VIRTUAL MICROSCOPY AND DIGITAL CYTOLOGY: FACT OR FANTASY?

How do young and senior cytopathologists interact with digital cytology?

In che modo i citologi giovani e quelli esperti interagiscono con la citologia digitale?

Maria Rosaria Giovagnoli^I; Enrico Giarnieri^I; Elisabetta Carico^I; Daniele Giansanti^II

^ISeconda Facoltà di Medicina e Chirurgia, Università Sapienza, Rome, Italy
^IIDipartimento di Tecnologie e Salute, Istituto Superiore di Sanità, Rome, Italy

Address for correspondence

SUMMARY

Today thanks to the technological advances in information technology the scenario of utilization of digital cytology has radically changed. New competitive systems, such as client-server architectures are now available in digital cytology. Their application in telemedicine should be investigated. A new interactive tool designed for the final destination user (the cytopathologist) has been proposed. Taking into account the different expertise of the subjects of the study, the investigation was focused both on the senior cytopathologist and on the younger student pathologist. The methodology was tested on 10 students of a Master in cytopathology and on 3 senior cytopathologists. The study showed that the use of digital cytology applications is effective and feasible for telediagnosis. In particular, the study on younger and senior expert investigators showed that, although they interacted with the novel technology of the virtual slide in a different manner, all of them reached the objective of a "correct diagnosis". This investigation, in consideration of the effectiveness of the digital cytology, also showed other indirect and tangible cost-beneft and quantitative advantages. In particular for the learning methodologies for the students of the Master itself and for the biomedical personnel involved in diagnosis.

Key words: telemedicine, digital cytology, tele-pathology.

RIASSUNTO

Oggi grazie ai progressi tecnologici nelle tecnologie dell'informazione e della comunicazione lo scenario dell'utilizzo della citologia digitale è radicalmente cambiato. Nuovi sistemi competitivi, come ad esempio le architetture client-server sono oggi disponibili in citologia digitale. Il problema dell' analisi di queste nuove tecnologie sta diventando pertanto un aspetto importante in telemedicina. Un nuovo strumento interattivo che si basa sulla figura dell'utente finale cito-patologo è stato proposto per questo scopo. In particolare, tenendo conto delle differenze nella pregressa formazione, la metodologia è stata focalizzata sia sul cito-patologo esperto che su quello giovane in formazione universitaria. La metodologia è stata testata su 10 studenti di un master in citopatologia e 3 esperti in citopatologia. Il lavoro ha dimostrato che l'uso delle applicazioni di citologia digitale è efficace e fattibile per la tele-diagnosi. In particolare, la ricerca ha dimostrato che nonostante i due gruppi interagissero con la tecnologia in modo diverso entrambi hanno raggiunto l'obiettivo della "corretta diagnosi". Questa ricerca in considerazione della provata efficacia della citologia digitale inoltre ha mostrato vantaggi tangibili. In particolare sono stati intravisti dei vantaggi costo-beneficio nell' ambito della formazione (nell'immediato nel master stesso) e per il personale addetto alla diagnosi nel laboratorio biomedico.

Parole chiave: telemedicina, citologia digitale, tele-patologia.

INTRODUCTION

Up to 5-10 years ago digital cytology was principally based on the design and construction of a few identical and expensive platforms with microscope units and software tools for the display and for the point to point tele-control (zooming, moving cutting of pieces of images). Today, thanks to the wonderful development of information technology, the diffusion of new visualization strategies and the availability of low cost or free visualization proprietary tools the scenario has radically changed [1-3]. New competitive systems as client-server architectures are now available. In a typical today's architecture there is an on site server with a scanner or alternatively a remote digital cytology third party server (as the web service such as the Leeds' centre (www.virtualpathology.leeds.ac.uk/index.php) for the creation of the slides and low cost (or free) user-friendly applications proprietary software tools for the navigation on the virtual slide which can be installed in remote clients. This new methodology especially the one based on a third party server is rapidly spreading and is becoming a useful means for:

PURPOSE

The investigation of the uses of new technologies is becoming a primary aspect of studies in telemedicine. The task to investigate the new technologies is becoming a core aspect in telemedicine. However very few studies have focused on this in digital cytology and mainly only on single aspects, and more importantly no one on the destination-user of the technology: cytopathologist.

