work. The laboratory session started with a 20-min demonstration of all structures. After that, students had the chance to inspect all structures independently for 90 min. All three groups had been divided into teams of 10–15 students. The exposure time of students in each modality was the same. In Group I, a fresh frozen cadaver was dissected using basic surgical instruments. The introductory dissection process was recorded and screened in front of the dissection table to aid amateur students. In Group 2, plastic models were used. Students had the opportunity to construct and deconstruct the plastic models separating the superficial structures to explore the structures underneath. In Group 3, students had the opportunity to inspect all structures using rotate, zoom in/out, dissect for removing structures, and isolate mode for segregating structures.

Examination process

On the conclusion of the learning process, students had to take part in the final examinations. One hundred questions were asked to assess each group. Half of the questions involved naming anatomical structures in images projected on the screen (tag questions) and the other 50% of questions were multiple-choice question (MCQ). Both tag question and MCQ were identical for all groups. All tag and MCQs were generated and reviewed by the department’s tutor. The whole examination process was of 100 min duration. The examination process was anonymous. A six-digit coding was employed and written on the answer sheet by each student. After the examination, students were requested to fill in an anonymous questionnaire to assess the method they used. The evaluation questionnaire comprised 10 questions.

In the first segment, 2 questions included demographic data (age and sex). The second segment consisted of 6 questions (4 for evaluating the teaching modality and 2 for inspecting students feeling). The third part comprised 2 questions. Part one investigated the participant’s intent to recommend the course to the upcoming students, and the second part consisted of an open-ended question requesting students to make a short comment to state their suggestions or opinions regarding the teaching process. The same 6-digit code was used in the evaluation questionnaire. All data were collected and analyzed.

Statistical analysis

The analysis of data was done in the following ways:

The level of statistical significance was kept at 0.05.

Student’s scores in the examination process were compared with one-way ANOVA.

Levene’s test was used to evaluate the homogeneity of variances. When homogeneity of variances was met, the Bonferroni post hoc test was used to analyze the difference between groups.

Student scores of the same group in different types of questions were compared with the Paired-samples t-test.

Difference between the scores of male and female students in each group was assessed with Student’s t-test.

Student’s answers to the Likert-scale questions were compared using the Kruskal–Wallis test. All pairwise comparisons were done by the Dwass–Steel–Critchlow–Fligner test.

RESULTS

In total, 270 students (96% of the initial samples) reached the end of the whole course. Ninety-one students performed dissection (Group 1), 90 students used the BioDigital Human 3D anatomy software (Group 2), 89 students used plastic models (Group 3), and 6 students quit the study.

The remaining 6 students were refrained from participation in the examinations as they had not attended either lectures or laboratories.

The mean age of the participants was 17.4 (±1) years.

About 46% were male students and 54% were female, as shown in Table 2.

Table 2.

Total samples and distribution of male and female

Modality	Total students	Male	Female
Plastic models	89	40	49
3D	91	40	51
Dissection	90	44	46
Total	270	124	146

3D=Three-dimensional

As far as student’s performance were concerned, in all types of questions (Tag, MCQ, Bloom 1 and 2 and total questions), students of 3D and Dissection group performed better than the plastic models group. There was no difference or a little bit difference in performance of 3D and dissection. The results of the analysis using ANOVA and Bonferroni post hoc test are presented in Table 3.

Table 3.

Analysis using ANOVA and Bonferronl post hoc test

Question type	ANOVA F-test P	Group	n	Mean±SD	Post hoc test (Bonferroni)
Tag questions	F=367.68	Plastic (1)	89	23.85±2.4	2, 3>1
	P<0.0001	Dissection (2)	91	31.11±1.88	No difference between 2 and 3
		3D (3)	90	31.17±1.86
MCQ	F=265.70	Plastic (1)	89	24.76±1.97	3, 2>1
	P<0.0001	Dissection (2)	91	30.54±1.81	No difference between 3 and 2
		3D (3)	90	30.58±2.05
Bloom 1 MCQ	F=163.57	Plastic (1)	89	13.28±1.19	3, 2>1
	P<0.0001	Dissection (2)	91	15.96±0.96	No difference between 2 and 3
		3D (3)	90	15.9±1.21
Bloom 2 MCQ	F=239.86	Plastic (1)	89	11.47±1.09	3, 2>1
	P<0.0001	Dissection (2)	91	14.62±1.14	No difference between 3 and 2
		3D (3)	90	14.7±1.12
Total questions	F=467.79	Plastic (1)	89	48.58±3.85	2, 3>1
	P<0.0001	Dissection (2)		61.69±3.11	No difference between 2 and 3
		3D (3)	90	61.76±2.92

