Evaluation of tumor measurements in oncology: use of film-based and electronic techniques

PURPOSE

To evaluate the variability in bidimensional computed tomography (CT) measurements obtained of actual tumors and of tumor phantoms by use of three measurement techniques: hand-held calipers on film, electronic calipers on a workstation, and an autocontour technique on a workstation.

MATERIALS AND METHODS

Three radiologists measured 45 actual tumors (in the lung, liver, and lymph nodes) on CT images, using each of the three techniques. Bidimensional measurements were recorded, and their cross-products calculated. The coefficient of variation was calculated to assess interobserver variability. CT images of 48 phantoms were measured by three radiologists with each of the techniques. In addition to the coefficient of variation, the differences between the cross-product measurements of tumor phantoms themselves and the measurements obtained with each of the techniques were calculated.

RESULTS

The differences between the coefficients of variation were statistically significantly different for the autocontour technique, compared with the other techniques, both for actual tumors and for tumor phantoms. There was no statistically significant difference in the coefficient of variation between measurements obtained with hand-held calipers and electronic calipers. The cross-products for tumor phantoms were 12% less than the actual cross-product when calipers on film were used, 11% less using electronic calipers, and 1% greater using the autocontour technique.

CONCLUSION

Tumor size is obtained more accurately and consistently between readers using an automated autocontour technique than between those using hand-held or electronic calipers. This finding has substantial implications for monitoring tumor therapy in an individual patient, as well as for evaluating the effectiveness of new therapies under development.

Voice recognition in radiology reporting

OBJECTIVE

Recent advances in computer hardware and software technology have improved voice recognition systems used for radiology reporting. We describe and analyze this technology and some of its costs as well as the prospects for using voice recognition systems in the transcription of radiology reports. Factors to be considered in choosing a system include language model, speech flow, vocabulary size, and methods of acoustic modeling.

CONCLUSION

Careful evaluation of individual radiology practice will permit proper introduction of voice recognition technology into mainstream clinical use. With this new technology, radiologists can use normal speech patterns to dictate while reviewing films. Also, this technology can use standard personal computers to reduce hardware costs to a level acceptable to radiology departments.

Analysis of prostate and seminal vesicle motion: implications for treatment planning

PURPOSE

To quantify prostate and seminal vesicle positional changes (target motion) between treatment planning and delivery, and to identify the factors contributing to target motion.

METHODS AND MATERIALS

Thirty patients with adenocarcinoma of the prostate were prospectively evaluated by analyzing two sequential planning computerized tomography (CT) scans (S1, obtained prior to treatment, and S2, obtained during the fourth week of treatment) for each patient. All anatomical volumes of interest (soft tissue and bony) were reconstructed from transverse CT images and projected onto anterior and lateral beam's-eye view projections. Positional changes between S1 and S2 were eliminated by applying a rigid body translation and rotation. Target motion was then measured by recording the positional change between S1 and S2 at the edges (right, left, superior, inferior). Potential correlation of target motion with bladder volume, rectal volume, and rectal diameter changes were evaluated by linear regression analysis.

RESULTS

Neither the prostate nor seminal vesicles remained fixed with respect to bony anatomy between S1 and S2. The distribution of positional changes were generally small (< 0.5 cm), but maximum displacements of 1.5-2.2 cm did occur, particularly in the lateral view. In this study, bladder volume changes between the scans were small and did not correlate with target motion (P = 0.67). Both rectal volume and rectal diameter changes correlated with target motion for both the prostate (p = 0.004 and 0.005, respectively) and seminal vesicles (p < 0.001 and < 0.001, respectively). However, neither the initial rectal volume nor the initial rectal diameter could be used to predict subsequent target motion when evaluated either singly or as part of a multiple regression model.

CONCLUSIONS

Target motion occurs during the course of treatment planning and delivery and should be considered when designing conformal radiation fields. Although the target position at the time of planning CT may differ substantially from the mean treatment position, target motion cannot be predicted by evaluating simply measured parameters from a single scan, or double scan sequence.

