ABSTRACT
Objective:To build a predictive model for symptomatic radiation pneumonitis(RP) using the pretreatment CT radiomics features, clinical and dosimetric data of lung cancer patients by using machine learning method.Methods:A retrospective analysis of 103 lung cancer patients who underwent radiotherapy in the Affiliated Hospital of Jiangnan University from November 2018 to April 2020 was performed. Total normal lung tissues were segmented as an interested volume in pretreatment CT images, and then 250 radiomics features were extracted. The correlations of RP and clinical or dosimetric features were firstly investigated with univariate analysis. Then all clinical data, dosimetric data and CT radiomics features were collected and considered as predictors for modeling of RP grade ≥ 2. Features were selected through LASSO machine learning method, and the predictive model was built. Finally, nomogram for risk of RP were obtained according to the selected features.Results:The result of univariate analysis showed that symptomatic RP was significantly correlated with lung dosimetric parameters including mean lung dose (MLD), V20 Gy and V30 Gy( t=2.20, 2.34 and 2.93, P<0.05). Four features, including lung dose volume percentage V30 Gyand three radiomics features, entropy feature of GLCM, mean and median feature of wavelet histogram were selected among all clinical, dosimetric features and radiomics features. AUC of the predicted model obtained from selected features reached 0.757. For convenient clinical use, the nomogram were obtained, and then personalized RP risk prediction and early intervention could be performed according to this nomogram. Conclusions:Pretreatment CT radiomics and dosimetric features can be used in predicting symptomatic RP, which will be useful for advanced intervention treatment.
ABSTRACT
Objective To research the effect of cavity under Bolus to anisotropic analytical algorithm (AAA) on calculation precision of dose in shallow tissue based on Monte Carlo method;Methods A 30 cm × 30 cm × 30 cm water phantom with the upper surface was constructed which was located at the source-axis distance (SAD) of the medical linear accelerator and the center as well as coincided with the central axis of the radiation field in Eclipse treatment panning system. Above the water phantom, a water film of 1 cm thick with or without different cavities was constructed or. AAA was used to calculate the dose distribution on the central axis and the x-axis of different depth of the water model with different cavities respectively. The accelerator model, the same water phantom and the water film were constructed and the dose distributions of the same positions were calculated with Geant4. Based on the Geant4 calculation result, the calculation precision of AAA with different cavity were compared;Results For cavities with area of 2 cm × 2 cm, if the thickness is smaller than 0.5 cm, the AAA calculation error is about 2%. with the cavity thickness increase, the AAA would overestimate the dose in the shallow area under the cavity. With the cavity area increase, the area where AAA overestimate the shallow dose gradually moved out until near the edge of the radiation field, and the calculation error on the central area gradually reduced until there is basically no error. Conclusions The shallow dose would be increased according to the cavity size when planning with AAA; If there are cavities with large volume, it is better to reposition.
