Real-time collaborative environment for radiation treatment planning virtual simulation

1444 IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 49, NO. 12, DECEMBER 2002

Real-Time Collaborative Environment for RadiationTreatment Planning Virtual Simulation

Efthymios Ntasis, Theofanis A. Maniatis, Member, IEEE, and Konstantina S. Nikita*, Senior Member, IEEE

Abstract—A virtual simulation (VS) system is a softwaresystem that enables the delineation of anatomical structures, andthe placement of irradiation fields for the purpose of radiationtreatment planning, making use of patient’s tomographic data,instead of the real patient. Since patient’s tomographic data canbe communicated between distinct radiotherapy departments,collaborative work on VS, connecting remote health care profes-sionals, becomes feasible. In this paper, an environment enablingreal-time collaboration on VS is presented. The environmentarchitecture is based on both offline and online communicationof data under a secure framework and can be directly integratedinto the infrastructure of a radiotherapy department. The onlinecollaboration relies on the simultaneous execution of all actions atboth collaborating sites, and prerequisites the offline communica-tion of the data set on which the collaboration will be performed.Analytical description of the custom-made layered service, whichsupports the offline communication is given, along with a detailedpresentation of the secure management of messages, which enablesthe real-time collaboration. The technical evaluation of the envi-ronment highlights the effectiveness of the proposed methodology,since real-time secure collaboration on VS is achieved.

Index Terms—Collaborative work, radiation treatment plan-ning, virtual simulation.

I. INTRODUCTION

RADIATION therapy aims to tailor a dose envelope to atarget volume, while avoiding the delivery of high radia-

tion dose to normal tissue [1]. Radiation therapy methods can becategorized into two major groups: external beam radiotherapy,where radiation produced by a radiation therapy unit is deliveredto the treatment volume, and brachytherapy, where radiation isdelivered at a short distance by radioactive sources, positionedat the vicinity of the treatment volume. The efforts of health careprofessionals (i.e., radiation oncologists and medical physicists)during the last decade have raised radiation therapy to the mosteffective nonsurgical treatment modality [2].

The process of treatment planning, involves the determina-tion of treatment parameters considered optimal in the manage-ment of patient’s disease. In external beam radiotherapy, whichis the subject of this paper, treatment planning is a procedurecomprising a number of activities, which follow from the de-

Manuscript received November 30, 2001; revised December 4, 2001. Thiswork was supported by the Greek General Secretariat of Research and Tech-nology, under the EPET II Programme.Asterisk indicates corresponding author

E. Ntasis and T. A. Maniatis are with the National Technical University ofAthens, Department of Electrical and Computer Engineering, Zografos 15780,Athens, Greece.

*K. S. Nikita is with the National Technical University of Athens, Depart-ment of Electrical and Computer Engineering, Zografos 15780, Athens, Greece(e-mail: [email protected]).

Digital Object Identifier 10.1109/TBME.2002.805450

cision to treat a patient with radiotherapy and from the statedobjectives of that treatment [3]. The entire procedure, called ra-diation treatment planning (RTP), is of high complexity, and theuse of dedicated hardware and specialized software by health-care professionals is required, in order to achieve a reasonable,or sometimes optimal treatment plan. The final product of thisactivity is a blueprint for the treatment, to be followed meticu-lously and precisely over several weeks.

The activities involved in the RTP procedure include: local-ization of the target volume, design of beams and field shaping,dose calculation, plan optimization and evaluation before thefinal verification of the treatment plan [4], [5]. Conventionaltechniques, make use of the treatment simulator, which is anapparatus that uses a diagnostic X-ray tube with duplicated ge-ometrical, mechanical and optical properties, in relation with aradiation treatment unit. The treatment simulator is mainly usedfor the following two purposes: firstly to determine the posi-tion of the target volume, and second to display the treatmentfields so that the target volume may be accurately encompassed,without delivering excessive irradiation to surrounding healthytissues. These goals are achieved using real-time fluoroscopicimaging of internal organs leading the physician to the correctpositioning of fields and shielding blocks, in relation to externallandmarks [6].

Alternatively, the simulator machine can be replaced by avirtual simulation (VS) system [7], [8], a software system thatmakes use of patient’s tomographic (CT or MRI) data, insteadof the real patient, enabling health care professionals, to delin-eate the anatomical structures, and to place the irradiation fields.The results of the design on VS can then be passed to dose cal-culation routines, and to dose distributions’ evaluation and op-timization modules. Under this application scenario, the use oftreatment simulator machine is limited to the verification proce-dure of complex cases, adding more confidence to the planningprocess.

