The prevalence of Barrett's esophagus in the US: estimates from a simulation model confirmed by SEER data.

Barrett's esophagus (BE) is the precursor and the biggest risk factor for esophageal adenocarcinoma (EAC), the solid cancer with the fastest rising incidence in the US and western world. Current strategies to decrease morbidity and mortality from EAC have focused on identifying and surveying patients with BE using upper endoscopy. An accurate estimate of the number of patients with BE in the population is important to inform public health policy and to prioritize resources for potential screening and management programs. However, the true prevalence of BE is difficult to ascertain because the condition frequently is symptomatically silent, and the numerous clinical studies that have analyzed BE prevalence have produced a wide range of estimates. The aim of this study was to use a computer simulation disease model of EAC to determine the estimates for BE prevalence that best align with US Surveillance Epidemiology and End Results (SEER) cancer registry data. A previously developed mathematical model of EAC was modified to perform this analysis. The model consists of six health states: normal, gastroesophageal reflux disease (GERD), BE, undetected cancer, detected cancer, and death. Published literature regarding the transition rates between these states were used to provide boundaries. During the one million computer simulations that were performed, these transition rates were systematically varied, producing differing prevalences for the numerous health states. Two filters were sequentially applied to select out superior simulations that were most consistent with clinical data. First, among these million simulations, the 1000 that best reproduced SEER cancer incidence data were selected. Next, of those 1000 best simulations, the 100 with an overall calculated BE to Detected Cancer rates closest to published estimates were selected. Finally, the prevalence of BE in the final set of best 100 simulations was analyzed. We present histogram data depicting BE prevalences for all one million simulations, the 1000 simulations that best approximate SEER data, and the final set of 100 simulations. Using the best 100 simulations, we estimate the prevalence of BE to be 5.6% (5.49-5.70%). Using our model, an estimated prevalence for BE in the general population of 5.6% (5.49-5.70%) accurately predicts incidence rates for EAC reported to the US SEER cancer registry. Future clinical studies are needed to confirm our estimate.

Monitoring of radio frequency tissue ablation in an interventional magnetic resonance environment. Preliminary ex vivo and in vivo results.

RATIONALE AND OBJECTIVES: The authors evaluate the feasibility of monitoring radio frequency (RF) ablation in an interventional, open-configuration, 0.5-tesla magnetic resonance (MR) environment. METHODS: Ex vivo and in vivo RF coagulation necrosis were induced in porcine paraspinal muscle tissue using a 300 kHz monopolar RF generator applying 5 to 20 W over 3 to 9 minutes. Images were acquired simultaneous to RF application, after RF application, and in an intermittent mode (60 seconds of RF followed by 15 seconds of MR imaging). Temperature changes were monitored based on amplitude (ex vivo) and phase alterations (in vivo) of a T1-weighted graded refocused echo (GRE) sequence enabling an update every 2.5 seconds. A standardized color-coded subtraction technique enhanced signal changes. Additionally, T2- and T1-weighted spin echo (SE) images were acquired with and without intravenous contrast. Macroscopic coagulation size was compared with lesion size seen on MR images. RESULTS: Although lesion diameters were related directly to applied RF power, the application mode had no significant impact on coagulation size (P > 0.05). As could be expected, MR imaging during RF ablation resulted in major image distortion. Radio frequency effects were seen on images acquired in the continuous and intermittent modes. Coagulation size seen on GRE images correlated well with macroscopy both ex vivo (r = 0.89) and in vivo (r = 0.92). Poorer correlation was found with postinterventional SE sequences (r = 0.78-0.84). CONCLUSIONS: Magnetic resonance monitoring of RF effects is feasible both ex vivo as well as in vivo using temperature-sensitive sequences in an open-configuration MR environment.

Evaluation of selected two-dimensional segmentation techniques for computed tomography quantitation of lymph nodes.

RATIONALE AND OBJECTIVES: As contrast agents that selectively target normal lymph nodes are undergoing development and evaluation, it has become important to accurately and reproducibly determine nodal boundaries to study the agents and determine such values as lymph node area or mean nodal contrast concentration. This study was performed to evaluate the accuracy of different two-dimensional computer segmentation methods, tested on acrylic phantoms constructed to imitate the appearance of lymph nodes surrounded by fat. METHODS: Five segmentation techniques (manual tracing, semiautomatic local criteria threshold selection, Sobel/watershed technique, interactive deformable contour algorithm and thresholding) were evaluated using phantoms. Subsequently, the first three methods were applied to the images of enhanced lymph nodes in rabbits. RESULTS: Minimum errors in phantom area measurement (< 5%) and interoperator variation (< 5%) were seen with the Sobel/watershed technique and the interactive deformable contour algorithm. These two techniques were significantly better than thresholding and semiautomated thresholding based on local properties. CONCLUSION: Methods based on Sobel edge detection offer more objective tools than thresholding methods for segmenting objects similar to lymph nodes in computed tomography images. Both methods, Sobel/watershed and interactive deformable contour algorithm, are fast and have simple user interfaces.

