A "Sequential"" optimization technique for forward treatment planning for IMRT

Radiation therapy treatment planning for intensity modulated dose delivery with user-defined beam angles and sub-fields (sometimes called forward-planning IMRT) is an alternative to inverse planning. This method can be used for cases where critical structures lie within cavities in the target, and has the advantage of building on the techniques conventionally used for conformal treatment. A number of different optimization approaches have been described in the literature, and each has advantages and disadvantages. Here, the authors apply the simple and fast linear programming method and, if necessary, follow it with the more rigorous Cimmino iterative algorithm to obtain optimized weights for the fields/sub-fields. The linear programming algorithm can produce a set of beam weights that allow the total dose for each point in the tumor or critical organ to meet its corresponding limit, while maximizing (minimizing) the dose to the tumor (organ). In this case, the problem is termed "feasible," and no further optimization is attempted. If a feasible solution is not found, the dose limits can be relaxed. However, it is not always easy to know how to manipulate the dose limits. In such situations, the Cimmino iterative algorithm is brought into play to find the best set of beam weights (weighted least squares solution to the linear feasibility problem) given the dose limits and importance weightings assigned to each organ. The authors have tested the technique on a number of patient cases and achieved substantial improvement over previously used techniques.