Even though the availability of three-dimensional (3-D) vi-sualization techniques as well as image processing tools facili-tates the design of the plan using VS, the complexity of definingan optimal treatment plan remains high, resulting to scenarioswhere a “second opinion” is needed. At this point, the bene-fits from the employment of VS in the RTP procedure are high-lighted, since the “virtualization” of the complete RTP proce-dure, does not only avoid patient’s transfers between differentdepartments and health care institutions, but also enables theavailability of the virtual patient on different sites than thoseof its physical location [9], [10]. Thus, most of the data re-quired in the RTP procedure, treatment execution and verifica-tion can be transferred and stored digitally, based upon standards

0018-9294/02$17.00 © 2002 IEEE

NTASIS et al.: REAL-TIME COLLABORATIVE ENVIRONMENT FOR RADIATION TREATMENT 1445

Fig. 1. GALENOS interface for the definition of anatomical structures.

and data formats, such as the Digital Imaging and Communica-tions in Medicine (DICOM) protocol [11]. Moreover, puttinginto practice the continuous advances in telecommunications,which have vastly contributed in the establishment of tele-radi-ology [12]–[16] networks, real time information flow can over-pass the limits of the local area network (LAN) of a clinic, andcollaboration between institutions can take place. In this way,with the introduction of VS, a wide area of possible telematicsapplications in radiation therapy has been initiated [17], [18].

In this paper, the VS system called GALENOS that supportssecure online tele-collaboration between health care profes-sionals is presented. A general overview of the functionalitiesprovided by the stand-alone VS system is given, followed bythe architecture description of the collaborative environment,and the evaluation results of the technical pilot study.

II. V IRTUAL SIMULATION

The aim of VS is to provide health care professionals withthe required tools in order to carry out target volume and othercritical organs delineation, as well as irradiation fields’ place-ment. Furthermore, the interoperability of a VS system with to-mographic scanners and Picture Archiving and CommunicationSystems (PACS), from which it receives data, and with the dosecalculation routines, to which the results of the design are sent,must be assured.

The connectivity of GALENOS VS with devices foundin a radiation oncology department is accomplished usingthe DICOM standard. More specifically, GALENOS VS iscapable of importing tomographic data in DICOM format,while data output, concerning anatomical structure description,beam placement and radiotherapy related images, is encodedaccording to the RT extensions of the DICOM standard. Thus,GALENOS VS consists an adaptable component to the LANof the radiation oncology department without any need forspecialized configurations.

The design architecture of the stand-alone VS system is per-formed in accordance with the dual aim of VS, hence two in-terfaces are incorporated: one dedicated to anatomical structuredefinition, and one dedicated to beam definition. Both of theseinterfaces interact with the GALENOS database, where all ofthe required data are stored.

A. Anatomical Structure Definition

Anatomical areas in the patient’s body, which are significantfor the RTP process including the tumor area [19], neighboringorgans at risk, as well as the external contour of patient’s bodyare characterized as volumes of interest (VOIs). The anatom-ical structure description consists of the definition of VOIs andthe localization of external markers, placed during the CT ex-amination, and used for the repositioning of the patient in thetreatment unit and/or in the simulator machine. The definition


Fig. 2. GALENOS interface for beam placement.

of VOIs takes place on tomographic (CT or MRI) images usinga set of automatic and semiautomatic contouring functionalities[20], [21]. In addition, the structure delineation process is facili-tated by basic image processing tools [22] and a 3-D interactivedisplay of defined VOIs based on surface rendering techniques[23], as shown in Fig. 1.

B. Beam Definition

Beam description consists of the set up of beams’ direction,beams’ placement with respect to a patient-based fixed coordi-nate system, and fields’ size and shape. During the interactiveprocedure of beam definition, all treatment machine motions,including collimator and couch angle are simulated, while thegeometric characteristics of the specific treatment units, whichare stored in GALENOS database, are taken into considera-tion. Although the default settings for the movements and scalesof treatment units simulated by GALENOS VS are in accor-dance with the IEC standard [24], user defined settings are alsosupported.

The visualization tools provided for the description of beamdefinition are Beam’s Eye View (BEV) images, 3-D views, 2-Dimages in axial and nonaxial planes, and a virtual view of thetreatment unit (Fig. 2). The design of field’s size and shape ismade using BEV images, where the perspective projections ofthe VOIs are shown, superimposed to the digital reconstructedradiograph (DRR). The extraction of DRR is performedusing ray casting [25], while a preview of DRR is availablefor real-time visualization, using Fourier volume rendering[26]–[28]. The provided 3-D views contain the results of either

volume [29] or surface rendering techniques [23], and serve asa tool for evaluating beam placement with respect to tumor andcritical structures. Tomographic images and reconstructed im-ages defined on nonaxial (sagittal, coronal, and oblique) planes,displaying superimposed beams projections and anatomicalstructures boundaries, are used for a detailed presentation ofirradiated areas. Finally, the virtual view of the treatment unitwith the patient in treatment position augments health careprofessional’s comprehension regarding machine settings.