Quantitative high-resolution measurement of cerebrovascular physiology with slip-ring CT.

PURPOSE: To implement and validate spiral slip-ring CT for use in cerebrovascular studies. METHODS: Continuous data were acquired from an experimental, first-pass, iodine contrast, bolus study by unidirectional X-ray tube rotation, and images were reconstructed at 100-millisecond intervals. Functional maps of cerebral blood volume (CBV) and cerebral blood flow (CBF) were constructed with voxel-by-voxel gamma variate fitting. Reproducibility studies, different injection volumes and sites, and CO2 challenge were applied to verify the technique. RESULTS: Average absolute cortical gray and white matter and basal ganglia results were reproducible within +/- 0.8 ml/100 g for CBV and +/- 20 ml/100 g per minute for CBF, CBV response to changing arterial CO2 tension was significant only in cortical gray matter and basal ganglia; CBF response was significant in gray and white matter, as well as in the basal ganglia. CONCLUSION: Functional CT and constructed functional maps provide an optimal, high-resolution tool with which to visualize cerebrovascular parameters and their changes.

Abdominal abscess. Imaging and intervention.

The diagnosis and management of patients with abdominal abscesses continue to challenge those involved in imaging and interventional techniques. In this article, the authors discuss the use of ultrasonography and CT scans in the work-up of patients suspected of having abdominal abscesses and describe techniques useful for their management preoperatively.

Imaging-guided percutaneous abdominal biopsy.

Guided percutaneous biopsy has become an important means of diagnosing abdominal lesions. This technique provides necessary pathologic information without resorting to open biopsy. Fluoroscopy, sonography, computed tomography or magnetic resonance imaging may be used for needle guidance, depending on the patient's condition and the size and location of the lesion. These variables also influence selection of needle gauge, tip configuration, and sampling mechanism. The authors describe appropriate techniques for percutaneous biopsy of hepatic, renal, pancreatic, adrenal, and retroperitoneal lesions.

Block-copolymer of polyethylene glycol and polylysine as a carrier of organic iodine: design of long-circulating particulate contrast medium for X-ray computed tomography.

In order to obtain small, polymer-stabilized particulate carriers for organic iodine to serve as a contrast agent for X-ray computed tomography (CT) an attempt was made to design a carrier based on polymeric micelles. Here we describe the synthesis of an iodine-containing amphiphilic block-copolymer which can micellize in aqueous solutions. The two blocks of the copolymer consisted of methoxypoly(ethyleneglycol) and poly[epsilon,N-(triiodobenzoyl)-L-lysine]. Upon dispersion in water, the block copolymer formed particles with average diameter 80 nm and iodine content up to 44.7%. The particles start to dissociate to the individual polymeric chains in the concentration range of 0.05-0.5 microM in water at 23 degrees C. Upon intravenous injection at 250 mg of iodine/kg (570 mg of the agent/kg) in rabbits the medium demonstrated exceptional 24 hr half-life in the blood substantiating corona/core structure of the particles with PEG chains protecting the iodine-containing core. The possible use of these particulates as contrast medium for X-ray computed tomography is discussed.

Representing both first- and second-order uncertainties by Monte Carlo simulation for groups of patients.

Actual implementation of probabilistic sensitivity analysis may lead to misleading or improper conclusions when it is applied to groups of patients rather than individual patients. The practice of combining first- and second-order simulations when modeling the outcome for a group of more than one patient yields an erroneous marginal distribution whenever the parameter values are randomly sampled for each patient while the results are presented as simulated means for the group of patients. This practice results in underrepresenting the second-order uncertainty. It may also distort the shape (especially the symmetry or extent of the tails) in the simulated distribution. As a result, it may lead to premature or incorrect conclusions of superiority. It may also result in inappropriate estimates of the value of further research to inform parameter values.

A cost--utility analysis of treatment options for inguinal hernia in 1,513,008 adult patients.

