CN102542153B - Method for introducing influence of biological radiation sensitivity parameters on normal tissue complication probability - Google Patents

Abstract

The invention discloses a method for introducing influence of biological radiation sensitivity parameters on normal tissue complication probability, which is characterized in that the method comprises the following steps: based on a simple normal tissue organ model, calculating to obtain the dose and BED (biologically effective dose) distribution of the organ model; performing box normalization on the BED distribution, and calculating to obtain a total SF (survival fraction); computing the SF to obtain an EUD (effective uniform dose) of the organ model; replacing a gEUD (generalized effective uniform dose) corresponding to 50% complications of an NTCPLKB (normal tissue complication probability lyman-kutcher-burman) model into the effective uniform dose EUD corresponding to nuclide 50% complications; and introducing the influence of the radiation sensitivity parameters on the normal tissue complication probability by using multiple radiation sensitivity parameters in the EUD model. By the method, the limitation that the conventional NTCP LKB model only reflects the influence of dose change on the damage of functions of a normal tissue organ, cannot reflect response difference of different biological individuals on the same dose distribution is overcome; and a valuable basis for radiation biological toxicity prediction can be provided for a radiation clinical worker.

Description

A kind of method of introducing influence of biological radiation sensitivity parameters on normal tissue complication probability

Technical field

The present invention relates to a kind of calculating normal structure complication probability (NTCP, Normal Tissue ComplicationProbability) model, this model can be used in nuclear medicine radioimmunotherapy, studies the impact of multiple radiation sensitivity parameters on normal tissue complication probability NTCP.

Background technology

Radiation effect on biological objects is carried out to radiobiology simulation, is another effective, the economic research approach outside traditional biological laboratory facilities that continues.Modern radiobiology simulation has had a set of complete theory, and follows Bioexperiment to find in upgrading constantly.Biological effective dose (BED, Biologically Effective Dose) has reflected the damage influence of radiation dose to biosome, and BED can be expressed as:

BED＝D×RE-RF (10)

In formula (10), RE is the relative efficiency factor (RE, Relative Effectiveness factor) of radiating particle, and RF is growth factor (RF, Regeneration Factor).The BED of i cell _ican be calculated as follows:

${\mathrm{BED}}_i=(\frac{\stackrel{\·}{D}(t=0)}{{\mathrm{\λ}}_{\mathrm{eff}}}-\frac{0.693}{{\mathrm{\α\λ}}_{\mathrm{eff}}{T}_{\mathrm{av}}})*(1+\frac{\stackrel{\·}{D}(t=0)}{(\mathrm{\μ}+{\mathrm{\λ}}_{\mathrm{eff}})(\mathrm{\α}/\mathrm{\β})})$

$-\left(\frac{0.693}{\mathrm{\α}{T}_{\mathrm{av}}}\right)(-\frac{1}{{\mathrm{\λ}}_{\mathrm{eff}}}\ln \left(\frac{0.693}{\stackrel{\·}{D}(t=0)\mathrm{\α}{T}_{\mathrm{av}}}\right))---\left(1\right)$

In formula (1), μ is that cell index is repaired constant, be the predose rate of nucleic, α, β are biological radiosusceptibility parameter, λ _effeffectively clean up constant, T _avfor average cell clone DT Doubling Time.

BED is carried out to branch mailbox processing, utilize its normalization distribution function P (ψ _j), just can obtain total existence mark SF (Survival Fraction) of whole biosome area-of-interest:

$\mathrm{SF}=\underset{j}{\overset{N}{\mathrm{\Σ}}}P\left({\mathrm{\ψ}}_j\right)e^{-\mathrm{\α}{\mathrm{\ψ}}_j}\mathrm{\Δ}{\mathrm{\ψ}}_j---\left(2\right)$

In formula (2), P (ψ _j) be BED _ithe normalization distribution function of the BED-volume histogram (BVH, BED VolumeHistogram) obtaining after returning case to process, ψ _jthe intermediate value of BED j case, Δ ψ _jit is the width value of j case.