Tools for the automatic placement of treatment unit isocenterto the center of VOIs (geometric or center of mass) and to patientskin are available, while the functionality of automatic adapta-tion of field boundaries, is provided to health care professionals,in order to accurately encompass any specific VOI.

III. COLLABORATIVE ENVIRONMENT

The collaborative environment of GALENOS VS extends thecapabilities of the stand-alone VS, giving the opportunity to re-mote health care professionals to cooperate for the delineationof VOIs and for the placement of irradiation fields. The imple-mentation is based on a point-to-point communication scheme(Fig. 3) and the main concept of the environment is to synchro-nize the two VS systems, in order to provide the collaboratingparties with the same view of the application [30], [31]. The in-dispensable condition for the realization of a collaboration ses-sion is that both collaborating parties possess the same data set,on which the treatment plan will be performed. Upon this con-dition, all the functionalities provided by the stand alone VScan be simultaneously executed in a collaborative environment.


Fig. 3. Architecture of the point-to-point communication scheme on VS.

The term “data set,” in the case of GALENOS collaborative en-vironment, includes patient demographic information, a set oftomographic images and information related to their acquisition(pixel spacing, slice position, etc.). The proposed scenario is de-composed into the following steps.

— A data set is scheduled to be communicated betweenthe collaborating parties.

— The transfer of the data set takes place.— A collaboration request is sent from one party to the

other.— In case of request’s approval, the online collaboration

session begins and the VS system of the collaboratingparty is initialized to the state of the VS system of thecalling party.

— online collaboration takes place, with the two partiessharing the same view of the VS, while voice commu-nication is also provided.

— The collaboration session ends, upon request of one ofthe collaborating parties.

In order to realize the above scenario, both asynchronous(offline transfer of data sets) and synchronous (online collab-oration) data exchange schemes are employed, while specialattention is given to security issues, which are raised by theintroduction of telecooperative functionalities. In a moreconcrete way, the information manipulated by the VS systemcontains many data related directly to identifiable persons,their illnesses and their treatment. Therefore, the introductionof electronic processing and transferring of such information,through the telecooperative work in VS, should be directlyfollowed by the implementation of a security frame, whicheliminates any possible violations of the authenticity, integrity,confidentiality and availability of data [32]–[35].

A. Offline Transfer of Data Sets

Collaboration on anatomical structure and beam definitionprerequisites a successful communication of the data set, onwhich the treatment plan will be based, between the collabo-rating parties. Data transfer is usually performed offline, con-sidering the size of the communicated data set, which can rangefrom 5–25 MB after compression (depending on the number of

Fig. 4. Architecture of the offline data exchange service.

tomographic images). To this end, the user can schedule dataexchange jobs, which are stored in GALENOS database, andinclude the location of the receiving party, an identifier for thedata set that is going to be sent, the scheduled time for thedelivery and the status of the job (“scheduled,” “in progress,”“sent” or “failed”). These data exchange jobs are handled bya custom-built data exchange service. The role of the serviceis to send/receive datasets to/from different workstations run-ning GALENOS VS without any intervention by the user. Atany given instance the service is in one of the following modes:idle, sending or receiving.

As shown in Fig. 4, the service is composed by four distinctlayers, each one responsible for a different task. The partition ofservice architecture into the specific layers is implied by the dis-crete functionalities of the service and is used for the indepen-dent implementation and redesign of each layer. The data layeris responsible for the communication with the GALENOS data-base, the compression layer is responsible for the compressionof the data set, the security layer handles the security function-alities and the socket layer establishes the required connectionand is responsible for the communication of data between theGALENOS sites.

Every layer is enabled to raise an error condition message.Thus, if an error is raised during sending mode, the delivery isinterrupted and the error condition is forwarded to the data layer.In the case where the error condition is raised during a receivingmode, apart from the interruption of the receiving process, anerror message is sent to the sender and forwarded to sender’sdata layer. In the data layer, a retry mechanism sets the job statusto “failed” after a number of unsuccessful attempts. On the other


hand, upon successful completion of data set communication,the receiver sends a message of transfer completion to the senderand through the sender’s data layer the job status is set to “sent.”A detailed description of each layer’s functionalities is given inthe following.

If the service is in idle mode, the data layer performs a peri-odic monitoring regarding the presence of a data exchange job.Whenever the scheduled time for a data exchange job arrives,the data layer sets the service into sending mode, assembles therequired data set from the GALENOS database, and changes thejob status to “in progress.” The data set is then forwarded to thecompression layer. In the case where the service is in receivingmode, the data layer receives a data set from the compressionlayer, stores the contained data in GALENOS database, and setsthe service in idle mode.

In the compression layer, the lossless Huffman algorithm [36]is used for the compression/decompression of the relative data.Then, the result is forwarded to the security or the data layer,respectively.

