Part 6: advanced cardiovascular life support. Section 7: algorithm approach to ACLS. 7C: a guide to the international ACLS algorithms. European Resuscitation Council.
Explore the source record for details and available documents.
Algorithms↗
SEARCH · Search PubMed
Results for “Algorithms”
Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Part 6: advanced cardiovascular life support. Section 7: algorithm approach to ACLS emergencies. 7D: the tachycardia algorithms. European Resuscitation Council.
Explore the source record for details and available documents.
Algorithms↗
Proposal for clinical algorithm standards. Society for Medical Decision Making Committee on Standardization of Clinical Algorithms.
Explore the source record for details and available documents.
Algorithms↗
Developing treatment algorithms for unipolar depression in Cyberspace: International Psychopharmacology Algorithm Project (IPAP).
Explore the source record for details and available documents.
Algorithms↗
Algorithm-directed care by nonphysician practitioners in a pediatric population: Part I. Adherence to algorithm logic and reproducibility of nonphysician practitioner data-gathering behavior.
The ability of quickly trained nonphysician practitioners to care for pediatric patients with upper respiratory tract infections (URIs) was evaluated in 3802 patients. These nonphysician practitioners (Army Pamosists) used an explicit treatment protocol and computerized audit of protocol adherence. Pamosists omitted protocol-suggested plans in 3.7 per cent of cases and ordered unnecessary treatment plans in 1.7 per cent of cases. They did not obtain a suggested MD consultation in 6.2 per cent of the cases. Agreement between Pamosists and pediatricians on data and management decisions (PM-MD study) was then compared with agreement on the same variables between pairs of a group of five pediatricians who saw a separate but comparable group of 103 pediatric URI patients (MD-MD study). The Pamosists demonstrated good (77 to 89 per cent) overall agreement with pediatricians, and the amounts of agreement between Pamosists and pediatricians in the PM-MD study did not generally differ significantly (p greater than 0.05) from the agreement on the same variables between pairs of pediatricians in the MD-MD study.
Allied Health Personnel↗
Monte Carlo algorithms with absorbing Markov chains: Fast local algorithms for slow dynamics.
Explore the source record for details and available documents.
Journal Article↗
The effect of reconstruction algorithm on conspicuity of polyps in CT colonography.
OBJECTIVE: CT colonography studies to date have used a standard CT algorithm. To determine whether nonstandard algorithms may result in better performance of CT colonography, we conducted a prospective, blinded-observer study of the effect of the reconstruction algorithm on the conspicuity of colonic polyps and folds. SUBJECTS AND METHODS: CT colonography of patients with proven polyps, masses, or polypoid folds was performed on an MDCT scanner, and the images were reconstructed using the standard, soft, lung, and detail algorithms. Two experiments were performed. The first used four patient data sets of a short segment of colon (30-60 images), each reconstructed using all four algorithms and then viewed on a workstation in a four-on-one format that allowed all four reconstructions to be viewed simultaneously. The second used six sets of cut-film images (four or eight images each); images within each set differed only in the reconstruction algorithm used to generate them (eight-image sets were prepared with two different level settings). Twenty-one reviewers with varying levels of experience who were unaware of the purpose of the study were asked to rank the images within each set according to their value in the detection of either polyps or masses. RESULTS: Reviewers showed statistically significant differences in preference for the four algorithms (p = 0.037 in the computer-based experiment; for the cut-film experiment, p = 0.029 for the four-image sets and p = 0.041 for the eight-image sets). In the computer-based experiment, reviewers preferred the detail algorithm to the standard algorithm with an estimated probability of 0.67 (95% confidence interval [CI], 0.57-0.75) and the soft algorithm over the standard algorithm with an estimated probability of 0.59 (95% CI, 0.51-0.66). However, reviewers with the most experience (having interpreted at least 250 cases) preferred the soft algorithm over the standard algorithm by the same two-to-one margin as observed for the detail algorithm. In contrast, the standard and detail algorithms were ranked similarly in the cut-film experiment, with the soft and lung algorithms ranked worst. CONCLUSION: To our knowledge, ours is the first observer study on the effect of the reconstruction algorithm on conspicuity of folds and polyps in CT colonography. Our results indicate significant differences in the reconstruction algorithms, with the soft and detail algorithms being preferred over the standard algorithm by experienced reviewers when interpreting images on a workstation. These results indicate the need for further research into the effect of reconstruction algorithms on CT colonography.
