3D surgical planning in patients affected by lipodystrophy

https://doi.org/10.1016/j.compmedimag.2014.12.007Get rights and content

Highlights

  • Use of 3D surface scanning for surgical planning (no radiation required).

  • Easy-to-use application for surgical planning of facial lipodystrophy.

  • Providing of a facial natural surface to guide surgeons in their interventions.

  • Computation of the volume to be injected during facial surgery interventions.

Abstract

Lipodystrophy is a pathological condition characterized by the focal or general absence of adipose tissue. Surgeons reset the patient's surface contours using injectable materials to recreate a normal physical appearance. However, due to difficulties in preoperative planning and intraoperative assessment, about 15% of the surgical procedures involved are reinterventions to improve volume or symmetry. This increases the need for an available, efficient tool capable of providing the surgeon with a good estimation of the volumes to be injected before the intervention proper begins. This work describes a virtual reality-based application for the surgical planning of facial lipodystrophy correction (FLIC). The tool uses points selected interactively by the surgeon to compute a curve that delimits the surface area to be operated on. It then automatically computes an estimated natural reconstructed surface and the quantity of volume that needs to be implanted during the intervention. Experiments have been carried out in which the filling volumes estimated using FLIC and ZBrush software were compared with the real volumes injected by the surgeon. ICCs higher than 0.97 indicate that there were no significant differences between the respective measurements, thus validating the tool proposed in this paper.

Introduction

HIV or AIDS is a disease that has spread in recent years. Thanks to advances in medication control, it is now considered a chronic illness rather than a fatal disease. The introduction of antiretroviral therapies (ART) has significantly reduced mortality and infection rates in patients affected by the human immunodeficiency virus (HIV). Anti-retroviral drugs have reduced the viral load to undetectable levels, allowing patients to lead normal lives. However, one of the side effects of AIDS treatment is the marked lipodystrophy [1], [2] it can cause in several parts of the body.

Lipodystrophy – the loss and/or accumulation of fat in specific areas of the body – is a major concern among HIV patients because of the discomfort and stigmatization produced by the physical changes it induces [1], [2].

One of the biggest concerns for patients affected by facial lipoatrophy (FLA) is the effect and visibility the pathology has in their faces. Facial contours are very complex and vary greatly from one individual to another. It is therefore particularly difficult to monitor volume changes over long periods in the FLA population, economic and non-invasive techniques being required to measure facial changes. The imaging technique of three dimensional (3D) surface scanning offers opportunities for the objective analysis of human facial features and for monitoring FLA evolution [3], with the added advantage that no radiation is required to generate a 3D model of the patient.

Therapeutic measures to limit or correct side effects associated with FLA include:

  • 1

    Replacement of the pharmaceuticals used in FLA treatment by others less related to these effects (if the patient's condition allows such a substitution) and,

  • 2

    Surgical techniques using polyacrylamide synthetic implants and autologous fat grafting (facial lipofilling) to correct facial lipoatrophy (FLA).

People with moderate or severe lipodystrophy in Andalusia (Spain) have been receiving reconstructive facial surgery since 2003 through the Andalusian Health System. However, this type of surgery is relatively new and reliable evidence is still lacking with regard to such important aspects as safety, tolerability and effectiveness in the medium and long term. One reason for this is the difficulty of quantizing the degree of improvement and persistence over time with existing diagnostic methods. There are two main problems in this type of surgery. On the one hand, surgeons do not have models of the patients’ faces as they were when healthy (free of lipodystrophy), and this makes it difficult to know how much volume is required in the facial area to give the face a natural appearance. On the other hand, to avoid having to perform follow-up operations later, surgeons would like to have an accurate idea of the liquid volume that needs to be injected prior to the intervention. For example, due to difficulties involved in preoperative planning and intraoperative assessment, at least in our series of approximately 275 consecutive procedures for correcting facial lipoatrophy with polyalkylamide synthetic facial implants, about 15% of the operations were reinterventions to improve either volume or symmetry.