The fulfillment of the requirements ofconfidentiality, de-fending the privacy of the treated patient,integrity, preventingchanges or damages of the communicated data andauthenticityof data, ensuring that the communicated data is “genuine,” isachieved with the employment of encryption techniques, dig-ital signatures and certificates in the security layer. Furthermore,audit trailsare maintained through the chronological recordingof all transactions. The adopted security approach, is based oncertification authority architecture [37]. The certification au-thority creates certificates, which are used to exchange publickeys without contacting a public key server and thus avoiding apossible bottleneck. The main features of the security layer aredescribed in the following.

— Data is symmetrically encrypted with the use ofblowfish algorithm [38], [39]. The choice of theparticular algorithm was based on the level of theprovided security, on the speed of the algorithm andon its widespread deployment [40]. The symmetricencryption key is negotiated in every session.

— The negotiation of the symmetric key is performedusing the public key RSA [41] algorithm.

— Digital signatures of the communicated data are ap-pended to the encrypted plain message. The SHA1hashing algorithm [42] along with the RSA algorithmis used to support the digital signature scheme.

— Certificates are used to securely distribute key bindingsand data related to health care professionals identity.

— Certificates revocation lists (CRLs) are used for certifi-cates, which have become invalid. Due to the relativesmall size of CRLs, instead of querying a distributionpoint in order to obtain the CRL, “push” mechanismsare employed [43], which multicast the CRL to all sitesof the GALENOS collaborating environment.

In case of sending a data set, the results of the security layer,consisting of the encrypted data set, the digital signature and thecertificate of the sender, are forwarded to the socket layer. Oth-erwise, the received data from the socket layer is manipulated,in order to verify sender’s identity, to confirm the integrity of

the received data and finally, to decrypt the compressed data setand forward it to the compression layer.

The socket layer is responsible for the establishment andgraceful termination of a TCP/IP socket connection betweentwo sites and the reliable communication of data over theconnection.

B. Online Collaboration

The online collaboration in GALENOS VS follows the strict“What You See Is What I See” (WYSIWIS) [44] paradigm.WYSIWIS is a basic Computer Supported CollaborativeWork (CSCW) paradigm [45], which recognizes that efficientreference to collaborating entities, depends on a commonview of the application. As a result of the selection of thespecific CSCW paradigm, a master/slave control relationship isemployed, where the health care professional playing the roleof the master is able to perform all actions on the VS system,while the slave participant can only view the results of master’sactions. The roles between the collaborating parties can changeupon master’s request, while any of the parties can end thesession.

Following the design of the stand-alone VS application, theonline collaboration can take place either on anatomical or beamdescription design. The initial step toward the strict implemen-tation of the WYSIWIS paradigm is the transmission of the ap-plication status of the calling party to the remote one, duringthe establishment of the online collaboration session. The trans-mission is performed using the same architecture as the one em-ployed in the offline data exchange service, and the only differ-ence in layers’ functionalities is met in the data layer, which, inthis case, is solely responsible for the retrieval of the anatomicalor beam design related data.

From this point further, and after the take-up of master’s rolefrom one of the health care professionals, the collaborative envi-ronment forces the execution of all actions, which take place onthe master’s VS system to the remote one. Moreover, a parallelvoice communication channel further supports the collaborationbetween the two parties. Minimization of latencies introducedby the network is achieved by the employment of a UDP data-gram channel for the communication of events strictly related tomouse movements, ensuring the smoothness of application op-eration in the remote party and avoiding an overhead caused bythe large amount of events of this type. However, for the commu-nication of other types of events, a TCP/IP channel is employedavoiding the packet loss and message reordering usually intro-duced by the use of a UDP channel.

Even though the messages exchanged during the onlinecollaboration session do not contain information directlylinked with identifiable patients, an extensive analysis of thecommunicated messages, could lead to assumptions regardingthe treatment, which is going to be followed. Furthermore, theauthenticity of the data should also be assured. Thus, securitymechanisms must be employed, without having a significantimpact on the interactivity of the online collaboration. Takinginto consideration the required speed of electronic transactionsduring the telecooperative session, the negotiation of sym-metric keys, along with electronic signatures and certificates,


TABLE IDATA SET SIZE AND TIME PERIODS RECORDED

DURING OFFLINE COMMUNICATION

is done periodically during the tele-cooperation, achieving acompromise between security and performance.

IV. RESULTS

For the purpose of evaluating the proposed collaborative en-vironment, a pilot study has been performed, comprising theradiotherapy departments of four hospitals, three of which arelocated in the same city, while the fourth is at a distance of200 Km. The VS system of the GALENOS collaborative en-vironment has been installed on PCs with double Pentium IIIprocessors operating at 600 MHz and 256 MB RAM connectedto the LAN of the radiotherapy departments, and configured toautomatically receive data from the tomographic scanners andto export treatment settings in DICOM RT format. The commu-nication between the collaborating departments has been sup-ported by ISDN lines, and a single data channel (64 Kb/s) isallocated to a given connection.