Algorithms↗
Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data.
OBJECTIVES: Implementation of the International Statistical Classification of Disease and Related Health Problems, 10th Revision (ICD-10) coding system presents challenges for using administrative data. Recognizing this, we conducted a multistep process to develop ICD-10 coding algorithms to define Charlson and Elixhauser comorbidities in administrative data and assess the performance of the resulting algorithms. METHODS: ICD-10 coding algorithms were developed by "translation" of the ICD-9-CM codes constituting Deyo's (for Charlson comorbidities) and Elixhauser's coding algorithms and by physicians' assessment of the face-validity of selected ICD-10 codes. The process of carefully developing ICD-10 algorithms also produced modified and enhanced ICD-9-CM coding algorithms for the Charlson and Elixhauser comorbidities. We then used data on in-patients aged 18 years and older in ICD-9-CM and ICD-10 administrative hospital discharge data from a Canadian health region to assess the comorbidity frequencies and mortality prediction achieved by the original ICD-9-CM algorithms, the enhanced ICD-9-CM algorithms, and the new ICD-10 coding algorithms. RESULTS: Among 56,585 patients in the ICD-9-CM data and 58,805 patients in the ICD-10 data, frequencies of the 17 Charlson comorbidities and the 30 Elixhauser comorbidities remained generally similar across algorithms. The new ICD-10 and enhanced ICD-9-CM coding algorithms either matched or outperformed the original Deyo and Elixhauser ICD-9-CM coding algorithms in predicting in-hospital mortality. The C-statistic was 0.842 for Deyo's ICD-9-CM coding algorithm, 0.860 for the ICD-10 coding algorithm, and 0.859 for the enhanced ICD-9-CM coding algorithm, 0.868 for the original Elixhauser ICD-9-CM coding algorithm, 0.870 for the ICD-10 coding algorithm and 0.878 for the enhanced ICD-9-CM coding algorithm. CONCLUSIONS: These newly developed ICD-10 and ICD-9-CM comorbidity coding algorithms produce similar estimates of comorbidity prevalence in administrative data, and may outperform existing ICD-9-CM coding algorithms.
Algorithms↗
The effects of the change in the NRMP matching algorithm. National Resident Matching Program.
CONTEXT: Following 2 years of heated controversy about the resident match, the National Resident Matching Program (NRMP) recently voted to replace the existing matching algorithm with a newly designed applicant-proposing algorithm. OBJECTIVE: To design an applicant-proposing algorithm for the match and compare it with the existing NRMP algorithm to determine how many applicants and residency programs could be expected to receive better or worse matches from the 2 algorithms, how the different algorithms influence the opportunity for strategic behavior, and what advice can be given to participants. DESIGN: Computational experiments compared the newly designed applicant-proposing algorithm with the existing NRMP algorithm on the rank order lists (ROLs) submitted by all applicants and residency programs in the 1987 and 1993 through 1996 NRMP matches. RESULTS: Differences in the matchings produced by the 2 algorithms are small: fewer than 1 in 1000 applicants would have received a different match. Most (but not all) of the few applicants who are matched to different positions by the 2 algorithms do better when the applicant-proposing algorithm is used; the opposite is true for programs. Opportunities for profitable strategic behavior are very rare for both applicants and programs under either algorithm. With either algorithm, both applicants and programs can be advised that trying to get a preferred match by behaving strategically is far more likely to harm than to help them. CONCLUSIONS: The existing NRMP algorithm and the newly designed applicant-proposing algorithm perform similarly. Both algorithms make it sensible for applicants and residency programs to arrange their ROLs based solely on their preferences for possible matches. The choice of algorithms will systematically affect the matches of only a small group of applicants (<0.1%). The NRMP's recent decision to use the applicant-proposing algorithm starting in 1998 reflects a judgment about the impact of this difference on applicants and programs.