MCQ=Multiple-choice questions, SD=Standard deviation, 3D=Three-dimensional

Year-wise analysis

The ANOVA results between the various tests conducted in three different years (2019, 2021, and 2022) revealed the following findings.

For students taught via 3D models, a year-by-year analysis revealed no significant statistical differences in marks obtained in Bloom’s level 1 questions (P = 0.1669). However, significant differences were observed over the years in Bloom’s level 2 questions (P = 0.0002), post hoc analysis via Bonferroni test revealed differences in marks obtained in the year 2021 and 2019, with P < 0.0001 and in between the years 2022 and 2019 (P = 0.027). Students in the year 2019 performed better than students in the later 2 years.

Statistical differences were also observed in the marks obtained for MCQs (P = 0.0045), with students in 2019 outperforming students in 2021 (P = 0.005) and 2022 (P = 0.045). Marks obtained in TAG question revealed differences over the 3 years (P = 0.0001), with students in 2021 and 2022 scoring less than students in 2019 (P = 0.003 and P < 0.0001, respectively). Total marks obtained in all the questions over the years revealed no significant statistical difference.

For students taught via dissection, ANOVA test failed to reveal significant statistical differences in marks obtained over the years in any type of evaluation method or in total marks obtained (Bloom’s level 1, P = 0.33, Bloom’s level 2, P = 0.57, MCQ, P = 0.69, TAG questions, P = 0.76, Total marks, P = 0.98).

Students taught with the help of plastic models showed significant differences in marks obtained in Bloom’s level 1 questions over the years (P = 0.0139), with students in 2019 scoring higher than students in 2022 (P = 0.012). Marks obtained in TAG questions also revealed a significant difference over the years (P = 0.0088), with students in 2022 having improved scores over students in 2019 (P = 0.007). Marks obtained in Bloom 2 questions, MCQs, or the total marks obtained showed no significant statistical differences.

One-way ANOVA was applied to analyze the scores of the students in the examination. There was a significant difference in the performance between the groups (P < 0.0001). Post hoc test by Bonferroni revealed the differences between individual groups.

In total scores, dissection and 3D groups performed significantly better than groups taught by plastic models (P < 0.0001). However, scores between dissection and 3D groups did not vary significantly. Figure 1 shows a boxplot of total scores in the examination for the three groups with whiskers from maximum to minimum. Segment in the box plot demonstrates the median scores of the three groups with dissection and 3D group scores higher than scores of groups taught via plastic models.

Figure 1.

A boxplot of students’ total scores in the examinations. Performance of students in total questions is shown. 3D = Three-dimensional

Tag questions showed a significantly higher score in dissections and 3D group compared to plastic model groups (P < 0.0001).

Figure 2 demonstrates a box and whisker plot of the same with whiskers extending from maximum to minimum and the segment demonstrating the median score.

Figure 2.

A boxplot of students’ performance in tag questions. 3D = Three-dimensional

MCQ questions also revealed a similar distribution with dissection and 3D groups performing better than plastic models group (P < 0.0001). Figure 3 demonstrates a box and whisker plot of scores in MCQ, whiskers showing marks from maximum to minimum and median scores demonstrated by the segment in the box.

Figure 3.

A boxplot of students’ performance in multiple-choice question. MCQ = Multiple-choice question, 3D = Three-dimensional

Bloom 1 questions also revealed a similar scenario with students in dissection and 3D model group scoring better than students taught by plastic models (P < 0.0001). Figure 4 shows the significantly higher mean scores of the two groups via box plot.