A technical pilot study has been performed, aiming at theevaluation of the architecture design of the collaborative envi-ronment. The technical pilot study is based, in general, uponquality assessment of system functionalities and upon the mon-itoring of system performance.

The quality assessment of the results during online collabora-tion in comparison to the ones during the stand-alone operationis meaningless, in contrast with other telematics applications[14], [46], [47], which transmit images during online collabo-ration. The later are employing lossy compression algorithms,in order to achieve a compromise between speed and quality ofthe final result. On the other hand, all the functionalities of theGALENOS online collaboration session are performed locallyand only the instructions of the commands that have to be exe-cuted are transmitted over the network.

On the contrary, the evaluation of the collaboration environ-ment performance is of high importance and has been carriedout by monitoring the amount of time required for the offlinecommunication of data sets, and the latencies, which are de-tected during the online collaboration. The average time periodsrequired for various phases of the offline communication of 44data sets between distinct radiotherapy departments have beenrecorded and the results are shown in Table I, as a function ofdata sets size. The time period dedicated to the initialization of

Fig. 5. Latencies recorded during the online collaboration session, where theminimum mean latency was observed.

Fig. 6. Latencies recorded during the online collaboration session, where themaximum mean latency was observed.

the secure channel was 2.4 s, including the exchange of certifi-cates between the collaborating parties, the secure communica-tion of the symmetric key using the RSA algorithm (2048 bit)and the calculation/verification of digital signatures. The av-erage throughput of the line regarding the encrypted commu-nicated data was 60.583 kb/s.

As a first evaluation test of the online communication, mousemovements were the only events communicated between thecollaborating parties via UDP, and the exact times of their ex-ecution on the distinct sites were recorded. The mean latencydetected was 0.06 s, while the same test over a TCP/IP connec-tion resulted to a mean latency of 0.27 s. The next step includedthe monitoring of all the events communicated between the col-laborating parties in GALENOS VS, for eleven periods of dis-tinct collaborating sessions, the mean time of which was 15 min.During the monitoring periods, the average latency between theexecution of the commands at the master’s side, and the re-mote one, varied according to the actions taking place duringthe collaborating session. The minimum-recorded mean latencywas 0.14 s, while the maximum-recorded mean latency was0.20 s. The latencies recorded during the collaborating periods,in which the minimum and maximum mean latencies were en-countered, are shown in Figs. 5 and 6, respectively. The greatestlatencies were detected in the case of continuous rotations of3-D surface rendering scenes and were considered to be causedby the difference in the performance of the respective graphicsdisplay adapters.


V. DISCUSSION

The described evaluation study of the collaborative environ-ment has been performed by the developers, in cooperation withhealth care professionals of four radiotherapy departments, andwas aiming at the verification of the developed collaborative en-vironment. The next step toward a complete evaluation study,includes the validation procedure of possible applications of thecollaborative environment, investigating the level of satisfac-tion of user’s requirements and needs. Two major applicationscenarios have been identified and their validation study is cur-rently underway: the establishment of a collaborative network ofhealth care professionals using VS, and the formation of a geo-graphically distributed educational framework. A controlled ef-fectiveness study and a cost-effectiveness study [48] are takingplace for both applications.

In the case of a collaborative network of health care profes-sionals, a controlled effectiveness study is expected to prove thatthe outcomes of a collaboration session on VS are superior incomparison to the results of a stand-alone process. Dose calcu-lation routines are employed as a measure of the quality of thebeam placement outcomes, while a board of experts is enrolledinto the evaluation of the quality of the overall treatment plans ofcertain cases, for both stand-alone and collaborative proceduresof the design. Moreover, the time dedicated to the collaborativework from all parties, in comparison to the time required, for astand-alone process is monitored.

The workflow of the collaboration in the proposed environ-ment presents similarities with the educational process, wherethe tutor of an educational session interactively introducesmethods for designing a given treatment plan, and the traineeis able to perform a design upon tutor’s request. Thus, a frame-work supporting medical training programs and continuouseducation of health care professionals is established, giving theopportunity to the participants to share the experience acquiredin different radiotherapy departments. The evaluation of theprovided service is expected to reveal an improvement in thequality of the educational programs.

Moreover, a great amount of our efforts is currently focusedon the transfer of the collaborative environment to the web [18].The take-up of web technologies, already in use by other med-ical applications [15], [16], [49], [50], can offer a significantnumber of advantages, since health care professionals are en-abled to collaborate on VS, regardless of their location, the plat-form and the RTP system they are using. The implementation isconcentrated on the quality of the provided service and resultswill be presented in the future.

