Perceptions on the Use of an Online Tool in the Teaching-learning

Process in Microscopy

Breno N. S. Keller

1 a

, Mariana T. Rezende

2 b

, Tales M. Machado

1 c

, Saul Delabrida

1 d

Claudia M. Carneiro

2 e

and Andrea G. C. Bianchi

1 f

Computing Department, Federal University of Ouro Preto, Ouro Preto, MG, Brazil

Clinical Analysis Department, Federal University of Ouro Preto, Ouro Preto, MG, Brazil

Keywords:

Online Learning, Cytology, Microscopy, Evaluation.

Abstract:

During the COVID-19 pandemic, remote learning was an alternative to maintaining student participation in

subjects, active learning, and knowledge development. This approach is necessary for the experimental de-

mands of the practical content of the Cervical Cytology class. This paper presents and discusses using an

online platform to learn practical content in the microscopy subject of Cervical Cytology class. The evaluated

scenarios demonstrated that the planning of the discipline and personal factors such as student interest and

availability could inﬂuence the student performance.

1 INTRODUCTION

Digital native is a term deﬁned by (Prensky, 2001)

to refer to a generation who were born in a world

greatly permeated by technology. That means they

expect digital tools to constantly mediate their inter-

actions with the world. Such assumption implies that

even contents covered in the classroom need to be

closer to their digital daily life to keep students inter-

ested and motivated. Therefore, there is a formal ex-

pectation that education at all levels (university, high

school, and elementary) incorporate educational tech-

nologies in their execution resulting in practices of

Digital Technologies of Information and Communica-

tion (Bahia et al., 2019). This approach aims to facil-

itate, enhance learning, and include innovative prac-

tices in education.

Thus, computational tools are necessary to ad-

dress that vision in several areas of knowledge since

these tools allow the learning process to be modiﬁed

to suit the student or explore different teaching meth-

ods for the same topic.

(Guze, 2015) classify these tools in the following

https://orcid.org/0000-0001-5414-6716

https://orcid.org/0000-0002-9514-9312

https://orcid.org/0000-0003-0603-823X

https://orcid.org/0000-0002-8961-5313

https://orcid.org/0000-0002-6002-857X

https://orcid.org/0000-0001-7949-1188

categories: computer-assisted learning, mobile de-

vices, digital games (or serious games), simulations,

and wearable equipment. Moreover, despite the sim-

ilarity between these tools due to the computational

resources used, each represents a different interaction

model, which explores users’ different senses and per-

ceptions, allowing access to activities analogous to

the experimental ones in the case of remote learning.

In digital pathology and cytopathology, these ap-

proaches range from different scholar levels, from in-

side university classrooms to conferences and spe-

cializations (Maley et al., 2008; Wiecha et al., 2010;

Bahia et al., 2019; Guiter et al., 2021). The technol-

ogy allowed a signiﬁcant transformation in the pro-

cess with the digitization of microscopic slides to gen-

erate images of whole slides (Whole slide imaging

(WSI) or digital slides), which can be manipulated by

the operator and used in the learning context (Hanna

et al., 2020; Guiter et al., 2021). However, such sys-

tems require high-value equipment and are not easy

to access. In addition, users (students) also need qual-

ity internet access and good computers to use these

resources well.

This work proposes a digital tool to support the

implementation of remote model microscopy subjects

and the discussion and evaluation of such interaction.

Hence two remote teaching scenarios of a cytology

discipline were implemented using a support tool for

assessing students. These scenarios were performed

during the COVID-19 pandemic in 2020 and 2021.

Keller, B., Rezende, M., Machado, T., Delabrida, S., Carneiro, C. and Bianchi, A.

Perceptions on the Use of an Online Tool in the Teaching-learning Process in Microscopy.

DOI: 10.5220/0011062200003179

In Proceedings of the 24th International Conference on Enterprise Information Systems (ICEIS 2022) - Volume 2, pages 325-331

ISBN: 978-989-758-569-2; ISSN: 2184-4992

 2022 by SCITEPRESS – Science and Technology Publications, Lda. All rights reser ved

325

This work is structured as follows: Section 2

presents some related works; Section 3 presents a

contextualization of the class used as the basis; Sec-

tion 4 introduces the digital the tool used; Section 5

presents the evaluated scenarios; Section 6 describes

the results observed and Section 7 presents the con-

clusions of this work.