BACKGROUND: The controversial issue of the cost-effectiveness of laparoscopic inguinal hernia repair is examined, employing a decision analytic method. MATERIALS AND METHODS: The NSAS, NHDS (National Center for Health Statistics), HCUP-NIS (Agency for Healthcare Research and Quality) databases and 51 randomized controlled trials were analyzed. The study group constituted of a total of 1,513,008 hernia repairs. Projection of the clinical, economic, and quality-of-life outcomes expected from the different treatment options was done by using a Markov Monte Carlo decision model. Two logistic regression models were used to predict the probability of hospital admission after an ambulatory procedure and the probability of death after inguinal hernia repair. Four treatment strategies were modeled: (1) laparoscopic repair (LR), (2) open mesh (OM), (3) open non-mesh (ONM), and (4) expectant management. Costs were expressed in US dollars and effectiveness in quality-adjusted life years (QALYs). The main outcome measures were the average and the incremental cost-effectiveness (ICER) ratios. RESULTS: Compared to the expectant management, the incremental cost per QALY gained was 605 dollars (4086 dollars, 9.04 QALYs) for LR, 697 dollars (4290 dollars, 8.975 QALYs) for OM, and 1711 dollars (6200 dollars, 8.546 QALYs) for ONM. In sensitivity analysis the two major components that affect the cost-effectiveness ratio of the different types of repair were the ambulatory facility cost and the recurrence rate. At a LR ambulatory facility cost of 5526 dollars the ICER of LR compared to OM surpasses the threshold of 50,000 dollars/QALY. CONCLUSIONS: On the basis of our assumptions this mathematical model shows that from a societal perspective laparoscopic approach can be a cost-effective treatment option for inguinal hernia repair.

A model of focal cancer in rabbit lymph nodes.

RATIONALE AND OBJECTIVES: We sought to develop a model of focal cancer in the rabbit lymph node. METHODS: Under computed tomography (CT) guidance, 4-5 million VX2 cells were directly injected into the popliteal nodes of 14 anesthetized New Zealand White rabbits. Fifteen to 18 days later, percutaneous lymphography was performed with CT scanning using radiopaque nanoparticulates and massage. Histologic correlation also was obtained. RESULTS: In 12 of the 14 animals, focal lesions were successfully created within (n = 6) and adjacent (n = 6) to the node, and all animals appeared to be healthy when euthanized. Within 15 min after massage, CT lymphography showed homogeneous enhancement of normal node regions and no enhancement of cancer. There was good agreement between histology and lymphography. CONCLUSION: This method is suitable as a model to test for diagnostic and therapeutic interventions.

Radio-frequency tissue ablation of VX2 tumor nodules in the rabbit lung.

RATIONALE AND OBJECTIVES: The authors investigated whether small pulmonary malignancies could be treated with computed tomography (CT)-guided, percutaneously placed radio-frequency (RF) electrodes. METHODS: Pulmonary tumors were created in 11 New Zealand white rabbits by using CT-guided injection of a VX2 sarcoma cell suspension into the lower portion of the right lung. Tumors were allowed to grow 14-21 days to achieve a diameter of 6-12 mm. Electrodes were placed coaxially into the tumors via insulated 19-gauge Turner needles. Seven tumors were treated with RF for 6 minutes at 90 degrees C. Four tumors served as controls and were not treated. Follow-up CT and histopathologic analysis were performed on days 0-28. Specimens from treated rabbits were examined histopathologically on days 0 and 3 (n = 2 each), and days 1, 5, and 28 (n = 1 each). RESULTS: Immediately following treatment, CT images showed rounded opacities enveloping the tumor. This corresponded histologically to coagulation necrosis of tumor and surrounding alveoli. In all cases, at least 95% of treated tumor nodules were necrotic at histopathologic analysis. Peripheral residual nests of histologically viable tumor were seen in three rabbits (43%). Control rabbits showed growing tumor nodules without necrosis at autopsy (mean survival, 23 days after inoculation). Two RF-treated rabbits (29%) and one control rabbit (25%) had pneumothoraces. CONCLUSION: Percutaneous RF tissue ablation can be used to successfully treat small parenchymal tumor nodules within the lung in an animal model.

Radiofrequency tissue ablation: increased lesion diameter with a perfusion electrode.