Utilize total existence mark SF, by formula (3), just can calculate effective uniform dose EUD:

$\mathrm{EUD}=-\frac{1}{\mathrm{\α}}\ln \left(\mathrm{SF}\right)---\left(3\right)$

Visible according to modern radiobiology modeling theory, EUD has comprised abundant radiosusceptibility parameter, as α (unit: Gy ^-1), α/β (unit: Gy), nucleic decay half life period, biology clean up half life period and cell repair half life period etc.

In radiation oncotherapy, tumour normal structure and organ around inevitably can be subject to radiation damage more or less, can cause the forfeiture of normal structure organ dysfunction when serious.Existed at present several to evaluate the radiobiological models of the suffered degree of injury of normal structure organ, such as critical first prime model, parallel model and LKB (Lyman-Kutcher-Burman) model etc., wherein LKB model is used comparatively extensively at present in radiation biological simulation, and LKB model formation is:

$\mathrm{NTCP}\left(\mathrm{EUD}\right)=\frac{1}{\sqrt{2\mathrm{\π}}}\underset{-\∞}{\overset{t}{\∫}}\exp ({-u}^2/2)\mathrm{du}---\left(11\right)$

t＝(gEUD-D ₅₀)/(D ₅₀m)

In formula (11), D ₅₀be that organ dysfunction is lost 50% needed dosage, m is the inverse of LKB response curve maximum slope, and u is integration variable.

The meaning of LKB model description is: only have and reach certain limit (as 50%) when the suffered average function damage accumulation of organ, the complication of organ just there will be.This model based on the effective uniform dose (gEUD of broad sense, Generalized meanEquivalent Uniform Dose) concept, for parallel organization organ (as lung, liver, brain, kidney etc.), also be in radionuclide therapy, conventionally to need the organ that jeopardizes of consideration, gEUD is exactly the suffered mean dose of organ.So in this definition of LKB model, except dose characteristics value, do not consider the impact of many radiation biological mathematic(al) parameters.

Summary of the invention

The present invention is for avoiding the existing deficiency of above-mentioned prior art, and a kind of method of introducing influence of biological radiation sensitivity parameters on normal tissue complication probability is provided.Realization, in radionuclide therapy, is studied the impact of multiple radiosusceptibility parameter on NTCP output, for nuclear medicine clinical position person provides valuable radiobiology prediction.

The present invention is that technical solution problem adopts following technical scheme:

The feature that the present invention introduces the method for influence of biological radiation sensitivity parameters on normal tissue complication probability is to carry out as follows:

Step 1, set up the naive model of normal structure organ

Set up the normal tissue cell group model A1 of a square, described normal tissue cell group model A1 includes a spherical organ model A2, with described spherical organ model A2, represent a human organ being surrounded by normal tissue cell group, the normal tissue cell group model A1 of whole square is consisted of the identical sphaerocyst of size, and the material of cell is even aqueous medium;

Step 2, utilize the cell S factor and convolution method to realize dose distributions computation

Set up two sphaerocyst Model B 1 and B2 that size is identical, utilize Monte Carlo algorithm simulation to obtain the cell S factor list under described two sphaerocyst Model B 1 centre distance situations of change different from B2, according to described cell S factor list correspondence, obtain the cell S factor distributed data of normal structure organ model A1, utilize three-dimensional fast Fourier convolution algorithm 3-DFFT to realize to take in spherical organ model A2 the dose distributions computation that cell is unit;

Step 3, calculating biological effective dose BED and the mark SF that always survives

Adopt clinical literature data, calculate in spherical organ model A2, the biological effective dose BED that the cell of take is unit distributes, and BED is distributed and returns case to process, and calculates total existence mark SF of spherical organ model A2;

The BED of i cell in described spherical organ model A2 _iby formula (1), calculate:

${\mathrm{BED}}_i=(\frac{\stackrel{\·}{D}(t=0)}{{\mathrm{\λ}}_{\mathrm{eff}}}-\frac{0.693}{{\mathrm{\α\λ}}_{\mathrm{eff}}{T}_{\mathrm{av}}})*(1+\frac{\stackrel{\·}{D}(t=0)}{(\mathrm{\μ}+{\mathrm{\λ}}_{\mathrm{eff}})(\mathrm{\α}/\mathrm{\β})})$

$-\left(\frac{0.693}{\mathrm{\α}{T}_{\mathrm{av}}}\right)(-\frac{1}{{\mathrm{\λ}}_{\mathrm{eff}}}\ln \left(\frac{0.693}{\stackrel{\·}{D}(t=0)\mathrm{\α}{T}_{\mathrm{av}}}\right))---\left(1\right)$

In formula (1), μ is that cell index is repaired constant, the predose rate of nucleic, α and the β object radiation susceptibility parameter of making a living, λ _efffor effectively cleaning up constant, T _avfor average cell clone DT Doubling Time;

Total existence mark SF calculates by formula (2):

$\mathrm{SF}=\underset{j}{\overset{N}{\mathrm{\Σ}}}P\left({\mathrm{\ψ}}_j\right)e^{-\mathrm{\α}{\mathrm{\ψ}}_j}\mathrm{\Δ}{\mathrm{\ψ}}_j---\left(2\right)$

In formula (2), P (ψ _j) be BED _ithe normalization distribution function of the BED-volume histogram obtaining after returning case to process, ψ _jthe intermediate value of BED j case, Δ ψ _jit is the width value of j case;

Step 4, calculate effective uniform dose EUD

Utilize the resulting total existence mark SF of step (3), by formula (3), calculate effective uniform dose EUD of spherical organ model A2:

$\mathrm{EUD}=-\frac{1}{\mathrm{\α}}\ln \left(\mathrm{SF}\right)---\left(3\right)$

In formula (3), the α object radiation susceptibility parameter of making a living, SF is total existence mark of all cells in spherical organ model A2;

Step 5, determine the effective uniform dose EUD of 50% complication ₅₀

According to EUD with mean dose variation relation curve if it is D that the dosage of 50% complication occurs ₅₀, by described D ₅₀described in contrast variation relation curve, determines the corresponding effective uniform dose EUD of 50% complication ₅₀;

Step 6, calculating normal structure complication probability NTCP

The LKB EUD model of normal structure complication probability NTCP is set suc as formula (4)

$\mathrm{NTCP}\left(\mathrm{EUD}\right)=\frac{1}{\sqrt{2\mathrm{\π}}}\underset{-\∞}{\overset{t}{\∫}}\exp ({-u}^2/2)\mathrm{du}---\left(4\right)$

In formula (4): t=(EUD-EUD ₅₀)/(EUD ₅₀m); EUD ₅₀it is the effective uniform dose of 50% complication; M is the inverse of curve maximum slope, and m is clinical experience parameter; U is integration variable;

The EUD that the EUD that step (4) is obtained and step (5) obtain ₅₀substitution formula (4) calculates NTCP;

Step 7, change respectively the value of each radiosusceptibility parameter, by formula (4), calculate corresponding NTCP;

Described radiosusceptibility parameter comprises: cell index is repaired constant μ, and biosome radiosusceptibility parameter alpha and β, effectively clean up constant λ _effwith average cell clone DT Doubling Time T _av.

The feature that the present invention introduces the method for influence of biological radiation sensitivity parameters on normal tissue complication probability is also:

The method of utilizing Monte Carlo algorithm simulation to obtain the cell S factor list in two sphaerocyst Model B 1 centre distance situations different from B2 in described step 2 is: the sphaerocyst model of setting up two same radius, these two sphaerocyst models infiltrate in even aqueous medium, using B1 as source cell, B2 is as target cell, and nucleic activity is evenly distributed in described source cell B1; Described source cell B1 center position is constant, the center position of described target cell B2 radially changes according to the step-length of 1 μ m, utilize Monte Carlo algorithm simulation to obtain not isocenter and, apart from the absorbed dose in target cell B2 in situation, obtain the list of the cell S factor; The described cell S factor refers to medical internal radiation dose MIRD regulation: in source cell B1 during a nuclear decay of radioactive nuclide generation per second, and the mean dose absorbing in target cell B2; The unit of the S factor is GyBq ^-1s ^-1.