Algorithms↗
An adaptive interval-based algorithm for withholding ICD therapy during sinus tachycardia.
Avoiding inappropriate ICD therapy during supraventricular tachycardia (SVT) while assuring 100% sensitivity for VT/VF remains a challenge. Inappropriate VT/VF therapy during sinus tachycardia (ST) is particularly distressing to the patient because the full sequence of ICD therapies is often delivered. ST or 1:1 atrial tachycardia (AT) with long PR intervals and ST or AT with atrial oversensing of far-field R waves cause the majority of inappropriate therapy in the Medtronic GEM DR (Model 7271) ICD. The goals of the present effort were to define an adaptive interval-based algorithm for withholding VT/VF therapy in dual chamber ICDs during ST and to compare performance of the adaptive algorithm with that of the original ST withholding algorithm in the GEM DR. The adaptive algorithm uses a combination of 1:1 atrial to ventricular conduction pattern, changes in RR intervals and changes in intrinsic PR intervals to establish evidence for or against the presence of ST. Performances of the adaptive and original ST withholding algorithms were compared on 3 databases collected by implanted GEM DR devices. The first database included 684 spontaneous VT/VF episodes. The second database included 216 spontaneous SVT episodes that received inappropriate VT/VF therapy. These databases included up to 2,000 atrial or ventricular sensed or paced events preceding the spontaneous tachycardias. The third database included 320 spontaneous ST/AT episodes for which therapy was appropriately withheld by the GEM DR. Performance of the adaptive algorithm on the third database was predicted rather than directly computed because of record length limitations. VT/VF therapy was classified as "withheld" if evidence of ST remained high for one algorithm (i.e., at least 7 more beats to VT/VF detection) at the point of VT/VF detection by the other algorithm. For the 684 true VT/VF episodes, the original algorithm withheld VT/VF therapy in 5 episodes and the adaptive algorithm withheld VT/VF therapy in 3 episodes. The 95% confidence interval for the difference in VT/VF sensitivity between the adaptive and original algorithms was [-0.5 to + 1.1%]. Twelve of the 320 ST/AT episodes (3.8%) that were appropriately classified by the original algorithm were predicted to receive inappropriate therapy by the adaptive algorithm. However, relative to the original algorithm, the adaptive algorithm appropriately withheld VT/VF therapy for 76 of 216 true SVT episodes (i.e., incremental specificity of 35.2%). For the specific SVT episodes that were the targets for improvement by the adaptive ST algorithm (ST/AT with long PR intervals and ST/AT with intermittent atrial oversensing of far-field R waves), the adaptive algorithm reduced inappropriate therapy by 63.2%.
Algorithms↗
Speed and convergence properties of gradient algorithms for optimization of IMRT.