RATIONALE AND OBJECTIVES: We sought to induce large zones of coagulation necrosis using radiofrequency (RF) with perfusion electrodes and to define optimal parameters for this system. METHODS: We developed RF electrodes with internal cannulas to enable tip perfusion. Lesions were created with monopolar RF in ex vivo and in vivo liver and muscle tissue with and without perfusion of the electrode tip using 0 degree C saline. In separate experiments, wattage, current, procedure duration, tip exposure, and perfused tip temperatures were studied. RESULTS: In ex vivo liver tissue, a maximum lesion diameter of 3.1 cm without charring occurred with perfusion at 12 min and 50 W. In in vivo liver tissue with perfusion (tip temperature = 25-35 degrees C) and a 3-cm tip exposure, 80 W were deposited in muscle tissue and 65 W in liver tissue for 12 min without inducing charring. Lesion diameters were 4.5 cm and 2.4 cm, respectively. By comparison, without perfusion a maximum of 20 W could be deposited into either tissue type, resulting in 1.8-cm muscle lesions and 1.2-cm liver lesions. Tip temperatures between 45 degrees C and 55 degrees C resulted in charring. Smaller but predictable lesion diameters were created with a lower power, a shorter tip exposure, or both. Of all the parameters, diameter correlated best with the current applied. CONCLUSION: Perfusion of RF electrodes with chilled saline allows for increased power deposition without tissue charring, increasing the volume of coagulation necrosis created with a single electrode insertion. Perfusion electrodes therefore might decrease the number of probe insertions required for percutaneous tumor ablation therapy or allow for the treatment of larger lesions.

Inpatient costs of routine endovascular repair of abdominal aortic aneurysm.

RATIONALE AND OBJECTIVES: The purpose of this study was to determine the inpatient cost of routine (ie, without emergent conversion to open repair during the hospital stay) endovascular stent-graft placement in a consecutive series of patients undergoing elective endovascular repair of abdominal aortic aneurysm (AAA) at a single institution. MATERIALS AND METHODS: Inpatient hospital costs of 91 patients who underwent initial elective endovascular repair of AAA were analyzed retrospectively. All patients had participated in clinical trials at the authors' institution during the previous 6 years. Financial data were derived from the hospital's cost-accounting system; additional procedural data were collected from a departmental database and with chart review. Stent-graft and professional costs were excluded. RESULTS: The mean total cost for endovascular repair was $11,842 (standard deviation [SD], $5,127), mean procedure time was 149 minutes (SD, 79 minutes), and mean length of stay was 3.5 days (SD, 2.3 days). Total cost depended on stent-graft type (means, $12,428 [bifurcated] vs $9,622 [tube]; P = .0002) and strongly correlated with procedure time and length of hospital stay (r = 0.78 and 0.66, respectively; P < .0001). Ninety-six percent of total costs for all patients were attributable to the following departments: operating theater (31%), radiology (31%), nursing (22%), and anesthesia (12%). CONCLUSION: Overall costs are greater with bifurcated than with tube stent-grafts. Total procedure-related costs are divided relatively equally between the operating theater, the radiology department, and the combination of the nursing and anesthesia departments.

Radiofrequency tissue ablation in the rabbit lung: efficacy and complications.

RATIONALE AND OBJECTIVES: We assessed the feasibility and safety of performing percutaneous radiofrequency ablation of pulmonary tissue in rabbits. METHODS: Using an aseptic technique and computed tomography (CT) guidance, insulated 19-gauge aspiration biopsy needles were inserted into the right lower lobe of eight New Zealand White rabbits. Radiofrequency was applied via a coaxial electrode for 6 min at 90 degrees C. Probe-tip temperature, tissue impedance, and wattage were recorded at baseline and at 60-sec intervals throughout the procedure. CT scanning was used to assess tissue destruction and the presence or absence of pneumothorax immediately after the procedure and at 24 hr, 3 days, 10 days, 21 days, and 28 days. Three rabbits were sacrificed immediately, and the remaining rabbits were euthanized at 24 hr and at 3 days. 10 days, and 28 days (two rabbits). Gross and microscopic pathology were obtained and correlated with CT findings. RESULTS: The mean initial tissue impedance was 509 +/- 197 omega, marked changes in tissue impedance were found during the procedure (240-1380 omega). Rigid temperature control required continuous manual fine-tuning of generator output. Increased respiratory rate was noted in one rabbit during the first 30 sec of radiofrequency application. Homogeneous, ovoid opacities 8.4 +/- 2.4 mm in diameter and 1.4 +/- 0.1 cm in length were found by CT scanning immediately after the procedure. These opacities showed maximal consolidation at 3 days, corresponding to coagulative necrosis and a peripheral acute inflammatory reaction. At 10 days, peripheral hyperattenuation with central hypoattenuation (early fibrosis surrounding degenerating blood products) was seen. Minimal residual fibrosis, pleural scarring, or both were noted by 28 days, suggesting a rapid, near-total recovery from the procedure. Lesion sizes were within 2 mm of gross pathologic findings. Pneumothoraces were noted in three of the eight rabbits (37.5%). CONCLUSION: Radiofrequency tissue ablation was safely performed in pulmonary parenchyma via a percutaneous, transthoracic approach using a coaxial needle technique. Tissue response to thermal injury was predictable and easily monitored by CT scanning with excellent radiologic-pathologic correlation.