In described step 2, utilize three-dimensional fast Fourier convolution algorithm 3-D FFT, realize and in spherical organ model A2, using the method that dosage that cell is unit distributes and be: the convolution kernel by the cell S factor distributed data of the normal structure organ model A1 obtaining in step 2 as 3-D FFT, by usining the nucleic activity that cell is unit, distribute as response function, describedly take nucleic activity that cell is unit and distribute and refer to: in whole normal structure organ model A1, the relative activity of each cell distributes, the convolution of described convolution kernel and response function utilizes the three-dimensional Fast Fourier Transform (FFT) 3-D fft algorithm of Matlab to calculate realization.

Compared with the prior art, beneficial effect of the present invention is embodied in:

BED has described under unlimited LDR radiation event, the biological effect that same dose produces, and it more highlights the impact of dose rate on tissue response; What EUD described is effective uniform dose of area-of-interest, and what its reflected is the impact that integral dose distributes on tissue response.BED is its biological effective dose of angle calculation from cell or volume elements individuality, if reflect total existence effect of whole organ or tissue, conventionally adopts the simple average value of BED, and EUD expresses effective uniform dose of whole tissue or organ by a value.Therefore in essence, be consistent aspect the homogeneity effect of EUD and gEUD non-uniform dose distribution in describing parallel organization organ (as lung, liver, brain, kidney etc.).Comparatively speaking, for parallel organization organ, be also in radionuclide therapy, conventionally need to consider jeopardize organ, EUD ratio meaning closer to gEUD.The present invention implants EUD parameter in the NTCP LKB model generally adopting at present, to introduce the impact of radiosusceptibility parameter on NTCP, overcome the damage influence that existing NTCP LKB model only reflects that dosage changes normal tissue organ dysfunction and can not reflect the limitation of different biology individual to same dose distribution response difference.Model of the present invention can be radiation clinical position person valuable radiobiology toxicity prediction foundation is provided.

Accompanying drawing explanation

Fig. 1 is a simple group model A1 of human normal tissue cell who sets up in the inventive method, wherein includes a spherical organ model A2, and spherical organ model A2 represents a human organ being surrounded by normal tissue cell group;

Fig. 2 is that effective uniform dose EUD in a simple human normal tissue cell group model of setting up in the inventive method is with mean dose variation relation;

Fig. 3 is cell proliferation in a simple human normal tissue cell group model of setting up in the inventive method impact on various nucleic NTCP curves, wherein EUD ₅₀get 34Gy, m gets 0.1;

Fig. 4 is that the various radiosusceptibility parameters in the simple human normal tissue cell group model that in the present invention, method is set up change the impact on NTCP curve.

Embodiment

Fig. 1 has shown that a simple group model A1 of human normal tissue cell wherein comprises a spherical organ model A2, and spherical organ model A2 represents a human organ being surrounded by normal tissue cell group; Fig. 2 has provided in spherical organ model A2 effectively uniform dose EUD with mean dose variation relation; Fig. 3 has provided the impact of cell proliferation on various nucleic NTCP curves, wherein EUD in spherical organ model A2 ₅₀get 34Gy, m gets 0.1.The feature that the present embodiment is introduced the method for influence of biological radiation sensitivity parameters on normal tissue complication probability is to carry out as follows:

Step 1, set up the naive model of normal structure organ

As shown in Figure 1, set up the normal tissue cell group model A1 of a square, normal tissue cell group model A1 includes a spherical organ model A2, with spherical organ model A2, represent a human organ being surrounded by normal tissue cell group, the normal tissue cell group model A1 of whole square is consisted of the identical sphaerocyst of size, and the material of cell is even aqueous medium;

Step 2, utilize the cell S factor and convolution method to realize dose distributions computation