Gradient algorithms are the most commonly employed search methods in the routine optimization of IMRT plans. It is well known that local minima can exist for dose-volume-based and biology-based objective functions. The purpose of this paper is to compare the relative speed of different gradient algorithms, to investigate the strategies for accelerating the optimization process, to assess the validity of these strategies, and to study the convergence properties of these algorithms for dose-volume and biological objective functions. With these aims in mind, we implemented Newton's, conjugate gradient (CG), and the steepest decent (SD) algorithms for dose-volume- and EUD-based objective functions. Our implementation of Newton's algorithm approximates the second derivative matrix (Hessian) by its diagonal. The standard SD algorithm and the CG algorithm with "line minimization" were also implemented. In addition, we investigated the use of a variation of the CG algorithm, called the "scaled conjugate gradient" (SCG) algorithm. To accelerate the optimization process, we investigated the validity of the use of a "hybrid optimization" strategy, in which approximations to calculated dose distributions are used during most of the iterations. Published studies have indicated that getting trapped in local minima is not a significant problem. To investigate this issue further, we first obtained, by trial and error, and starting with uniform intensity distributions, the parameters of the dose-volume- or EUD-based objective functions which produced IMRT plans that satisfied the clinical requirements. Using the resulting optimized intensity distributions as the initial guess, we investigated the possibility of getting trapped in a local minimum. For most of the results presented, we used a lung cancer case. To illustrate the generality of our methods, the results for a prostate case are also presented. For both dose-volume and EUD based objective functions, Newton's method far outperforms other algorithms in terms of speed. The SCG algorithm, which avoids expensive "line minimization," can speed up the standard CG algorithm by at least a factor of 2. For the same initial conditions, all algorithms converge essentially to the same plan. However, we demonstrate that for any of the algorithms studied, starting with previously optimized intensity distributions as the initial guess but for different objective function parameters, the solution frequently gets trapped in local minima. We found that the initial intensity distribution obtained from IMRT optimization utilizing objective function parameters, which favor a specific anatomic structure, would lead to a local minimum corresponding to that structure. Our results indicate that from among the gradient algorithms tested, Newton's method appears to be the fastest by far. Different gradient algorithms have the same convergence properties for dose-volume- and EUD-based objective functions. The hybrid dose calculation strategy is valid and can significantly accelerate the optimization process. The degree of acceleration achieved depends on the type of optimization problem being addressed (e.g., IMRT optimization, intensity modulated beam configuration optimization, or objective function parameter optimization). Under special conditions, gradient algorithms will get trapped in local minima, and reoptimization, starting with the results of previous optimization, will lead to solutions that are generally not significantly different from the local minimum.
Algorithms↗
Are decisions using cost-utility analyses robust to choice of SF-36/SF-12 preference-based algorithm?
BACKGROUND: Cost utility analysis (CUA) using SF-36/SF-12 data has been facilitated by the development of several preference-based algorithms. The purpose of this study was to illustrate how decision-making could be affected by the choice of preference-based algorithms for the SF-36 and SF-12, and provide some guidance on selecting an appropriate algorithm. METHODS: Two sets of data were used: (1) a clinical trial of adult asthma patients; and (2) a longitudinal study of post-stroke patients. Incremental costs were assumed to be 2000 dollars per year over standard treatment, and QALY gains realized over a 1-year period. Ten published algorithms were identified, denoted by first author: Brazier (SF-36), Brazier (SF-12), Shmueli, Fryback, Lundberg, Nichol, Franks (3 algorithms), and Lawrence. Incremental cost-utility ratios (ICURs) for each algorithm, stated in dollars per quality-adjusted life year (dollars/QALY), were ranked and compared between datasets. RESULTS: In the asthma patients, estimated ICURs ranged from Lawrence's SF-12 algorithm at 30,769 dollars/QALY (95% CI: 26,316 to 36,697) to Brazier's SF-36 algorithm at 63,492 dollars/QALY (95% CI: 48,780 to 83,333). ICURs for the stroke cohort varied slightly more dramatically. The MEPS-based algorithm by Franks et al. provided the lowest ICUR at 27,972 dollars/QALY (95% CI: 20,942 to 41,667). The Fryback and Shmueli algorithms provided ICURs that were greater than 50,000 dollars/QALY and did not have confidence intervals that overlapped with most of the other algorithms. The ICUR-based ranking of algorithms was strongly correlated between the asthma and stroke datasets (r = 0.60). CONCLUSION: SF-36/SF-12 preference-based algorithms produced a wide range of ICURs that could potentially lead to different reimbursement decisions. Brazier's SF-36 and SF-12 algorithms have a strong methodological and theoretical basis and tended to generate relatively higher ICUR estimates, considerations that support a preference for these algorithms over the alternatives. The "second-generation" algorithms developed from scores mapped from other indirect preference-based measures tended to generate lower ICURs that would promote greater adoption of new technology. There remains a need for an SF-36/SF-12 preference-based algorithm based on the US general population that has strong theoretical and methodological foundations.