This work presents a novel virtual reality-based software application for assisting surgeons during the planning of facial surgery. The tool gives the estimated facial volume that needs to be injected in the patient to correct the effects of FLA, thereby reducing the number of reinterventions stemming from problems with the initial treatment and/or patient dissatisfaction with the results. The current state of the art in surgical planning tools does not provide a tool that effectively helps surgeons during the planning of surgical interventions in patients with facial lipodystrophy. Admittedly, much work has been done in the field of surgical planning [4], [5] and many studies addressing facial surgical planning in particular can be found in fields like maxillo-facial surgical planning [6] and Orthognathic Surgery [7]. In these works, 3D bone and soft-tissue models are usually obtained by applying marching cubes to the CT data available. Each coordinate on the bone surface has its own unique projection on the soft-tissue mesh. When a point on the bone surface is changed, the corresponding soft tissue coordinate is changed accordingly. This makes it possible to simulate soft tissue movement caused by bone movement. However, all these works rely on the use of CT imaging, which implies radiating the patient each time a new intervention is necessary. Moreover, the techniques described in these works are intended mainly for bone interventions and not just for soft tissue surgery. In our work, we use a cost-effective, economic, non-invasive 3D surface scanning and imaging technique, which makes it possible to monitor FLA evolution [3] over long periods of time, generating a 3D model of the patients without having to resort to any radiation. The surgical interventions are simulated directly on the patient's surface and not on bone structures.

Some other authors have focused on developing simulator tools using finite-element modeling to create real-time virtual environments for skin surgery. In [8], a deformable soft-tissue model for skin that responds in real time to user manipulation was implemented and used to perform tensile stress tests on human skin samples. In Sifakis et al. [9], a real-time virtual surgical environment was implemented to enable the surgeon to practice cutting and manipulating tissue. The main drawback of these simulators is that they do not operate on real patient models and focus almost exclusively on obtaining accurate models of skin tissue rather than providing a fast method to help surgeons plan interventions.

Many other works propose software applications which freely modify specific parts of the body (face, breasts, ears, neck, etc.) [10], [11], [12], [13]. These works do not consider real skin models but are intended to provide an approximate idea of what the patient would look like after surgery. Most of the programs are essentially photograph editors, intended only as a means of communication between patient and surgeon. Obviously, the results they produce have to be approved by the specialist who is going to perform the surgery.

In the work presented here, the body region subject to intervention is not freely modified by the user. The user (the surgeon) merely selects a number of points around the area to be operated on, and a new surface is then computed in order to give the patient's face a natural appearance, ideally similar to what it looked like before the illness.

The tool was validated using 3D images obtained by scanning the facial surface of several patients affected by HIV- and ART-associated facial lipodystrophy. It has a very manageable interactive graphical interface. The tool is described in greater detail in the next section. The third section describes some of the material employed and presents some experimental results. Finally, the last section contains a discussion of the results obtained and some conclusions.

Section snippets

Methods

The proposed FLIC software tool uses a surface scanner to create a model of the patient's facial surface. This imaging technique was chosen for its cost-effectiveness, and also because, unlike other alternatives, such as Computer Tomography, it involves no use of radiation.

To develop the tool we used VTK class libraries [14], which operate in the programming language C++. VTK (Visualization Toolkit) is an open source object-oriented programming code for displaying and processing images.

The set

Material and experimental results

The process of validating and evaluating our proposed tool consisted of three stages. 25 subjects were involved in all the experiments carried out. All of the patients in our study were male and Caucasian. Their average age was 48.3 years (range: 30–68 years). Informed consent was obtained from the patients and the in-house review board approved the study protocol. A scan of each patient's cervicofacial region was obtained using a 3D CSI white light scanner. All the scans were performed using

Conclusions and discussion

Newly available drugs have considerably reduced the impact of their side effects on HIV patients to such an extent that HIV is now seen more as a chronic disease than as an inevitably fatal illness. However, one of the most obvious side effects of medication is facial lipodystrophy. So far, surgery is the most effective solution, but the difficulties of preoperative planning and intraoperative assessment give rise to a reintervention rate of about 15%, the aim of most reinterventions being to

Conflict of interest statement

None declared.

Acknowledgements

The authors wish to thank the Plastic Surgery Unit at the Virgen del Rocio Hospital for providing the three-dimensional models of the patients used in this work. This work was supported in part by the FIPSE project “Realidad Virtual aplicada a la lipodistrofia facial a T.A.R. en pacientes con VIH” (“Virtual Reality Applied to Facial Lipodystrophy and A.R.T. in HIV Patients”) and by project P11-TIC-7727, Junta de Andalucia (Spain).