Set up two sphaerocyst Model B 1 and B2 that size is identical, utilize Monte Carlo algorithm simulation to obtain two cell S factor lists under sphaerocyst Model B 1 centre distance situations of change different from B2, according to cell S factor list correspondence, obtain the cell S factor distributed data of normal structure organ model A1, utilize three-dimensional fast Fourier convolution algorithm 3-D FFT to realize to take in spherical organ model A2 the dose distributions computation that cell is unit;

In concrete enforcement, the method of utilizing Monte Carlo algorithm simulation to obtain the cell S factor list in two sphaerocyst Model B 1 centre distance situations different from B2 is: the sphaerocyst model of setting up two same radius, these two sphaerocyst models infiltrate in even aqueous medium, using B1 as source cell, B2 is as target cell, and nucleic activity is evenly distributed in source cell B1; Source cell B1 center position is constant, and the center position of target cell B2 radially changes according to the step-length of 1 μ m, utilizes Monte Carlo algorithm simulation to obtain not isocenter and, apart from the absorbed dose in target cell B2 in situation, obtains the list of the cell S factor; The cell S factor refers to medical internal radiation dose MIRD (Medical Internal Radiation Dose) regulation: in source cell B1 during a nuclear decay of radioactive nuclide generation per second, and the mean dose absorbing in target cell B2; The unit of the S factor is GyBq ^-1s ^-1.

In concrete enforcement, utilize three-dimensional fast Fourier convolution algorithm 3-D FFT, realize and in spherical organ model A2, using the method that dosage that cell is unit distributes and be: the convolution kernel by the cell S factor distributed data of the normal structure organ model A1 obtaining in step 2 as 3-D FFT, by usining the nucleic activity that cell is unit, distribute as response function, the nucleic activity that the cell of take is unit distributes and refers to: in whole normal structure organ model A1, the relative activity of each cell distributes, the convolution of convolution kernel and response function can utilize the three-dimensional Fast Fourier Transform (FFT) 3-D fft algorithm of Matlab to calculate realization.

Step 3, calculating biological effective dose BED and the mark SF that always survives

Adopt clinical literature data, calculate in spherical organ model A2, the biological effective dose BED that the cell of take is unit distributes, and BED is distributed and returns case to process, and calculates total existence mark SF of spherical organ model A2;

The BED of i cell in spherical organ model A2 _iby formula (1), calculate:

${\mathrm{BED}}_i=(\frac{\stackrel{\·}{D}(t=0)}{{\mathrm{\λ}}_{\mathrm{eff}}}-\frac{0.693}{{\mathrm{\α\λ}}_{\mathrm{eff}}{T}_{\mathrm{av}}})*(1+\frac{\stackrel{\·}{D}(t=0)}{(\mathrm{\μ}+{\mathrm{\λ}}_{\mathrm{eff}})(\mathrm{\α}/\mathrm{\β})})$

$-\left(\frac{0.693}{\mathrm{\α}{T}_{\mathrm{av}}}\right)(-\frac{1}{{\mathrm{\λ}}_{\mathrm{eff}}}\ln \left(\frac{0.693}{\stackrel{\·}{D}(t=0)\mathrm{\α}{T}_{\mathrm{av}}}\right))---\left(1\right)$

In formula (1), μ is that cell index is repaired constant, the predose rate of nucleic, α and the β object radiation susceptibility parameter of making a living, λ _efffor effectively cleaning up constant, T _avfor average cell clone DT Doubling Time;

Total existence mark SF calculates by formula (2):

$\mathrm{SF}=\underset{j}{\overset{N}{\mathrm{\Σ}}}P\left({\mathrm{\ψ}}_j\right)e^{-\mathrm{\α}{\mathrm{\ψ}}_j}\mathrm{\Δ}{\mathrm{\ψ}}_j---\left(2\right)$

In formula (2), P (ψ _j) be BED _ithe normalization distribution function of the BED-volume histogram obtaining after returning case to process, ψ _jthe intermediate value of BED j case, Δ ψ _jit is the width value of j case;