Adult↗
High-performance exact algorithms for motif search.
OBJECTIVE: The human genome project has resulted in the generation of voluminous biological data. Novel computational techniques are called for to extract useful information from this data. One such technique is that of finding patterns that are repeated over many sequences (and possibly over many species). In this paper we study the problem of identifying meaningful patterns (i.e., motifs) from biological data, the motif search problem. METHODS: The general version of the motif search problem is NP-hard. Numerous algorithms have been proposed in the literature to solve this problem. Many of these algorithms fall under the category of heuristics. We concentrate on exact algorithms in this paper. In particular, we concentrate on two different versions of the motif search problem and offer exact algorithms for them. RESULTS: In this paper we present algorithms for two versions of the motif search problem. All of our algorithms are elegant and use only such simple data structures as arrays. For the first version of the problem described as Problem 1 in the paper, we present a simple sorting based algorithm, SMS (Simple Motif Search). This algorithm has been coded and experimental results have been obtained. For the second version of the problem (described in the paper as Problem 2), we present two different algorithms--a deterministic algorithm (called DMS) and a randomized algorithm (Monte Carlo algorithm). We also show how these algorithms can be parallelized. CONCLUSIONS: All the algorithms proposed in this paper are improvements over existing algorithms for these versions of motif search in biological sequence data. The algorithms presented have the potential of performing well in practice.
Algorithms↗
Symptom-based, algorithmic approach for handling the initial encounter with victims of a potential terrorist attack.
OBJECTIVES: This study intended to create symptom-based triage algorithms for the initial encounter with terror-attack victims. The goals of the triage algorithms include: (1) early recognition; (2) avoiding contamination; (3) early use of antidotes; (4) appropriate handling of unstable, contaminated victims; and (5) provisions of force protection. The algorithms also address industrial accidents and emerging infections, which have similar clinical presentations and risks for contamination as weapons of mass destruction (WMD). METHODS: The algorithms were developed using references from military and civilian sources. They were tested and adjusted using a series of theoretical patients from a CD-ROM chemical, biological, radiological/nuclear, and explosive victim simulator. Then, the algorithms were placed into a card format and sent to experts in relevant fields for academic review. RESULTS: Six inter-connected algorithms were created, described, and presented in figure form. The "attack" algorithm, for example, begins by differentiating between overt and covert attack victims (A covert attack is defined by epidemiological criteria adapted from the Centers for Disease Control and Prevention (CDC) recommendations). The attack algorithm then categorizes patients either as stable or unstable. Unstable patients flow to the "Dirty Resuscitation" algorithm, whereas, stable patients flow to the "Chemical Agent" and "Biological Agent" algorithms. The two remaining algorithms include the "Suicide Bomb/Blast/Explosion" and the "Radiation Dispersal Device" algorithms, which are inter-connected through the overt pathway in the "Attack" algorithm. CONCLUSION: A civilian, symptom-based, algorithmic approach to the initial encounter with victims of terrorist attacks, industrial accidents, or emerging infections was created. Future studies will address the usability of the algorithms with theoretical cases and utility in prospective, announced and unannounced, field drills. Additionally, future studies will assess the effectiveness of teaching modalities used to reinforce the algorithmic approach.
Algorithms↗
An objective procedure for evaluation of adaptive antifeedback algorithms in hearing aids.