VI. CONCLUSION

A CSCW environment has been presented enabling healthcare professionals to perform online collaboration on the de-sign of radiotherapy treatment plans using a VS system. The im-plementation of the proposed architecture is based on the strictWYSIWIS paradigm and supported by a custom-built servicefor the offline communication of data. Special attention is givento security issues, based on the employment of encryption tech-niques, digital signatures and certificates, without having a sig-nificant impact on the performance of the environment. Stan-

dardized technologies are used, such as DICOM, TCP/IP, andISDN protocols, for the interconnection with the LAN of theradiotherapy department and the LAN-to-LAN connection be-tween the remote radiotherapy departments. However, innova-tive techniques in the layered structured service of the offlinecommunication of data-sets and in the secure management ofmessages during the online collaboration, result into an effec-tive approach, which can also be applied to similar cases of col-laborative environments manipulating medical data.

The evaluation of the collaborating environment indicated thesuitability of the proposed architecture, since high interactivityhas been achieved, without compromising the quality of the VSfunctionalities. The validation procedure is currently underwayand is expected to reveal the practical value of the environ-ment toward the promotion of the education and the collabora-tion between health care professionals in radiotherapy treatmentplanning.

ACKNOWLEDGMENT

The authors would like to thank Dr. Mouravliansky,M. Gletsos, E. Zacharaki, and C. Vasios for their valuablecontribution in the development of the GALENOS VS system,as well as all partners in the GALENOS consortium: UniversityHospital of Patras, The Radiology Department of the Universityof Athens at Evgenidion Hospital, Metaxa Memorial HospitalPiraeus, St. Savvas Anticancer Hospital Athens, and PhilipsHellas Medical Systems.

REFERENCES

[1] S. Webb, The Physics of Three-Dimensional RadiationTherapy. London, U.K.: Institute of Physics Publishing, 1993.

[2] J. Einhorn and J. E. Frödin, “Radiotherapy for cancer,”Acta Oncologica,vol. 1 and 2, Suppl. 6 and 7, 1996.

[3] The Royal College of Radiologists’ Clinical Oncology InformationNetwork,Guidelines for External Beam Radiotherapy. London, U.K.:Springer-Verlag, 1999.

[4] J. A. Purdy and E. Klein, “External photon beam dosimetry and treat-ment planning,” inPrinciples and Practice of Radiation Oncology, C.A. Perez and L. W. Brady, Eds. Philadelphia, PA: Lippincott-Raven,1997.

[5] J. A. Purdy, “Three-dimensional physics and treatment planning,” inPrinciples and Practice of Radiation Oncology, C. A. Perez and L. W.Brady, Eds. Philadelphia, PA: Lippincott-Raven , 1997.

[6] F. M. Khan, The Physics of Radiation Therapy. Baltimore, MD:Williams & Wilkins, 1994.

[7] G. W. Sherouse, J. D. Bourland, K. Reynolds, H. L. McMurry, T. P.Mitchell, and E. L. Chaney, “Virtual simulation in the clinical setting:Some practical considerations,”Int. J. Rad. Oncol. Biol. Phys., vol. 19,pp. 1059–1065, 1991.

[8] G. W. Sherouse and E. L. Chaney, “The portable virtual simulator,”Int.J. Rad. Oncol. Biol. Phys., vol. 21, pp. 475–482, 1993.

[9] W. Cai, S. Walter, G. Karangelis, and G. Sakas, “Collaborative virtualsimulation environment of radiotherapy treatment planning,” inCom-puter Graphics Forum, (Eurographics’00), 2000, vol. 19.

[10] D. Olsen, O. S. Bruland, and B. Davis, “Telemedicine in radiotherapytreatment planning: Requirements and applications,”Radiotherapy andOncology, vol. 54, no. 3, pp. 255–259, 2000.

[11] Digital Imaging and Communications in Medicine (DICOM) Version3.x, Final Text, NEMA Standards Publication PS3.x, 1993.

[12] D. Caramella, J. Repnen, F. Fabbrini, and C. Bartolozzi, “Teleradiologyin Europe,”Eur. J. Radiol., vol. 33, no. 1, pp. 2–7, 2000.

[13] G. Glombitza, H. Evers, S. Hassfeld, U. Engelmann, and H. Meinzer,“Virtual surgery in a (tele-)radiology framework,”IEEE Trans. Inform.Technol. Biomed., vol. 3, pp. 186–196, Sept. 1999.


[14] J. Reponen, S. Lahde, O. Tervonen, E. Ilkko, T. Rissanen, and I.Suramo, “Low-cost digital teleradiology,”Eur. J. Radiol., vol. 19, no.3, pp. 226–231, 1995.