Step 4, calculate effective uniform dose EUD

Utilize the resulting total existence mark SF of step (3), by formula (3), calculate effective uniform dose EUD of spherical organ model A2:

$\mathrm{EUD}=-\frac{1}{\mathrm{\α}}\ln \left(\mathrm{SF}\right)---\left(3\right)$

In formula (3), the α object radiation susceptibility parameter of making a living, SF is total existence mark of all cells in spherical organ model A2;

From formula (1) and formula (3), what BED emphasized is the impact of dose rate on tissue response, is from individual its biological effectivenesses of angle calculation such as cells; And EUD reflection is the impact that integral dose distributes on tissue response, EUD has expressed the biological effectiveness of whole tissue or organ by a value.

Step 5, determine the effective uniform dose EUD of 50% complication ₅₀

According to EUD with mean dose variation relation curve if it is D that the dosage of 50% complication occurs ₅₀, by D ₅₀contrast variation relation curve, determines the corresponding effective uniform dose EUD of 50% complication ₅₀;

Due to the physical half time of each nucleic, the differences such as power spectrum (seeing Fig. 2) of decay β particle, its relation curve is different, therefore corresponding same D ₅₀eUD ₅₀be discrepant (in Table 1a), even if this explanation normal structure organ has absorbed same mean dose, different nucleic are different to its degree of injury.In studied nucleic, the EUD of Y-90 ₅₀value minimum (34Gy), the EUD of Cu-67 ₅₀value maximum (44Gy).Can infer that thus the toxicity threshold of Y-90 is minimum in studied nucleic, toxicity is the strongest, and this is relevant with longer physical half time (64.1 hours) with the high energy β particle of its decay.

Step 6, calculating normal structure complication probability NTCP

The LKB EUD model of normal structure complication probability NTCP is set suc as formula (4)

$\mathrm{NTCP}\left(\mathrm{EUD}\right)=\frac{1}{\sqrt{2\mathrm{\π}}}\underset{-\∞}{\overset{t}{\∫}}\exp ({-u}^2/2)\mathrm{du}---\left(4\right)$

In formula (4): t=(EUD-EUD ₅₀)/(EUD ₅₀m); EUD ₅₀it is the effective uniform dose of 50% complication; M is the inverse of curve maximum slope, and m is clinical experience parameter; U is integration variable;

The EUD that the EUD that step (4) is obtained and step (5) obtain ₅₀substitution formula (4) calculates NTCP;

Step 7, change respectively the value of each radiosusceptibility parameter, by formula (4), calculate corresponding NTCP;

Radiosusceptibility parameter comprises: cell index is repaired constant μ, and biosome radiosusceptibility parameter alpha and β, effectively clean up constant λ _effwith average cell clone DT Doubling Time T _av.

For example:

Consider the restriction of personal computer internal memory, general normal tissue cell group model A1 gets 280 * 280 * 280, represents that A1 contains 2 * 10 ⁷individual cell, the radius of establishing cell is 50 μ m, the size of the whole model of A1 is 2.8cm * 2.8cm * 2.8cm, this size meet the nucleic of studying (I-131, Cu-67, Re-186, Re-188 and Y-90) space of the β particle that discharges of decay transports equilibrium condition.

Take holonephros irradiation as example, and the people such as Emami sum up clinical data suggestion for 1991: holonephros irradiates the dosage threshold value D of 5 years 50% complication ₅₀for 28Gy.According to clinical literature data, choose biosimulation parameter (in Table 1a, table 1b), calculate (seeing Fig. 1) different mean doses in corresponding spherical organ model A2 time EUD value, obtain relation curve between the two (seeing Fig. 2).According to D ₅₀=28Gy obtains the EUD that multiple nucleic is corresponding ₅₀(in Table 1a).