2 RELATED WORK

(Maley et al., 2008) propose a computational tool

used to assist teaching. The authors evaluate how a

web system helps in the pathology learning process.

It was observed how these students behave in the pro-

posed scenario and how this impacts their learning

process. One behavior observed is that users prefer

a face-to-face setting and are not fulﬁlling the appli-

cation’s response deadlines during the school term.

However, only a small group of students maintained

this behavior. That was associated with individuals’

characteristics, who focus on the result as a study

strategy.

Another use of computational resources as teach-

ing tools was shown by (Wiecha et al., 2010) where

they discuss the use of the Second Life (SL) platform

for teaching and training. The SL is a virtual envi-

ronment focused on providing communication and in-

teraction resources to users. The authors created a

test case based on a seminar on the use of insulin for

patients with type 2 diabetes. The scenario allowed

participants to be more conﬁdent about evaluating the

need to apply the discussed treatment after participat-

ing in the dynamics. In addition, participants also re-

ported that the interaction model was as good or su-

perior to a traditional face-to-face model.

(Krasne et al., 2013) shows a combination of

learning tools and algorithms, which presents a sys-

tem to aid the study of histopathology based on a

web system, which implements an adaptive learning

model. The user is provided with a study scenario

that presents different cases related to the course con-

tents to improve students’ long-term content reten-

tion. Moreover, the system presents a case study and

options to identify which type of problem (injury) the

case study addresses. Furthermore, the order in which

the case studies are delivered to the user is constructed

using an algorithm that learns from the user’s perfor-

mance. It was observed how students’ content reten-

tion was through their reassessment in later periods,

which showed that users who used the system had

greater content retention.

(Darici et al., 2021) report the implementation of

a histology course in a remote model for the context

of the COVID-19 pandemic. Students from two dif-

ferent academic periods (second and third) were eval-

uated in the scenario developed. These students were

introduced to the course content and performed inter-

active activities during remote classes. At the end of

the course, the students’ performance was evaluated

as 71% (second period) and 74% (third period). How-

ever, the students’ evaluation of the model was posi-

tive, on a 100-point scale with 1 being “very good”

and 100 being “very bad” students reported a median

of 21 (second semester) and 22.5 (third semester).

The works described in this section show that the

remote teaching-learning process is possible and fea-

sible. However, they present a great variety in how

they work and are applied, consequently impacting

their performance.

3 CERVICAL CYTOLOGY

The Cervical Cytology class aims to present to un-

dergraduates the dimension of cervical cancer and

how screening for this neoplasm is essential to ensure

women’s health. This class aims to enable students

to develop logical, critical, and analytical reasoning

in the face of cervical cancer screening. All screen-

ing steps are characterized, from the collection of the

material to the release of the diagnosis report, includ-

ing staining techniques, assembly, reading of the cyto-

logical smear, interpretation, quality monitoring, and

report. In addition, there is content focused on the

physiology of the female reproductive system, his-

tology, cytology of the cervix, hormones, menstrual

cycle, inﬂammatory processes, microbiota, and cyto-

morphological criteria of malignancy.

Most of the concepts and contents presented in the

class are linked to identifying and recognizing ele-

ments in the Pap smears when viewed under an op-

tical microscope or in images originated from them.

Traditionally, the professor explains the content by

exposing many images illustrating cervical-vaginal

smears, infectious agents, microbiota, inﬂammatory

processes, cytological changes, and other important

biological structures to understand the theory studied

fully. So, the visualization of images within this disci-

pline’s teaching is crucial for better student learning.

The course evaluation is split into two parts: theo-

retical and practical. The practical classes of the dis-

cipline are carried out under an optical microscope

for visualization of cytology slides after each theo-

retical class, covering the topics and biological struc-

tures taught for a better understanding of the students.