OBJECTIVE: This study evaluated the performance of nine adaptive antifeedback algorithms. There were two goals: first, to identify objective procedures that are useful for evaluating these algorithms, and second, to identify strengths and weaknesses of existing algorithms. DESIGN: The algorithms were evaluated in behind-the-ear implementations on the Knowles Electronics Manikin for Acoustic Research (KEMAR). Different acoustic conditions were created by placing a telephone handset or a hat on KEMAR. Electroacoustic techniques were devised to measure the following performance aspects of each algorithm: (1) additional gain made available before oscillation, (2) gain lost in specific frequency regions, (3) reduction of suboscillatory peaks in the frequency response, (4) speed of adaptation to changing acoustic conditions, and (5) robustness in the presence of tonal input signals. RESULTS: For each measurement, performance varied widely across algorithms. No single algorithm was clearly superior or inferior to the others. Generally, the feedback cancellation algorithms were less likely to sacrifice gain in specific frequency regions and better at reducing suboscillatory peaks, whereas the algorithms that used noncancellation techniques were more tolerant of tonal input signals. For those algorithms equipped with special operational modes intended for music listening, the music mode improved the response to tonal inputs but sometimes sacrificed other performance aspects. Algorithms that required an acoustic measurement for initialization purposes tended to perform poorly in acoustic conditions dissimilar to the condition in which initialization was performed. CONCLUSIONS: The objective methods devised for this study appear useful for evaluating the performance of adaptive antifeedback algorithms. Currently available algorithms demonstrate a wide range of performance, and further research is required to develop new algorithms that combine the best features of existing algorithms.
Acoustics↗
Improved learning algorithms for mixture of experts in multiclass classification.
Mixture of experts (ME) is a modular neural network architecture for supervised learning. A double-loop Expectation-Maximization (EM) algorithm has been introduced to the ME architecture for adjusting the parameters and the iteratively reweighted least squares (IRLS) algorithm is used to perform maximization in the inner loop [Jordan, M.I., Jacobs, R.A. (1994). Hierarchical mixture of experts and the EM algorithm, Neural Computation, 6(2), 181-214]. However, it is reported in literature that the IRLS algorithm is of instability and the ME architecture trained by the EM algorithm, where IRLS algorithm is used in the inner loop, often produces the poor performance in multiclass classification. In this paper, the reason of this instability is explored. We find out that due to an implicitly imposed incorrect assumption on parameter independence in multiclass classification, an incomplete Hessian matrix is used in that IRLS algorithm. Based on this finding, we apply the Newton-Raphson method to the inner loop of the EM algorithm in the case of multiclass classification, where the exact Hessian matrix is adopted. To tackle the expensive computation of the Hessian matrix and its inverse, we propose an approximation to the Newton-Raphson algorithm based on a so-called generalized Bernoulli density. The Newton-Raphson algorithm and its approximation have been applied to synthetic data, benchmark, and real-world multiclass classification tasks. For comparison, the IRLS algorithm and a quasi-Newton algorithm called BFGS have also been applied to the same tasks. Simulation results have shown that the use of the proposed learning algorithms avoids the instability problem and makes the ME architecture produce good performance in multiclass classification. In particular, our approximation algorithm leads to fast learning. In addition, the limitation of our approximation algorithm is also empirically investigated in this paper.
Journal Article↗
Comparison of glaucomatous visual field defects using standard full threshold and Swedish interactive threshold algorithms.
OBJECTIVES: To compare the severity, size, and depth of glaucomatous visual field defects using standard full threshold (FT), Swedish interactive threshold algorithm (SITA) standard (SS), and SITA fast (SF) algorithms of the Humphrey perimeter. METHODS: A prospective observational case series of 77 patients with glaucoma performed FT, SS, and SF 30-2 white-on-white testing programs on the same day on 2 occasions for 1 month. The severity of defects was compared using the mean deviation, pattern standard deviation, Advanced Glaucoma Intervention Study, and Hodapp-Anderson-Parrish severity scores. The sizes of defects were compared using the total number of abnormal points on the pattern deviation plot that fit standard criteria for glaucomatous visual field defects. The depths of the defects were compared using the sum of the threshold values for points identified in the pattern deviation plot as fitting criteria for glaucomatous defects. RESULTS: The mean deviations were slightly better using the SS (-9.6 +/- 7.1 dB) or the SF (-9.1 +/- 6.7 dB) algorithm compared with the FT algorithm (-10.3 +/- 7.1 dB) (P<.005). There were no significant differences in pattern standard deviations between SS (8.6 +/- 4.0, P =.08) and SF (8.1 +/- 3.6, P =.19) compared with FT (8.3 +/- 3.3), although the pattern standard deviation was higher in SS fields compared with SF fields (P<.001). Advanced Glaucoma Intervention Study scores were slightly better when the SS (7.5 +/- 5.6) or SF (7.2 +/- 5.4) algorithm was used compared with the FT algorithm (8.6 +/- 5.4) (P<.001). The sizes of glaucomatous defects were slightly larger using the SS (20.9 +/- 10.7) algorithm compared with the FT algorithm (19.2 +/- 10.9) (P =.004) but not the SF algorithm (20.0 +/- 10.6) (P =.11). The depth of defects measured by the SS (220.4 +/- 108.0 dB) and SF (219.8 +/- 101.3 dB) algorithms was significantly shallower compared with that measured by the FT algorithm (152.3 +/- 79.1 dB) (P<.001). There were no significant differences in Hodapp-Anderson-Parrish severity scores among algorithms (P =.12). CONCLUSIONS: Glaucomatous defects are measured significantly shallower using the new SITA algorithms but are approximately the same size and severity compared with FT measurements. Care should be taken when using threshold values to compare glaucomatous defects in a patient when converting from FT to SITA algorithms.