Functional CT with an experimental intravascular contrast agent in the assessment of liver vascular physiology.

RATIONALE AND OBJECTIVES: We evaluated liver vascular physiology with a functional spiral computed tomography (CT) technique and an intravascular contrast agent. METHODS: Eleven rabbits were studied by means of continuous 40-second single-section data acquisition after bolus injection of an experimental contrast agent. Sequential images were reconstructed at 200-msec intervals. Aortic, portal and hepatic venous, and liver time-HU curves were obtained. From these, hepatic blood volume and flow, tissue transit times, and arterial and portal contributions to total liver blood supply were assessed. RESULTS: The following measures were obtained: hepatic blood volume fraction, 0.33 +/- 0.03 (mean +/- standard error); total flow, 241.1 mL/min +/- 33.6 per 100 g of tissue (arterial component, 11.3 mL/min +/- 3.0 per 100 g of tissue; portal component, 226.4 mL/min +/- 30.7 per 100 g of tissue); arterial transit time, 8.7 seconds +/- 1.6; portal transit time, 8.7 seconds +/- 1.3; arterial to portal perfusion ratio, 0.06 +/- 0.01; and calculated arterial and portal perfusion indexes, 0.05 +/- 0.01 and 0.95 +/- 0.01, respectively. CONCLUSION: Functional CT is a promising, high-resolution tomographic imaging technique for evaluating liver perfusion.

Liver tumor ablation: real-time monitoring with dynamic CT.

RATIONALE AND OBJECTIVES: To determine whether incomplete contact of ethanol with tumor limits the success of percutaneous ethanol injection therapy. MATERIALS AND METHODS: Percutaneous ethanol injection was performed in seven normal New Zealand white rabbits and 18 rabbits with 1-3-cm liver tumors 10-14 days after percutaneous implantation of suspended tumor cells. A 3-5 mL dose of ethanol was injected at a rate of 0.2 mL/sec either into normal liver remote from large vessels or directly into tumor. During and immediately after injection, axial, 2-mm-thick, contrast material-enhanced computed tomography scans were obtained at reach of three levels every 9 seconds. RESULTS: In normal animals, virtually all injected ethanol tracked to the hepatic capsule. As ethanol was injected into tumors, peripheral tracking, similar to that seen in normal livers, or extratumoral puddling was observed. Ethanol-tumor contact was incomplete in 16 of 18 animals (89%). Histopathologic analysis showed incomplete tumor necrosis. CONCLUSION: In this model of hepatic carcinoma metastasis, the tumor failed to hold sufficient ethanol for successful ablation by means of percutaneous ethanol injection therapy.

Percutaneous computed tomography lymphography in the rabbit by subcutaneously injected nanoparticulates.

RATIONALE AND OBJECTIVES: Surgical lymphangiography is infrequently used in staging cancer because of its inherent limitations. Radiopaque nanoparticulates target lymph nodes draining interstitial tissues and could make percutaneous lymphography feasible. METHODS: Experimental nanoparticulate contrast agent formulations were injected subcutaneously in the forepaw or hindpaw of normal rabbits or rabbits with induced reactive nodal hyperplasia. Axillary and popliteal nodes were imaged with thin-section computed tomography (CT) using quantitative methods to measure node enhancement. Dose-response (0.1-2.0 ml) and time course (4 hr to 10 weeks) of enhancement were assessed. RESULTS: Nodal enhancement above 100 Hounsfield units was consistently obtained. Enhancement was significantly related to dose and peaked at 10 hr with slow washout over the observation period. Nodes with reactive hyperplasia were larger and had heterogeneous enhancement patterns distinctly different from normal nodes. CONCLUSION: Percutaneous CT lymphography effectively depicts the macroscopic intranodal architecture in rabbits.

Time-lapse quantitative computed tomography lymphography: assessing lymphatic function in vivo.