Thus, each student has an optical microscope and

slides available to visualize the subjects teach in the

ICEIS 2022 - 24th International Conference on Enterprise Information Systems

326

theoretical classes and are present in the slide. Also,

the practical classes are accompanied by the teacher

and allow the students to look at different examples

equivalent to multiple slide images over each lecture’s

content.

The periods evaluated were remote, so applying

the traditional practical test under an optical micro-

scope was impracticable. Therefore, the tool pre-

sented in Section 4 was essential to assess students re-

garding the visual requirements required in the practi-

cal test, which are fully linked to the theoretical con-

tent taught. In addition to the tool presenting a content

review, it was responsible for learning tests.

4 THE SYSTEM

This section describes the system used during the

class as a support tool. The computational system was

constructed based on the interaction process for iden-

tifying and recognizing the elements present in the

Pap smear. The students perform these activities in

optical microscopes during classes and practice tests.

Thus, the online platform was built to replicate the in-

teraction and experience from those situations. More-

over, due to accessibility, the system was designed to

be used in mobile device browsers. Also, the system

interface is in the students’ native language. Figure 1

shows screens of the implemented system.

Figure 1a presents the list of exercises (or activi-

ties) that students must answer on the topics studied

in the theoretical class. When the user clicks on one

exercise of the list, the system shows a screen similar

to Figure 1b. The image to be evaluated is presented

with the regions of interest (denominated questions)

to be identiﬁed and named. Multiple regions of inter-

est in the same image make up the activity. The stu-

dent is also able to zoom in on the image to see more

details of the analyzed region, as shown in Figure 1c.

Finally, the student clicks on the region, and multi-

ple options are presented to select which alternative

best describes the highlighted (or marked) region, as

shown in Figure 1d. The student has prompt feedback

about his answer and can view the correct answer for

a region if he has selected the wrong option. The ex-

ercise is completed and corrected by answering all the

questions in the image.

Besides, considering the return of face-to-face ac-

tivities, this tool can also be used in the routine of

practical classes to establish knowledge, not only in

cytology but in any subject involving microscopy.

It is necessary to point out that the presentation

and description of this tool are not essential for this

article, but their use in the process of interaction in

(a) Exercise List. (b) Evaluated Image.

tives.

Figure 1: System Screens.

online pathology classes. To facilitate understanding,

in the next section, we will describe the experimenta-

tion scenario used to evaluate the interaction between

students and their performance in the tool.

5 TEST SCENARIOS

In this section, the observed scenarios are described.

They correspond to two approaches of remote learn-

ing activities carried out during the COVID-19 pan-

demic between 2020 and 2021. In both scenarios, the

system described in Section 4 was used, in which 87

activities and a total of 249 questions were consid-

ered.

Perceptions on the Use of an Online Tool in the Teaching-learning Process in Microscopy

327

5.1 Scenario 1: Shorter Academic

Period

The ﬁrst scenario corresponds to a model of a reduced

academic period, designed as an alternative for re-

motely carrying out the subject during the pandemic.

The class was developed in a shorter time than the tra-

ditional one, in this case, nine weeks instead of eigh-

teen weeks. Thus, in addition to the remote execution

model, it was also a faster and more compact model

than the traditional one. The classes were four times

a week instead of twice. However, the participation

of students in this period was optional, so the student

enrolled in the course had some interest or motivation

in carrying out this process.

As for the course dynamics, system access, and

evaluation, the students had only one deadline at the

end of the school term. This deadline includes all the

activities of the system.

5.2 Scenario 2: Regular Academic

Period

The second scenario corresponds to a regular model

of academic terms carried out remotely. In this case,

the period comprises fourteen weeks, and the partici-

pation of students in the period was mandatory.

In this scenario, the dynamics, system access, and

evaluation were divided into two deliveries: around

the ﬁrst half of the course, which means around 50

days, and the other half at the end of the school term.

6 RESULTS AND DISCUSSIONS

This section presents the results observed in the sce-

narios described above. Only the data from students

who have completed the course are considered, that

is, those who have completed the discipline.