Figure 4.

A boxplot of students’ performance in Bloom 1. 3D = Three-dimensional

Bloom 2 questions reveal a significantly higher score of dissection and 3D group as compared to the plastic group (P < 0.0001). Figure 5 demonstrates the scores in the form of a box plot with whiskers spanning from maximum to minimum and the segment in the box demonstrating the median.

Figure 5.

A boxplot of students’ performance in Bloom 2. 3D = Three-dimensional

As far as gender-based analysis was concerned, no significant difference between male and female students was observed, in all types of questions in the plastic, dissection, and 3D groups. The results are shown in Table 4.

Table 4.

Gender based analysis by Student's t-test in all three groups

Group	Question type	Mean score (male)	Mean score (female)	SD (male)	SD (female)	t-test (P)
Plastic	Bloom 1 level questions	13.6	13.1	1.13	1.215	0.049
	Bloom 2 level questions	11.4	11.5	1.25	0.960	0.757
	MCQ	25.0	24.6	2.02	1.934	0.311
	Tag questions	23.1	24.5	1.77	2.707	0.004
	Total questions	48.0	49.1	3.38	4.181	0.180
Dissection	Bloom 1 level questions	15.9	16.0	1.03	0.910	0.784
	Bloom 2 level questions	14.5	14.7	1.12	1.172	0.419
	MCQ	30.5	30.7	1.77	1.876	0.513
	Tag questions	31.0	31.3	1.96	1.819	0.460
	Total questions	61.4	62.0	3.14	3.096	0.407
3D	Bloom 1 level questions	16.1	15.7	1.38	1.05	0.137
	Bloom 2 level questions	15.0	14.5	1.16	1.06	0.030
	MCQ	31.1	30.2	2.09	1.95	0.034
	Tag questions	30.4	31.8	1.71	1.73	<0.001
	Total questions	61.5	62.0	2.94	2.91	0.369

For all types of questions. MCQs=Multiple-choice questions, SD=Standard deviation, 3D=Three-dimensional

Qualitative results, i.e., students perceptions about teaching modality, are shown in Table 5.

Table 5.

Comparison of students perceptions using Kruskal–Wallis test and Dwass-Steel-Critchlow-Fligner pairwise comparisons

Questions	Kruskal–Wallis test P	Rank sum			Dwass-Steel-Critchlow-Fligner pairwise comparison
		Plastic (1)	3D (2)	Dissection (3)
Satisfaction with teaching modality	<0.001	8424	13,483	14,677	3>2>1
Anticipation (students feeling)	<0.001	8069	12,942	15,574	3>2, 1
Suggest course to upcoming students	<0.001	6781	12,728	17,075	3>2, 1
Any fear/stress before laboratory	<0.001	10,692	7493	18,399	3>2, 1
Any fear/stress after laboratory	<0.001	11,527	7456	17,601	3>2, 1

3D=Three-dimensional

All together 66% of all students were well pleased with their teaching modality. Sixty-seven percentage of students found their teaching modality of use for their clinical ventures in the near future. However, 20% of students felt severe distress or anxious before the laboratory. Fear/stress also noticed in 10% of students after laboratory. Finally, 85% of all students would suggest the course to forthcoming students. The results are presented in Table 5.

Higher rates were noticed in the dissection group when compared to the 3D and plastic groups, in all questions. In response to the questions, it was suggested that the dissection was more preferred than 3D and plastic course to upcoming students. In the dissection group, students were more stressed before and after the laboratory when compared to plastic models and 3D software. Finally, the dissection group outperformed the plastic models and 3D software groups in the question suggesting courses to upcoming students.

In open-ended questions, the majority of students (95%) needed more time in the laboratory and also wanted to try other teaching modalities (83%) for better knowledge. However, students of the dissection group preferred 3D software during clarification of anatomy in the upcoming semesters. At the same time, students of the 3D and plastic models group asked for cadaveric dissection. About 30% of students of the dissection group described the experience as the start of their surgical career.