José-Antonio Pérez-Carrasco is assistant professor in the Theory and Signal Processing Department within the University of Seville. He received the Ph.D. degree in electronics and signal processing from the University of Sevilla, Sevilla, Spain, in 2011. His Ph.D. dissertation was titled “Simulation Tool for Assembling and Analyzing Hierarchical AER Systems for Visual Processing”. His current research interests include image processing and its medical applications, visual perception, and

References (24)

  • Plastic Surgery Simulator

    (2015)
  • Online Plastic Surgery Simulator. Online Software

    (2015)
  • Cited by (0)

    José-Antonio Pérez-Carrasco is assistant professor in the Theory and Signal Processing Department within the University of Seville. He received the Ph.D. degree in electronics and signal processing from the University of Sevilla, Sevilla, Spain, in 2011. His Ph.D. dissertation was titled “Simulation Tool for Assembling and Analyzing Hierarchical AER Systems for Visual Processing”. His current research interests include image processing and its medical applications, visual perception, and real-time processing.

    Begoña Acha received the Ph.D. degree in communication engineering from the University of Sevilla, Sevilla, Spain, in 2002. She has been teaching and conducting research with the University of Sevilla since 1996. She is currently a Tenured Professor with the Signal Processing and Communications Department, University of Sevilla. Her current research activities include works in the field of color image processing and its medical applications, skin images, retinal images, and computed tomography images. She has published numerous papers in journals and conferences and she co-authored Color Image Processing with Biomedical Applications (SPIE Press, 2011).

    Tomás Gómez Cía was born in Pamplona, Spain, in 1958. He studied medicine in the University of Navarra. In 1982 he got a position as Resident Medical Intern in the Virgen del Rocío Hospital (HUVR), finishing his training as Specialist in Plastic, Reconstructive and Aesthetic Surgery in 1986. During 1987 he was a fellow of the Fund for Health Research at this center. In 1988 he started his professional career as specialist in the mentioned hospital, assigned to the Burn Unit. In 1989 he received the Ph.D. degree in Medicine and Surgery from the University of Sevilla. His Ph.D. dissertation was titled “Analysis by digital simulation of alterations in capillary permeability of burn patients” and it was held at the School of Industrial Engineering. In 1999 he became Clinical Head of the Burn Unit of the HUVR. In 2000 he became Head of the Plastic Surgery and Burn Service in the same Institution. Since 2003 he is also Director of the Clinical Management Unit of Plastic Surgery and Burns in the HUVR. He has published 43 articles in international journals, 36 publications in national journals, over 30 book chapters, 87 papers in international conferences, 162 communications at national conferences, and he has participated in and directed several research projects funded by various agencies. Since 2005 he leads the VirSSPA (Virtual Reality Applied to Surgical Planning) project. This project is funded by the Ministry of Health of the Government of Andalusia. VirSSPA is an open platform for surgical simulations where it is possible to perform a detailed analysis of the patients, and to plan and simulate the surgical procedure.

    Rafael Antonio López-García works as Resident Medical Intern in the Virgen del Rocío Hospital (HUVR). He is working towards obtaining his Ph.D. degree in Medicine in the University of Seville.

    Carlos Delgado received the degree in Telecommunications Engineering from the University of Seville, Spain, in 2011. In 2011 he got a contract to develop software in biomedical applications.

    Carmen Serrano received the Ph.D. degree in communication engineering from the University of Sevilla, Sevilla, Spain, in 2002. She has been teaching and conducting research at the University of Sevilla since 1996. She is currently a Tenured Professor with the Signal Processing and Communications Department, University of Sevilla. Her research interest is mainly focused on image processing. In particular, she conducts research on the detection of pathological signs in medical images and on the segmentation of anatomical structure for surgical planning. Her main research field is color image processing. She has developed algorithms for computer-assisted diagnosis of burn images and pigmented lesions of the skin and retinal images. Her research has been applied to medical images, including skin images as well as radiological images from modalities such as computed tomography and magnetic resonance imaging. She co-authored Color Image Processing with Biomedical Applications (SPIE Press, 2011) and she has published numerous papers in journals and prestigious conferences.

    View full text