The reliability of optimization under dose-volume limits

PURPOSE

An optimization algorithm improves the distribution of dose among discrete points in tissues, but tolerance depends on the distribution of dose across a continuous volume. This report asks whether an exact algorithm can be completed when enough points are taken to accurately model a dose-volume constraint.

METHODS AND MATERIALS

Trials were performed using a 3-dimensional model of conformal therapy of lung cancer. Trials were repeated with different limits placed on the fraction of lung which could receive > 20 Gy. Bounds were placed on cord dose and target dose inhomogeneity. A mixed integer algorithm was used to find a feasible set of beam weights which would maximize tumor dose. Tests of feasibility and optimality are introduced to check the solution accuracy.

RESULTS

Solutions were optimal for points used to model tissues. An accuracy of 3-4% in a volume condition could be obtained with models of 450-600 points. The error improved to 2% with 800 points to model the lung. Solution times increased six-fold at this level of accuracy.

CONCLUSION

The mixed integer method can find optimum weights which respect dose-volume conditions in usually acceptable times. If constraints are violated by an excessive amount, the optimization model should be rerun with more points.

CCRT (computer controlled radiation therapy) for non-small cell lung cancer: sensitivity of clinical gains to organ tolerance restrictions

A method is described for allowing the entrance angles of beams to vary along the longitudinal axis of the target volume using a computer controlled radiation therapy (CCRT) treatment technique. The increase in tumor dose potentially available with this technique is evaluated for the case of a non-small cell carcinoma of the lung for which treatment is constrained by a clinically relevant tolerance condition. Tolerance is expressed in terms of that volume of unaffected contralateral lung allowed to receive greater than 20 Gy. The evaluation is repeated with small changes made to the constraining condition in order to explore the sensitivity of the dose gain to the definition of tolerance. For the chosen example, at least an additional 10 Gy could be prescribed to the tumor with the CCRT technique when the 20 Gy contralateral lung volume was held to less than 30%. A gain of 7 Gy was found when the 20 Gy volume was held to 27% and 1 Gy when the 20 Gy volume was reduced to 25%. CCRT treatment can result in a significant dose gain under a clinically relevant condition for tolerance. Its measured advantage strongly depends on the dose constraints specified. Sensitivity analysis is necessary to meaningfully evaluate the merits of alternative treatment techniques.

The effect on minimum tumor dose of restricting target-dose inhomogeneity in optimized three-dimensional treatment of lung cancer

An examination was made of the effect upon the minimum tumor dose of a limit placed on the variation of dose across target. If the required level of target dose uniformity is slightly relaxed, a substantial improvement in the minimum tumor dose might appear. It was conjectured that this effect could be seen with treatments optimally planned and evaluated in three-dimensions. A model of advanced carcinoma of the lung treated with a computer controlled accelerator was used to test this hypothesis. A mathematical program for optimizing beam weights was used to determine the largest minimum tumor dose possible. In the six cases tested, a minimum tumor dose of greater than 80 Gy could be delivered if a 20% inhomogeneity limit was accepted. The minimum tumor dose fell to the range 44-64 Gy when the inhomogeneity limit was tightened to 13-17%. The results imply a need to examine the choice of a required level of dose uniformity from the range of values suggested in the 2-dimensional planning literature. If a strict bound-on-dose uniformity is preserved, mechanisms--such as formal optimization--which can reduce target dose inhomogeneity will be valuable.

Perspective display of patient external and internal contours

The two-dimensional perspective display of three-dimensional patient contour data is useful in radiotherapy treatment planning as it provides a comprehensive view of the relative positions of patient internal (organ and target) and external contours. The advent of minicomputers and video display systems has made possible the integration of such display techniques into the treatment planning routine. This paper describes and gives an example of the methods we have found useful in producing perspective displays of patient contour data. Included is a solution to the problem of removing hidden line segments from the displayed image.