DISCUSSION

This is the first study conducted in Jharkhand where three learning modalities were compared for teaching anatomy.

On analysis of data, it is clear that dissection and 3D software are more efficacious than plastic models in anatomy teaching.

The best part of BioDigital Human software was isolation and 360° view for every anatomical structure which was not feasible with dissection, which helped to enhance spatial anatomy acquisition. On the other hand, dissection provides a touch-mediated understanding of the human body, which is not compared with simulated and virtual world dissection.

Codd and Choudhury[1] failed to differentiate between this 3D software and dissection using a small pool of 36 students with 10 questions and a control group. Their findings were consistent with the results obtained in our study as our study questions were same in the two groups.

Hisley et al.[2] performed 3D visualization technology. They reported 3D magnetic resonance imaging – computed tomography (CT) was superior to dissection in acquisition of spatial knowledge. However, the sample was small (only 16 students). Their finding did not match with our finding, since in tag questions, students of the 3D group and the dissection group performed equally. As far as student preferences, satisfaction, and anticipation are concerned, dissection group received greater grades in comparison to 3D and plastic model groups.

In open-ended questions, dissection outmarked the other teaching modalities as astonishing experiences by students. The gold standard in anatomy teaching for students is undoubtedly dissection.[3,4,5,6]

Student’s perception regarding clinical usefulness had been found same for 3D and dissection. For future clinical activities, both modalities are useful because Dissection is related to practical skills and 3D software. Dissection is related to practical skills while 3D is related to software skills. 3D software can be freely used at home, and in self-study and would be very handy in a vertically integrated medical syllabus, making it a promising teaching tool in anatomy.

Sinclair[7] studying the anatomy of the lower limb in a sample of 219 students reported a minor edge of dissection in tag and MCQ, but no difference in open-ended questions.

Peppler et al.[8,9] failed to determine any significant statistical variation in a sample of 30 students studying anatomy of the upper and lower limbs. However, 5 years later, a repeat study with a larger sample size revealed dissection to be better in 5 of the 24 tests.

Peeler et al.[10] found no significant difference between dissection and 3D Software.

The plastic models group achieved a lower score when compared to all other groups in total, MCQ, and tag questions.

Khot et al.[11] compared plastic replicas of the pelvis to 2D Images and a 3D reconstruction of a CT scan. They found plastic models better than 3D in tag questions. However, this is not true with findings obtained in our study, since Khot et al. did not use a software which was commercially available but used a CT, 3D reconstruction. A recent meta-analysis demonstrated the effectiveness of plastic models in acquiring spatial knowledge and long term retention of knowledge.[12,13] However, plastic models do not accurately represent the accurate shape and surface of the human body. Our study finding matches with the above study because the plastic models group performed less than the other groups.

Regarding students’ perceptions, plastic models were rated lower than dissection and 3D groups. This opinion is clearly indicated by the student’s answer. Human anatomical variation seen during dissection cannot be observed in plastic models. One thing more is that practical skills cannot be developed by plastic models.

In the analysis based on gender, no significant differences were seen between three groups.

The other outcomes which have been studied in collaboration of more than one laboratory method for anatomy teaching.[1,3,4,7,14]

• Time interval increased in the anatomy laboratory

• Preparation of the students for human body care and doctor–patient relationship in the dissection of cadaver which cannot be done in 3D and plastic models.[15,16,17]

Limitation of the study

In this study, students were limited to one study group, so it is difficult to safely exclude the other one.

Finally, commercially available 3D software and models are of different quality available in the market which results in different findings. Hence, the results cannot be generalized. This study finding cannot be implemented in all types of anatomy learning like cross-sectional anatomy and surface anatomy.

CONCLUSION

The conclusion of this article is to use a mixture of the above teaching methods. Dissection is one of the gold standard modalities of teaching anatomy, but it is a fact that due to a lack of cadavers and teaching time in the laboratory, this mode of teaching cannot be used always. 3D software and plastic models are easy to use, cheap, and available at home during the study of anatomy and in future semesters to clear doubt regarding anatomy.

To conclude, all teaching modalities should be appointed to attain better educational goals.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.