RATIONALE AND OBJECTIVES: Immobility and massage produce different local limb lymph flow rates. We studied their influence on accumulation of radiopaque nanoparticulates in regional lymph nodes of normal rabbits. METHODS: Quantitative lymphography at 10-min intervals was used to follow the transport of subcutaneous (s.c.) nanoparticulates produced from insoluble esters of diatrizoic acid. In one design, both hindpaws received 0.5 ml of nanoparticulate s.c., and one hindpaw was massaged. In a second design, one hindpaw was injected and massaged while imaging the popliteal, presacral, and paraaortic nodes every 10 min. RESULTS: Gentle massage rapidly increased popliteal node accumulation in comparison with the immobile limb. On the massaged side, mean Hounsfield (HU) units, maximum Hounsfield units, and calculated iodine were significantly greater at 10 min and all subsequent times. In the node transfer experiments, it took 12, 30, and 45 min, respectively, to obtain 100-HU mean attenuation; 200-HU maximum attenuation thresholds were achieved at 20, 47, and 69 min, respectively. CONCLUSION: Quantitative computed tomography lymphography reflects local lymph physiology. Gentle massage of the s.c. injection site is a powerful lymphotropic stimulus.

Nanoparticulate computed tomography contrast agents for blood pool and liver-spleen imaging.

RATIONALE AND OBJECTIVES: We investigated the properties of a group of iodine-containing, insoluble compounds formulated as nanoparticles for use as potential blood pool and liver-spleen contrast agents. METHODS: High-resolution, quantitative computed tomography (CT) was performed prior to and at intervals following the intravenous administration of the contrast agents to rabbits. Time-density characteristics for three organs were evaluated. RESULTS: Excellent enhancement of blood (< or = 232 Hounsfield units [HU]), liver (< or = 263 HU), and spleen (< or = 350 HU) was achieved at the administered dose of 3.0 ml/kg. The composition of the agents influenced the biodistribution, as well as the residence time in blood, and time to peak enhancement in liver. CONCLUSION: Iodinated nanoparticulate compounds are promising CT contrast agents. Development of agents with desirable pharmacokinetic and biodistribution profiles may permit application-specific contrast enhancement.

Radiofrequency tissue ablation: importance of local temperature along the electrode tip exposure in determining lesion shape and size.

RATIONALE AND OBJECTIVES: We determined whether heat distribution along a radiofrequency (RF) electrode would be uniform when longer tip exposures are used and whether local temperature effects would influence the shape of induced tissue coagulation. METHODS: Thermistors were embedded within 18-gauge RF electrodes at both ends and in the middle of the exposed tip. The length of tip exposure varied from 1 to 7 cm. RF was applied in vitro to pig liver for 6 min using a constant tip temperature, which was varied in 10 degrees C increments from 60 degrees C to 110 degrees C. Experiments were performed in triplicate. The 3- and 5-cm probes were used at a 90 degrees C tip temperature to create lesions in live pig liver and muscle using similar parameters. Temperature was measured throughout the procedure. Observable coagulation necrosis was measured at the end of the treatment. Regression analysis was used to evaluate the local temperature-lesion diameter relationship. RESULTS: Temperatures were not uniform along the tip exposure for any given trial. Temperature variation increased with higher tip temperatures and longer tip exposures. The diameter of local coagulation necrosis was a function of the local mean temperature. For in vitro trials, no coagulation was seen when the local temperature was less than 50 degrees C. Temperatures above this threshold resulted in progressively greater lesion diameter, with a minimum of 1 cm of necrosis occurring at 71 degrees C. Additional increases in lesion diameter (1.4-1.6 cm) were observed at approximately 90 degrees C. Mathematical modeling demonstrated a best-fit curve: lesion diameter (in cm) = ¿1.4 + 0.03 (tip exposure)¿ ¿1 - e [-0.067(local temp - 49.5 degrees C)]¿, r2 = .986, SD = 0.14 cm for each curve. In living tissue, less uniformity in the shape of coagulation necrosis was seen around the electrodes. Local temperature-lesion diameter data fit the same logarithmic relation, but the threshold for coagulation necrosis was 8.5 degrees C higher than for in vitro specimens. CONCLUSION: Using a single-probe technique for RF-induced tissue necrosis, the diameter of tissue coagulation may be predicted by the local temperature along the exposed electrode. The uniformity of temperature decreases with increased tip exposures. This effect may be partially corrected by creating lesions at higher tip temperatures, where necrosis diameter is increased. Because effects are more pronounced in vivo, uniform volumes of tissue necrosis are limited to tip exposures of 3 cm or less.