The data area that radiosusceptibility parameter provides according to table 1b changes, the impact for research parameters on NTCP.For example Fig. 3 has provided cell proliferation to the impact of various nucleic NTCP curves (EUD wherein ₅₀got in above-mentioned studied nucleic EUD ₅₀corresponding minimum value 34Gy (being Y-90's), the reliability of the nucleic radiation safety that possessed some special knowledge to guarantee).For dosage, change slowly as seen, its cultivation effect is obvious.And for dosage ratio of transformation more sharply, the gradient ratio that exposure dose increases is larger, cultivation effect just almost can be ignored.Fig. 4 has provided various radiosensitive parameters and has changed the impact on NTCP curve, as (note: 28Gy) in the situation that mean dose is certain, Fig. 4 a shows that NTCP curve reduces along with the increase of α/β, so the dosage of 50% such complication probability transfinites for the little organ of α/β, arouse attention especially.Fig. 4 b is presented in the certain situation of mean dose, and NTCP curve reduces along with the increase of α, so for the organ that has little α, the dosage line of its 50% complication probability also should decrease.And if considered cell proliferation (shown in Fig. 4 c), the organ of left and right, α=0.2 should cause concern especially.For long organ of cell repair half life period, also corresponding increase of its risk factor (shown in Fig. 4 d).

Table 1a studies nucleic parameter

Note: ^*biology is cleaned up half life period T _cget 64 hours; Effectively clean up constant λ _eff=ln (2) * (T _p+ T _c)/(T _p* T _c)

Table 1b simulates parameter used and scope (take holonephros radiation parameters as example)

Note: ^*cell index is repaired constant μ=ln (2.)/T _μ

Claims (3)

1. introduce a method for influence of biological radiation sensitivity parameters on normal tissue complication probability, it is characterized in that carrying out as follows:

Step 1, set up the naive model of normal structure organ

Set up the normal tissue cell group model A1 of a square, described normal tissue cell group model A1 includes a spherical organ model A2, with described spherical organ model A2, represent a human organ being surrounded by normal tissue cell group, the normal tissue cell group model A1 of whole square is consisted of the identical sphaerocyst of size, and the material of cell is even aqueous medium;

Step 2, utilize the cell S factor and convolution method to realize dose distributions computation

Set up two sphaerocyst Model B 1 and B2 that size is identical, utilize Monte Carlo algorithm simulation to obtain the cell S factor list under described two sphaerocyst Model B 1 centre distance situations of change different from B2, according to described cell S factor list correspondence, obtain the cell S factor distributed data of normal tissue cell group model A1, utilize three-dimensional fast Fourier convolution algorithm 3-DFFT to realize to take in spherical organ model A2 the dose distributions computation that cell is unit;

Step 3, calculating biological effective dose BED and the mark SF that always survives

Adopt clinical literature data, calculate in spherical organ model A2, the biological effective dose BED that the cell of take is unit distributes, and BED is distributed and returns case to process, and calculates total existence mark SF of spherical organ model A2;

The BED of i cell in described spherical organ model A2 _iby formula (1), calculate:

$\begin{array}{c}{\mathrm{BED}}_i=(\frac{\stackrel{\•}{D}(t=0)}{{\mathrm{\λ}}_{\mathrm{eff}}}-\frac{0.693}{{\mathrm{\α\λ}}_{\mathrm{eff}}{T}_{\mathrm{av}}})*(1+\frac{\stackrel{\•}{D}(t=0)}{(\mathrm{\μ}+{\mathrm{\λ}}_{\mathrm{eff}})(\mathrm{\α}/\mathrm{\β})})\\ -\left(\frac{0.693}{\mathrm{\α}{T}_{\mathrm{av}}}\right)(-\frac{1}{{\mathrm{\λ}}_{\mathrm{eff}}}\mathrm{In}\left(\frac{0.693}{\stackrel{\•}{D}(t=0)\mathrm{\α}{T}_{\mathrm{av}}}\right))\end{array}----\left(1\right)$

In formula (1), μ is that cell index is repaired constant, the predose rate of nucleic, α and the β object radiation susceptibility parameter of making a living, λ _efffor effectively cleaning up constant, T _avfor average cell clone DT Doubling Time;

Total existence mark SF calculates by formula (2):