6.1 Scenario 1: Shorter Academic

Period

In Scenario 1, 12,164 responses were obtained from

50 students (97.70% of all possible answers, 98.11%

of all possible activities), which resulted in an aver-

age of 243.28 responses per student. Students had an

average hit rate of 85.47% concerning answered ques-

tions and 83.51% concerning all available questions.

Figure 2 shows the distribution of student re-

sponses during the period considering the class. In-

terestingly, most responses are concentrated in the last

two weeks of the period. However, some answers are

in the middle of the observed period, which corre-

sponds to the end of complementary activities to the

discipline.

Figure 2: Accumulated Activities Answered in Scenario 1.

Figure 3 shows the distribution of student re-

sponses during the period. Students with less than

70% hit rate were considered outliers and excluded

from the analysis (resulting in 45 students out of 50),

the remaining students’ results were divided into 3

groups of size 15 based on their performance. Thus,

the group 1 corresponds to students with a hit rate

between 100% to 89.5%; group 2 corresponds to the

correctness range from 89.5% to 83.5% and group 3

corresponds to the correctness range from 83.5% to

75%. Each color in the Figures 3a, 3b and 3c repre-

sents a student and in case of overlap only one colored

point is shown.

By observing the Figures 3a, 3b and 3c, we can

identify two behaviors among students: (I) resolution

of activities throughout the course and (II) resolution

of activities at the end of the course. These behaviors

occur in all groups in different proportions. However,

the behavior II seems to be connected to better stu-

dent performance. Given that in group 1, which con-

tains the users with the best success rates, at least two-

thirds of the students did the activities at the end of the

course, and in one or two interactions, they answered

87 activities (247 questions) in two days.

6.2 Scenario 2: Regular Academic

Period

In the second scenario, 2,113 responses were ob-

tained from 9 students (94.29% of all possible an-

swers, 90.42% of all possible activities), resulting in

a rate of 234.78 responses per student. On average,

these students had a hit rate of 80.69% concerning

what they answered and 70.08% concerning all ac-

tivities.

Figure 4 shows the distribution of student re-

sponses during the academic period. It is interesting

ICEIS 2022 - 24th International Conference on Enterprise Information Systems

328

(a) Group 1.

(b) Group 2.

Figure 3: Students Complete Activity Distribution in Sce-

nario 1.

to note that, as in this scenario, two delivery dates

were made for the activities (50 and 84 days elapsed

since the beginning). The distribution of responses is

concentrated around these two dates.

Figure 5 shows the distribution of student re-

sponses during the period. The responses are clus-

tered around the 50th (day of the ﬁrst delivery) and

84th (second delivery), with a smaller concentration

in the last one. However, the number of responses

decreased in the second delivery. This behavior can

be related to the student’s feeling of having already

passed the discipline and redirecting his efforts to

other courses.

Figure 4: Accumulated Activities Answered in Scenario 2.

Figure 5: Students Complete Activity Distribution in Sce-

nario 2.

6.3 Discussion about Both Scenarios

In general, the methodology utilized in Scenario 1 al-

lowed it to have the highest success rate. This behav-

ior can be associated with the fact that the students an-

swered all the activities at the evaluation period end.

The course structure favors that behavior as the sub-

ject is passed to the students by cumulative exposi-

tion. This increases the students’ ability to assess the

elements on the images over the period. In this way,

students could assimilate more information about the

structures and the exposure during classes.

Figure 6 presents a summary of user success rates

in both scenarios. Considering the ﬁrst delivery of

Scenario 2 as a splitting point, the data was separated

into two halves (the ﬁrst and the second delivery of

Scenario 2). In the ﬁrst half, it is observed that both

scenarios were obtained answers to almost all ques-

tions. However, the hit rates are both relative to what

was answered, and the total is proportional to each

other.

In the second half, there is again a lower delivery

rate of activities by students in Scenario 2. However,

the success rates relative to what was answered by the

students is high, on the order of 80%. The low de-

livery of activities reinforces the hypothesis that stu-

dents preferred not to respond to the activities, either

because they have already passed the subject or have

Perceptions on the Use of an Online Tool in the Teaching-learning Process in Microscopy

329

Figure 6: Answers Distributions in Both Halves of the Test

Scenarios.

other priorities (other courses). The observed hit rates

indicate they knew the subject; however, they pre-

ferred not to answer the activities.