The aim of the study is to:

MATERIALS AND METHODS

A workgroup that included the authors has designed two electronic questionnaires based on interactive forms. One of these was to investigate the acceptance of the technologies for scanning and navigating the virtual slide; the other to evaluate the acceptance of the digital cytologic diagnosis, both starting from a previous version centered in tele-pathology [9] (see in [9] for more details about the constructions of the forms).

Questionnaire for the acceptance of the digital cytology technology

The first questionnaire has been designed to investigate the software package tool itself. It has been organized into 6 sections. Table 1 details each section with the correspondent description. The second point concerning the methodology is centred on the figure of the cytopathologist: in spite of the fact that a wide range of objective quantitative methods of evaluations are available, the subjective aspects in the field of evaluation have always been considered the most important ones: it is on the opinion of the destination-users (based on their internal neural mechanisms) that is constructed the diagnosis and not on the basis of an objective tool for the image quality assessment. Figure 1 details a section (3) of the questionnaire (translated into English).

Table 2 shows in details the content of the section 5 "Requirements of the platform" for the standardization of the personal computer. The conditions of the test have been standardized according to this section.

Questionnaire for the acceptance of the digital cytologic diagnosis

The second questionnaire has been designed to investigate the acceptance of the digital cytologic diagnosis. Table 3 details the sections with the correspondent objective.

For the design of the second section it should be taken into account that in a virtual-slide a region can be univocally detected by means of the following three methods of the software ImageScope or Web Viewer designed by Aperio (Aperio, CA) (www.aperio.com):

a) three types of information: X, Y coordinates with the Zooming factor ( Image Scope viewer;

b) snapshot (usually a small size TIF image) (ImageScope Viewer) (Figure 2);

c) an internet link address pointer to the region of the slide (Spectrum Web Viewer) allowing the remote exchange of a selected image (Figure 3).

Selection criteria of virtual slides

Pap-test glasses have been chosen to be converted into virtual slides. The Scanscope XT equipment has been used for the digitalization.

Thin layer Pap-tests were used instead of the traditional ones The advantages of this choice were the small surface that needs to be scanned and thus a shorter time frame for the testing, homogeneity of the sample and the decreased amount of time for inspection by the cyopathologist. Other considered parameters have been the staining and the absence of artifacts or bubbles. Finally, samples with different diagnosis have been chosen (from negative to HSIL), in order to investigate by means of digital cytology different cytotypes and their nuclear and cytoplasmatic characteristics (physiological and not) and microbic flora.

Protocol and subjects

The selection was driven, as explained above, by the need to investigate subjects with a different expertise but who could be potential users of digital cytology.

Two groups were selected for the study:

- group A) 3 senior cytopathologists who had more than a 10 year working experience;

- group B) 10 students who were completing their Master in "diagnostic cytopathology and screening of population" after their Bachelor of Science or Medical Degree.

The defined protocol instruction was : assess the time spent to complete the diagnosis; select cytologic significant fields (optional). Time spent for selection was not calculated.

The protocol has been applied to 6 different scanned cytological glasses (virtual slides) chosen in consideration of the difficulty in the diagnosis. These slides came from 6 glasses with a full detailed diagnosis by standard microscopy, performed by other cytopathologists not involved in this investigation.

Architecture of the tele-cytology application

Different manufacturers were invited in this study. Aperio agreed to participate in this specific investigation. Figure 4 details the architecture:

RESULTS AND DISCUSSION

Above all the function allowing the extraction of an internet path as a pointer of an analyzed region of the tool Web-Viewer installed in the clients computers at the Sant'Andrea Hospital was found very useful. Comments, in this case, showed the utility of exchanging this path by email or even by a simple chat on the web. Moreover was found that, also with pc arranged by a very basic configuration (sufficient to run the Microsoft Office tools) different from the standardized one shown in Table 2, it was possible to navigate the virtual slide without too many delays.

All the subjects involved in the study reached the objective of a "correct diagnosis". Figure 4 shows some relevant snapshots from the virtual slides attached to the questionnaires. However younger and senior investigators related themselves with the digital cytology application in a different manner depending on their previous biological experience/knowledge and informatics background. Even if the younger investigators had more practice with navigation / informatics tools the experts, during comparison reached the target within a very short time (Table 4). Table 5 details the outcome from the different sections of the questionnaire about the acceptance of the tool.