ABSTRACT
ABSTRACT OBJECTIVE：To systematically evaluate the efficacy and safety of echinococcins in the prevention of invasive candidiasis，and to provide evidence-based reference for clinical treatment. METHODS：Retrieved from PubMed，Embase， Medline，Cochrane library，CJFD，VIP and Wanfang database，randomized controlled trials（RCTs）about the effectiveness and safety of echinococcins （trial group） versus conventional fungal drugs （amphotericin B and triazole antifungal agents，control group）in the prevention of invasive candidiasis were collected during database establishment to Jul. 2019. After data extraction of clinical studies met inclusion criteria and quality evaluation with Cochrane system evaluator manual 5.0.2，Meta-analysis was performed for breakthrough invasive fungal infection rate，fungal infection mortality rate，all-cause mortality rate and the incidence of drug withdrawal due to adverse reactions by using Rev Man 5.2 statistical software. RESULTS：Totally 7 RCTs were included， involving 3 219 patients. Results of Meta-analyses indicated that the incidence of breakthrough invasive fungal infection [OR＝0.58， 95％CI（0.40，0.85），P＝0.004]，fungal infection mortality rate[OR＝0.68，95％CI（0.51，0.92），P＝0.01] and the incidence of drug withdrawal due to adverse reactions [OR＝0.52，95％CI（0.40，0.67），P＜0.001] in trial groups were significantly lower than control group，with statistical significance. There was no statistical significance in the all-cause mortality rate [OR＝0.84，95％CI（0.67， 1.05），P＝0.13]. CONCLUSIONS：Compared with amphotericin B and triazole antifungal agents，echinococcins used for the prevention of invasive candidiasis can reduce the incidence of breakthrough invasive fungal infection，fungal infection mortality rate and the incidence of drug withdrawal due to adverse reactions
ABSTRACT
Objective@#To analyze the magnetic resonance imaging (MRI) radiomic performance of hypoxic area in nasopharyngeal carcinoma patients, aiming to provide a reference for identification and analysis of hypoxic area.@*Methods@#The MRI-T1, MRI-T2, MRI-T1+ and PET/CT images of 32 patients initially diagnosed with nasopharyngeal carcinoma were retrospectively analyzed. The gross tumor volume (GTV) of nasopharynx was countoured and the hypoxic areas (GTV-H) were identified by 18F-FMISO-PET images. The non-hypoxic areas (GTV-NH) were defined as the rest of areas removed GTV-H from GTV. The radiomic features of GTV-H and GTV-NH were extracted and compared.@*Results@#The average volume of GTV-H and GTV-NH was (10.92±11.02) cm3 and (7.21±5.70) cm3, respectively. The maximum rate of change was 46% for intensity direct-global min (ID-GM) on MRI-T1(P<0.05, AUC>0.7 and Youden index>0.5). The average rate of change was 136% for long run emphasis (LRE), long run high gray level emphasis (LRHGLE) and long run low gray level emphasis (LRLGLE) on MRI-T2(P<0.05, AUC>0.7 and Youden index>0.5). The high change rates was greater than 90% on MRI-T1+ (P<0.05, AUC>0.7 and Youden index>0.5) for ID-GM, LRE, LRHGLE and LRLGLE.@*Conclusions@#The hypoxic area of tumor target can be reflected by MRI radiomics on T1/T2/T1+ . Quantifying and tracking the variations of these features can bring benefit to recognize the hypoxic area of nasopharyngeal carcinoma tumor target.
ABSTRACT
Objective To evaluate the effect of the thickness and area of the cavity between bolus and skin upon the dose deposition in the shallow tissues.Methods The linear accelerator head model of 6 MV X ray was constructed using Geant4,which simulated the accelerator working condition of 10 cm× 10 cm field and recorded the information of emergent particles as phase space file.A water phantom of 30 × 30 × 30 cm3 was designed at the SAD level.A 30 cm × 30 cm × 1 cm water film was constructed adjacent to or with different cavities to the upper surface of the phantom.The phase space file was utilized as particle source to simulate the particle transport process.The axis depth dose distribution and lateral dose profiles in the center area of the field at a depth of 1 mm,5 mm,9 mm,15 mm and 21 mm of the phantom were obtained,respectively.The simulated data obtained from water film with different cavities were compared with those of adjacent to the water phantom.Results When the cavity was ≤ 5 mm,the cavity exerted slight effect.When the thickness was increased,the maximum dose depth (Dmax) was increased,the PDD at Dmax (PD Dmax) became less,the depth of influenced lateral dose profiles was increased and the dose in the center area was decreased.Along with the increase of cavity area,the Dmax was initially increased and then decreased,the PD Dmax was at first decreased and subsequently increased,the depth of influenced lateral dose profile was initially increased and then decreased,the dose in the center area was first decreased and subsequently increased.The lateral dose profile distant from the cavity or located at a depth ≥ 15 mm was almost not affected by the cavity.Conclusion The thickness of the cavity between bolus and skin should be less than 5 mm and the cavity area should be small as possible.