Figure 7 presents the confusion matrix of the stu-

dents’ answers for the main concepts addressed in the

discipline. A confusion matrix is a form of visualiza-

tion in which the columns represent the instances’ ac-

tual category while each row represents the instances’

answered category. Since the activities contain mul-

tiple alternatives, they were categorized into four the-

matic groups based on their concepts, these groups

being: Artifacts (1), Epithelial Cells (2), Inﬂamma-

tions (3), and Lesions (4). The Artifacts group gath-

ers the alternatives for materials foreign to the cytol-

ogy sample, such as dye precipitate and bubbles. The

Epithelial Cells group gathers alternatives represent-

ing the cells collected from the cervical epithelium.

The Inﬂammations group gathers alternatives asso-

ciated with the cellular changes that indicate an in-

ﬂammatory process associated with some infectious

agents. And Lesions group gathers the alternatives

representing the classiﬁcations for different changes

in cervical epithelial cells. Also, if a student answered

a wrong alternative but the alternative was in the same

group as the correct answer, it will be considered a

“hit” for this graph since the user had a small concep-

tual error.

In Figure 7a the confusion matrix for Scenario 1

is presented, and it is interesting to observe that users

confuse responses from category 1 with category 3.

However, the reverse process almost does not happen.

The results may indicate a learning effect during the

course span since the group 3 alternatives are content

from the middle of the course while the category 1 al-

ternatives are from the beginning. There is also some

confusion between category 2 and 3 alternatives, in-

ﬂammations, and lesions.

Figure 7b presents the confusion matrix of the re-

sponses obtained for Scenario 2. Like Scenario 1,

there are also errors between categories 1 and 3, but

(a) Scenario 1.

(b) Scenario 2.

Figure 7: Confusion matrices for responses categories.

with a volume greater than occurred in the ﬁrst sce-

nario. This may be a consequence of carrying out

the activities in two stages since, in this scenario, the

students had less exposure to the contents when they

were answering the ﬁrst half of the activities (com-

pared to Scenario 1), which may have caused this

higher error rate. Similar to the ﬁrst scenario, there

is also some confusion between category 2 and 3 al-

ternatives.

As categorized above, the response options were

generalized into four groups. However, the user may

have answered some alternative in the same category

as the correct answer, which marks the answer as in-

correct. The table 1 presents the hit and miss rates

within each category. Interestingly, there is a constant

error rate above 7% across all categories and scenar-

ios. Furthermore, in Scenario 2, there is a higher error

rate compared to Scenario 1.

ICEIS 2022 - 24th International Conference on Enterprise Information Systems

330

Table 1: Categories’ hit and miss rate.

Scenario 1 Scenario 2

Category Hits Miss Category Hits Miss

Artifacts 92.78% 7.22% Artifacts 89.04% 10.96%

Epithelial Cells 91.37% 8.63% Epithelial Cells 86.54% 13.46%

Inﬂammations 91.92% 8.08% Inﬂammations 90.91% 9.09%

Lesions 92.67% 7.33% Lesions 86.94% 13.02%

7 CONCLUSIONS

The COVID-19 pandemic has created the need to in-

corporate remote education experiences into the cer-

vical cytology discipline quickly. This proposal de-

veloped a framework that allowed students to interact

with images and subjects similar to those in a practi-

cal class. Also, we evaluated how different classroom

scenarios impacted student performance.

The data collected demonstrate that in Scenario

1 a better average performance of students was ob-

tained. This performance can be associated with

both the discipline’s methodology who performed in

less time with more condensed content, and students’

commitment since participation in the term contain-

ing Scenario 1 was optional. Also, it is important to

emphasize that the two scenarios present a difference

in the participants’ composition, motivations, and the

number of students in each scenario. So some differ-

ences in the performance are related to that.