[15] U. Engelmann, A. Schroter, U. Baur, O. Werner, M. Schwab, H. Muller,and H. Meinzer, “A three-generation model for teleradiology,”IEEETrans. Inform. Technol. Biomed., vol. 2, pp. 20–25, Mar. 1998.

[16] A. Abrardo and A. L. Casini, “Embedded java in a web-based teleradi-ology system,”IEEE Internet Comput., vol. 2, no. 3, pp. 60–68, 1998.

[17] D. Hunter, J. Brustrom, B. Goldsmith, L. Davis, M. Carlos, E. Ashley,G. Gardner, and I. Gaal, “Teleoncology in the department of defense: Atale of two systems,”Telemed. J., vol. 5, no. 3, pp. 273–283, 1999.

[18] E. Ntasis, T. S. Maniatis, and K. S. Nikita, “Web-based radiotherapytreatment planning,” inProc. World Congr. Medical Physics andBiomedical Engineering, Chicago, IL, July 2000.

[19] International Commission on Radiation Units and Measurements,“Prescribing, Recording, and Reporting Photon Beam Therapy,” ICRU,Bethesda, MD, ICRU Rep. 50, 1993.

[20] R. M. Haralick and L. G. Shapiro, “Survey: Image segmentation tech-niques,”Comp. Vis. Graph. Image Proc., vol. 29, pp. 100–132, 1985.

[21] R. Zucker, “Region growing: Childhood and adolescence,”Comp.Graph Image Process., vol. 5, pp. 382–399, 1976.

[22] R. C. Gonzalez and R. E. Woods,Digital Image Processing, 2ed. Reading, MA: Addison Wesley, 1993.

[23] H. Fuchs, Z. M. Kedem, and S. P. Uselton, “Optimal surface reconstruc-tion from planar contours,”Commun. ACM, vol. 20, no. 10, pp. 693–702,1977.

[24] International Electrotechnical Commission (IEC) Subcommittee 62C,“Radiotherapy Equipment—Coordinates, Movements and Scales,”, IECDraft 1217, Mar. 26, 1993.

[25] M. Levoy, “Display of surfaces from volume data,”IEEE Comput.Graph. Appl., vol. 8, pp. 135–143, Oct./Nov. 1988.

[26] T. Malzbender, “Fourier volume rendering,”ACM Trans. Graph., vol.12, no. 3, pp. 233–250, 1993.

[27] T. Totsuka and M. Levoy, “Frequency domain volume rendering,” inProc. 20th Annu. ACM Conf. Comp. Graphics, 1993, pp. 271–278.

[28] E. Ntasis, W. Cai, G. Sakas, and K. S. Nikita, “Real time Digital Recon-structed Radiograph (DRR) rendering in frequency domain,” inProc.BMES - EMBS, Atlanta, GA, Nov. 1999.

[29] P. Lacroute and M. Levoy, “Fast volume rendering using a shear-warpfactorization of the viewing transformation,” inProc. 21st Annu. ACMConf. Computer Graphics, 1994, pp. 451–458.

[30] J. N. Stahl, J. Zhang, C. Zellner, E. V. Pomerantsev, T. M. Chou, andH. K. Huang, “Teleconferencing with dynamic medical images,”IEEETrans. Inform. Technol. Biomed., vol. 4, pp. 88–96, June 2000.

[31] J. Zhang, J. N. Stahl, H. K. Huang, X. Zhou, S. L. Lou, and K. S. Song,“Real-time teleconsultation with high-resolution and large-volume med-ical images for collaborative healthcare,”IEEE Trans. Inform. Technol.Biomed., vol. 4, pp. 178–185, June 2000.

[32] I. Bashir, E. Serafini, and K. Wall, “Securing network software applica-tions,” Commun. ACM, vol. 44, no. 2, pp. 29–30, 2001.

[33] American College of Radiology, Reston, VA,ACR Standard for Telera-diology, Rev. 1998.

[34] E. Ntasis, K. S. Nikita, and G. Matsopoulos, “Security servicesfor telematics applications in clinical radio-oncology,” inRecentAdvances in Signal Processing and Communications, N. Mastorakis,Ed. Singapore: World Scientific, 1999, pp. 369–373.

[35] E. Ntasis, T. Maniatis, and K. S. Nikita, “Secure tele-cooperative workin virtual simulation procedure of radiotherapy treatment planning,” inIFMBE Proc. MEDICON, 2001, pp. 115–118.

[36] D. A. Huffmann, “A method for the construction of minimum-redun-dancy codes,”Proc. IRE, vol. 40, pp. 1098–1101, 1952.

[37] L. Kohnfelder, “Toward a practical public-key cryptosystem,” B.S.Thesis, MIT, 1978.

[38] B. Schneier, “Description of a new variable-length key, 64-bit blockcipher (blowfish),” inFast Software Encryption, Cambridge SecurityWorkshop Proceedings. New York: Springer-Verlag, 1994, pp.191–204.