ABSTRACT
Objective To investigate the effect of cavity thickness, area and distance under the bolus upon the dose in the superficial tissues. Methods An accelerator model was constructed based on Geant4.The model accuracy was validated by the comparison of the calculated data with the measured data. A 30×30×30 cm3 water phantom with the upper surface located at the isocenter level and a 30×30×1 cm3 water film were constructed. Different models with the water film close to or different cavities with the water phantom were established. Under the 10×10 cm2 field with 6 MV X-ray beam,the central axis depth dose distribution and the lateral dose profiles at a depth of 0. 1 cm ( profile1) of the models with different cavities were calculated. The calculated data of different model with the water film close to or different cavities with the water phantom were statistically compared. Results When the cavity thickness was ≤ 0. 5 cm, the cavity exerted slight effect upon the depth of maximum dose ( Dmax ) and superficial dose. As the cavity thickness was increased,the Dmax was also increased,the PDD at 0. 1 cm ( PDD1) was decreased rapidly and the profile1 was increased from the cavity center to the edge. Along with the increase of cavity area,the Dmax was initially increased and then decreased,whereas the PDD1 was first decreased followed by an increase. When the cavity area was small,the profile1 was gradually increased from the cavity center to the edge. When the cavity area was large,the profile1 was initially decreased and subsequently increased. When the distance was ≥0. 2 cm,it was qualified for the clinical requirement and it exerted no effect when the distance was≥1. 0 cm. The profile1 distant from the cavity was not affected. Conclusion The cavity under the bolus should be minimized to reduce the cavity thickness,area and distance as possible.
ABSTRACT
Objective To explore the identification of the hypoxia regions within the primary foci and imaging time selection in 18F-fluoromisonidazole (FMISO) imaging on patients with nasopharyngeal carcinoma (NPC).Methods From July 2013 to July 2014,44 NPC patients (33 males,11 females,age range:18-74(53.45± 12.88) years) underwent 18 F-FMISO PET/CT imaging,including 3 cases with twice imaging (totally 47 case times).Inaging data were acquired and reconstructed 2 and 4 h after the injection of 18 F-FMISO.1s F-FMISO PET/CT images were merged with MRI images obtained 1 week before to construct fusion images.The boundary of primary tumor was determined based on MRI.Visual analysis was performed and SUVmax of posterior cervical muscles,NPC primary foci was measured by 2 observers respectively.The uptake ratio of primary tumor to muscle (TMR) was calculated.The identify consistency of hypoxic region between two observers were evaluated by Kappa test and intraclass correlation coefficient (ICC).The image contrast was evaluated by Wilcoxon paired rank sum test of TMR.Results PET images and MRI images of NPC primary foci were successfully fused.Positive non-NPC tissues were identified by MRI.The visual recognition of hypoxic regions of the two observers for 2 and 4 h imaging were highly consistent (Kappa =0.931 and 0.965,both P＜0.001).There was a high degree of consistency between the SUVmax of posterior cervicalmuscles and that of primary tumors.ICCs of posterior cervical muscles in 2 and 4 h were 0.896 (95％ CI:0.814-0.942) and 0.924 (95％ CI:0.865-0.958),respectively.ICCs of primary tumors in 2 and 4 h were 0.991 (95％ CI:0.985-0.995) and 0.998 (95％ CI:0.996-0.999),respectively.TMRs (M(P25,P75))in 2 and 4 h were 1.560 (1.341,3.015) and 1.675 (1.387,3.001) respectively in 24 positive case times,and the difference was statistically significant (z=-2.557,P＜0.05).Conclusions Using fusion images of 18F-FMISO PET and MRI,hypoxic tissues within NPC primary foci can be accurately identified.There is a high degree of consistency within the visual and quantitative analysis of two observers.The image contrast of 18F-FMISO PET at 4 h is superior to that at 2 h.