It has been shown that all the items from Tables 1 and 3 received a high degree of acceptance. Younger investigators showed a high uniformity of responses in the section 3 on navigation. This intuitively depends on their previous and recent background of informatics courses specifically designed on image navigation [9] so that they had familiarity with similar navigation interfaces. When compared to senior investigators they showed a high variance in section 2 on the subjective perceived quality; this depended from the fact that not every subject has the same expertise thus the perceived quality is different. The experts assigned the lower scores to the questions of section 4 on the selection of details and on a specific question "speed of refresh" of section 3.

This can be explained by the fact that the subjects with higher expertise with traditional microscopy are not used to the latency time due to the "speed of refresh".

Experts showed a high uniformity of responses in the section 4 on selection of details. Section 5 was not submitted to the investigators but compiled by the authors of the study.

Statistical significance

The statistical significance of the two outcomes from the questionnaires has been also investigated using two different methodologies Wilkoxon and Anova.

The first one (Wilkoxon) has been used to investigate the significance of the difference between the two groups (students, experts).

The second one (Anova) has been used to investigate the significance of the data obtained by means of the two tests.

For the investigation two different software tools have been used:

The significance obtained by means of the Wilkoxon test used to investigate the difference in the time used for scanning the virtual slide between the two groups was > 99%.

The significance obtained by means of the Wilkoxon test used to investigate the differences in the responses provided by the two groups in the questionnaire designed to investigate the acceptance of the technology was always > 95% as shown in Table 6.

Anova showed a significance always > 95% for the investigated data.

CONCLUSIONS

The work conducted at the Sant'Andrea Hospital in Rome showed that the use of digital cytology applications is effective and feasible for tele-diagnosis. The investigation based on the proposed methodology on younger and senior expert investigators showed that, although they interacted with the novel technology for the virtual slide in a different manners, they all reached the target "correct diagnosis". This investigation on the effectiveness of tele-cytopatology also showed other indirect and tangible cost-benefts and quantitative advantages especially for the learning methodologies [4-8] for the students of the Master itself and for the biomedical personnel involved in diagnosis, in details:

Conflict of interest statement

There are no potential conflicts of interest or any financial or personal relationships with other people or organizations that could inappropriately bias conduct and findings of this study.

References

1. Demichelis F, Barbareschi M, Boi S, Clemente C, Dalla Palma P, Eccher C, Forti S. Robotic telepathology for intraoperative remote diagnosis using a still-imaging-based system. Am J Clin Pathol 2001;116(11):744-52.         

2. Dalla Palma P, Morelli L, Forti S, Barbareschi M, Piscioli F, Galvagni M, Boi S, Demichelis F. Telemetric intraoperative diagnosis among hospitals in Trentino: first evaluations and optimization of the procedure. Pathologica 2001;93(1):34-8.         

3. Demichelis F, Barbareschi M, Dalla Palma P, Forti S. The virtual case: a new method to completely digitize cytological and histological slides. Virchows Arch 2002;441(2):159-64.         

4. Giansanti D, Castrichella L, Giovagnoli MR. Telepathology training in a master of cytology degree course. J Telemed Telecare 2008;14(7):338-41.         

5. Giansanti D, Castrichella L, Giovagnoli MR. E-learning for laurea in biomedical laboratory technicians: presentation of a pilot study. Conf Proc IEEE Eng Med Biol Soc 2007;11:3745-7.         

6. Giansanti D, Castrichella L, Giovagnoli MR. How do young and senior cytopathologists interact with tele-cytopathology applications? Procedings of the 35^th Congress of cytology. Lisboa: Blackwell publishing; 2009. 20 (Suppl. 1) p. 470.         

7. Giansanti D, Castrichella L, Giovagnoli MR. New models of e-learning for healthcare professionals: a training course for biomedical laboratory technicians. J Telemed Telecare 2007; 13(7):374-6.         

8. Giansanti D, Castrichella L, Giovagnoli MR. Tele-pathology requires specific training for the technician of the biomedical laboratory. Telemed J E Health 2008;14(8):801-7.         

9. Giansanti D, Castrichella L, Giovagnoli MR. The design of a health technology assessment system in telepathology. Telemed J E Health 2008;14(6):570-5.         

Address for correspondence:
Maria Rosaria Giovagnoli
Università Sapienza, Seconda Facoltà di Medicina e Chirurgia
Azienda Ospedaliera Sant' Andrea, Via di Grottarossa 1035-1039
00189 Rome, Italy.
E-mail: mariarosaria.giovagnoli@uniroma1.it

Submitted on invitation.
Accepted on 5 May 2010.