[39] , Applied Cryptography: Protocols, Algorithms, and Source Codein C. New York: Wiley, 1996.

[40] M. C. Caloyannides, “Encryption wars: Early battles,”IEEE Spectrum,vol. 37, no. 4, pp. 37–43, 2000.

[41] R. L. Rivest, A. Shamir, and L. M. Adleman, “A method for obtainingdigital signatures and public-key cryptosystems,”Commun. ACM, vol.21, no. 2, pp. 20–126, 1978.

[42] National Institute of Standards and Technology, Reston, VA,NIST FIPSPUB 180, Secure Hash StandardWashington, DC, 1993, U.S. Dept.Commerce.

[43] P. D. McDaniel and S. Jamin, “Windowed certificate revocation,” inProc. IEEE INFOCOM, 2000, pp. 1406–1414.

[44] M. Stefik, D. G. Bobrow, G. Foster, S. Lanning, and D. Tatar,“WYSIWIS revised: Early experiences with multiuser interfaces,”ACM Trans. Office Informat. Syst., vol. 5, no. 2, pp. 147–167, 1987.

[45] W. Reinhard, J. Schweitzer, G. Völksen, and M. Weber, “CSCW tools:Concepts and architectures,”IEEE Comput., 1994.

[46] M. H. Weinstein and J. I. Epstein, “Telepathology diagnosis of prostateneedle biopsies,”Hum. Pathol., vol. 28, pp. 22–29, 1997.

[47] L. Makris, I. Kamilatos, E. V. Kopsacheilis, and M. G. Strintzis,“Teleworks: A CSCW application for remote medical diagnosis supportand teleconsultation,”IEEE Trans. Inform. Technol. Biomed., vol. 2,pp. 62–74, June 1998.

[48] R. Holle and G. Zahlmann, “Evaluation of telemedical services,”IEEETrans. Inform. Technol. Biomed., vol. 3, pp. 84–91, June 1999.

[49] J. Bai, Y. Zhang, and B. Dai, “Design and development of an interac-tive medical teleconsultation system over the World Wide Web,”IEEETrans. Inform. Technol. Biomed., vol. 2, pp. 74–79, June 1998.

[50] Y. Kim, J. H. Choi, J. Lee, M. K. Kim, N. K. Kim, J. S. Yeom, andY. O. Kim, “Collaborative surgical simulation over the Internet,”IEEEInternet Comput., vol. 5, pp. 65–73, Sept. 2001.

Efthymios Ntasis was born in Athens, Greecein 1975. He obtained the Diploma in Electricaland Computer Engineering in 1998 and the Ph.D.degree in biomedical engineering from the NationalTechnical University of Athens, Greece in 2001.Currently, he is working towards the Master of PublicHealth degree in Health Policy and Management atColumbia University, New York, NY.

His research interests include medical imaging,medical informatics and health care policy. He is amember of the ACM.

Theofanis A. Maniatis (M’00) was born in Athens, Greece, in 1967. He ob-tained the Diploma in Electrical Engineering from the National Technical Uni-versity of Athens, Greece in 1990, the M.A.Sc. degree from the University ofToronto, Canada in 1993 and the Ph.D. degree from the National Technical Uni-versity of Athens, Greece in 1998.

Since 1999, he has been working as a Researcher at the Institute of Commu-nication and Computer Systems, National Technical University of Athens. Hisresearch interests include inverse scattering, computational electromagnetics,and medical imaging.

Dr. Maniatis is a member of the Technical Chamber of Greece and the Hel-lenic Society of Biomedical Engineering.

Konstantina S. Nikita (M’96-SM’00) received theDiploma in electrical engineering and the Ph.D.degree from the National Technical University ofAthens, Greece, in 1986 and 1990, respectively. Shethen received the M.D. degree from the Universityof Athens, Greece, in 1993.

Since 1990, she has been working as a Researcherat the Institute of Communication and ComputerSystems, National Technical University of Athens.In 1996, she joined the Department of Electrical andComputer Engineering, National Technical Univer-

sity of Athens, where she is currently an Associate Professor. She has authoredor coauthored 65 papers in refereed international journals and chapters inbooks, and more than 100 papers in international conference proceedings. Shehas also been the Technical Manager of a number of European and NationalResearch and Development projects in the field of biomedical engineering.Her current research interests include medical imaging and image processing,medical informatics, telematics applications in medicine, applications ofelectromagnetic waves in medicine, simulation of biological and physiologicalsystems.

Dr. Nikita is a member of the Technical Chamber of Greece, the Athens Med-ical Association, and the Hellenic Society of Biomedical Engineering.

Documents

Real-time collaborative environment for radiation treatment planning virtual simulation