The virtual environment allows the student to

choose what and when to do it. For example, he can

organize himself, focusing on more difﬁcult content

(individualization of learning). Although the student

is not in a real laboratory, the diagnostic context is

covered and the situations created are very close to

routine. Another clear point in remote teaching is that

even if the laboratory does not have slides with a spe-

ciﬁc lesion, it is possible to create a repository with

different images from different locations and the most

varied lesions. Thus, it allows the student to see, de-

tect and learn in the widest possible way.

Finally, it is noteworthy that the external factors

also impact in the teaching-learning process (and con-

sequentially, their performance). Some of the exter-

nal factors include the time, the performance on other

subjects or disciplines studied parallel and personal

context.

Further experimental investigations are needed to

better estimate the impact of this learning model. One

improvement is using the system in more academic

periods to obtain more reliable data. Moreover, the

system can be improved to make it easier to access

and better interact with the different actors (students

and teachers) without impacting the model’s perfor-

mance.

ACKNOWLEDGMENTS

The authors would like to thank the Coordenac¸

de Aperfeic¸oamento de Pessoal de N

ıvel Superior

- Brasil (CAPES) - Finance Code 001, Fundac¸

de Amparo

a Pesquisa do Estado de Minas Gerais

(FAPEMIG), Conselho Nacional de Desenvolvimento

Cient

ıﬁco e Tecnol

ogico (CNPq) for supporting the

development of the present study. Also like to thanks

the Universidade Federal de Ouro Preto (UFOP),

the Center for Recognition and Inspection of Cells

(CRIC), the Extended Reality for Good Laboratory

(XR4Good) and the Laborat

orio Multiusu

arios de Mi-

croscopia Avanc¸ada e Microan

alise do N

ucleo de

Pesquisas em Ci

encias Biol

ogicas (NUPEB) for also

supporting this research.

REFERENCES

Bahia, N. S., da Silva, W. R., Vianna, J. B., Rodrigues,

H. G., Silva, M. T. B., and Bacchi, R. R. (2019). O

uso das tdic’s como estrat

egia para aprendizagem em

morfologia microsc

opica. Inform

atica na educac¸

ao:

teoria & pr

atica, 22(2).

Darici, D., Reissner, C., Brockhaus, J., and Missler, M.

(2021). Implementation of a fully digital histology

course in the anatomical teaching curriculum during

covid-19 pandemic. Annals of Anatomy-Anatomischer

Anzeiger, 236:151718.

Guiter, G. E., Sapia, S., Wright, A. I., Hutchins, G. G.,

and Arayssi, T. (2021). Development of a remote on-

line collaborative medical school pathology curricu-

lum with clinical correlations, across several interna-

tional sites, through the covid-19 pandemic. Medical

Science Educator, 31(2):549–556.

Guze, P. A. (2015). Using technology to meet the challenges

of medical education. Transactions of the American

clinical and climatological association, 126:260.

Hanna, M. G., Reuter, V. E., Ardon, O., Kim, D., Sirin-

trapun, S. J., Sch

ufﬂer, P. J., Busam, K. J., Sauter,

J. L., Brogi, E., Tan, L. K., et al. (2020). Validation

of a digital pathology system including remote review

during the covid-19 pandemic. Modern Pathology,

33(11):2115–2127.

Krasne, S., Hillman, J. D., Kellman, P. J., and Drake, T. A.

(2013). Applying perceptual and adaptive learning

techniques for teaching introductory histopathology.

Journal of pathology informatics, 4.

Maley, M. A., Harvey, J. R., Boer, W. B. d., Scott, N. W.,

and Arena, G. E. (2008). Addressing current problems

in teaching pathology to medical students: blended

learning. Medical teacher, 30(1):e1–e9.

Prensky, M. (2001). Digital natives, digital immigrants. on

the horizon. mcb university press. 9 (october).

Wiecha, J., Heyden, R., Sternthal, E., and Merialdi, M.

(2010). Learning in a virtual world: experience with

using second life for medical education. Journal of

medical Internet research, 12(1):e1.

Perceptions on the Use of an Online Tool in the Teaching-learning Process in Microscopy

331