$\mathrm{SF}=\underset{j}{\overset{N}{\mathrm{\Σ}}}P\left({\mathrm{\ψ}}_j\right)e^{-\mathrm{\α}{\mathrm{\ψ}}_j}\mathrm{\Δ}{\mathrm{\ψ}}_j---\left(2\right)$

In formula (2), P (ψ _j) be BED _ithe normalization distribution function of the BED-volume histogram obtaining after returning case to process, ψ _jthe intermediate value of BED j case, Δ ψ _jit is the width value of j case;

Step 4, calculate effective uniform dose EUD

Utilize the resulting total existence mark SF of step 3, by formula (3), calculate effective uniform dose EUD of spherical organ model A2:

$\mathrm{EUD}=-\frac{1}{\mathrm{\α}}\mathrm{In}\left(\mathrm{SF}\right)---\left(3\right)$

In formula (3), the α object radiation susceptibility parameter of making a living, SF is total existence mark of all cells in spherical organ model A2;

Step 5, determine the effective uniform dose EUD of 50% complication ₅₀

If it is D that the dosage of 50% complication occurs ₅₀, by described D ₅₀eUD is with mean dose in contrast variation relation curve determine the corresponding effective uniform dose EUD of 50% complication ₅₀;

Step 6, calculating normal structure complication probability NTCP

The LKB EUD model of normal structure complication probability NTCP is set suc as formula (4)

$\mathrm{NTCP}\left(\mathrm{EUD}\right)=\frac{1}{\sqrt{2\mathrm{\π}}}\underset{-\∞}{\overset{t}{\∫}}\exp (-u^2/2)\mathrm{du}---\left(4\right)$

In formula (4): t=(EUD-EUD ₅₀)/(EUD ₅₀m); EUD ₅₀it is the effective uniform dose of 50% complication; M is the inverse of NTCP curve maximum slope, and m is clinical experience parameter; U is integration variable;

The EUD that the EUD that step 4 is obtained and step 5 obtain ₅₀substitution formula (4) calculates NTCP;

Step 7, change respectively the value of each radiosusceptibility parameter, by formula (4), calculate corresponding NTCP;

Described radiosusceptibility parameter comprises: cell index is repaired constant μ, and biosome radiosusceptibility parameter alpha and β, effectively clean up constant λ _effwith average cell clone DT Doubling Time T _av.

2. the method for introducing influence of biological radiation sensitivity parameters on normal tissue complication probability according to claim 1, it is characterized in that the method for utilizing Monte Carlo algorithm simulation to obtain the cell S factor list in two sphaerocyst Model B 1 centre distance situations different from B2 in described step 2 is: the sphaerocyst model of setting up two same radius, these two sphaerocyst models infiltrate in even aqueous medium, using B1 as source cell, B2 is as target cell, and nucleic activity is evenly distributed in described source cell B1; Described source cell B1 center position is constant, the center position of described target cell B2 radially changes according to the step-length of 1 μ m, utilize Monte Carlo algorithm simulation to obtain not isocenter and, apart from the absorbed dose in target cell B2 in situation, obtain the list of the cell S factor; The described cell S factor refers to be stipulated in medical internal radiation dose MIRD: in source cell B1 during a radioactive nuclide generation per second nuclear decay, and the mean dose absorbing in target cell B2; The unit of the S factor is G _yb _q ^-1s ^-1.

3. the method for introducing influence of biological radiation sensitivity parameters on normal tissue complication probability according to claim 1, it is characterized in that utilizing in described step 2 three-dimensional fast Fourier convolution algorithm 3-D FFT, realize and in spherical organ model A2, using the method for the dose distributions computation that cell is unit and be: the convolution kernel by the cell S factor distributed data of the normal structure organ model A1 obtaining in step 2 as 3-D FFT, by usining the nucleic activity that cell is unit, distribute as response function, describedly take nucleic activity that cell is unit and distribute and refer to: in whole normal structure organ model A1, the relative activity of each cell distributes, the convolution of described convolution kernel and response function utilizes the three-dimensional Fast Fourier Transform (FFT) 3-D fft algorithm of Matlab to calculate realization.