Adult↗
Comparison of 2D and 3D algorithms for adding a margin to the gross tumor volume in the conformal radiotherapy planning of prostate cancer.
PURPOSE: To evaluate the adequacy of tumor volume coverage using a three-dimensional (3D) margin-growing algorithm compared to a two-dimensional (2D) margin-growing algorithm in the conformal radiotherapy planning of prostate cancer. METHODS AND MATERIALS: Two gross tumor volumes (GTV) were segmented in each of 10 patients with localized prostate cancer; prostate gland only (PO) and prostate with seminal vesicles (PSV). A predetermined margin of 10 mm was applied to these two groups (PO and PSV) using both 2D and 3D margin-growing algorithms. The 2D algorithm added a transaxial margin to each GTV slice, whereas the 3D algorithm added a volumetric margin all around the GTV. The true planning target volume (PTV) was defined as the region delineated by the 3D algorithm. The adequacy of geometric coverage of the GTV by the two algorithms was examined in a series of transaxial planes throughout the target volume. RESULTS: The 2D margin-growing algorithm underestimated the PTV by 17% (range 12-20) in the PO group and by 20% (range 13-28) for the PSV group when compared to the 3D-margin algorithm. For the PO group, the mean transaxial difference between the 2D and 3D algorithm was 3.8 mm inferiorly (range 0-20), 1.8 mm centrally (range 0-9), and 4.4 mm superiorly (range 0-22). Considering all of these regions, the mean discrepancy anteriorly was 5.1 mm (range 0-22), posteriorly 2.2 (range 0-20), right border 2.8 mm (range 0-14), and left border 3.1 mm (range 0-12). For the PSV group, the mean discrepancy in the inferior region was 3.8 mm (range 0-20), central region of the prostate was 1.8 mm ( range 0-9), the junction region of the prostate and the seminal vesicles was 5.5 mm (range 0-30), and the superior region of the seminal vesicles was 4.2 mm (range 0-55). When the different borders were considered in the PSV group, the mean discrepancies for the anterior, posterior, right, and left borders were 6.4 mm (range 0-55), 2.5 mm (range 0-20), 2.6 mm (range 0-14), and 3.9 mm (range 0-45), respectively. Underestimation of the required margin with the 2D algorithm occurred when the transaxial definition of the GTV shifted in position significantly between successive adjacent slices, resulting in transaxial discrepancies of up to 22 mm and 55 mm, respectively, for the PO and PSV groups. In the superior regions, the 2D algorithm was inadequate, often providing a margin of less than 3 mm compared to the 10 mm margin delineated by the 3D algorithm. CONCLUSION: This study illustrates that target margins added by a laminar method in the transaxial plane are inadequate for covering a 3D tumor volume so that a margin-growing algorithm which fully takes into account the 3D shape of the GTV should be used. If a 2D-margin method is utilized, an appreciation of spatial margins in 3D is required.
Aged↗