was calculated for all the patients from the two groups. The patient acute hematologic toxicity (e.g. patient with symptoms of leukopenia, neutropenia, anemia, or thrombocytopenia) was recorded according to RTOG scoring scale. Independent sample t test were applied for the dose-volume parameters, and card square analysis for clinical parameters. Results: In terms of the OAR dose-volume parameters, the V 5 wV 35 and D mean for the pelvic bone, left femoral head and the right femoral head significantly (p<0.05) decreased for the PBMS-IMRT Group compared with the IMRT group. There was no significant difference in any dosevolume parameters for the bladder. As for PTV dose distribution, there was a significant (pZ0.001) improvement of the target conformity in the PBMS-IMRT group. In addition, the dose homogeneity was also improved by 14.5% for this group, although not significantly (PZ0.108). For the following acute hematologic toxicity observation, the numbers of the patients with grade 0, 1, 2, 3, 4 were 12, 6, 14, 10, and 0 respectively for the IMRT group. Meanwhile the corresponding numbers were 10, 19, 12, 1, and 0 for PBMS-IMRT group. The incidence of grade 2 or upper acute hematologic toxicity in the PBMS-IMRT group was significantly reduced (31%, 13/42 vs. 57.1%, 24/42, pZ0.027). There was no incidence of grade 2 or upper acute cystitis in both of the two groups. Conclusion: The marrow-sparing of the pelvic bone significantly reduces its dosage as well as the left and right femoral head. Patients with pelvic bone marrow-sparing had lower incidence of the grade 2 or upper acute hematologic toxicity compared with those without marrow-paring. Pelvic bone marrow-sparing can benefit patients with rectal cancer undergoing chemo-radiotherapy.

Colorectal cancer is one of the major fatal diseases in Taiwan, and the age of patients is decreasing yearly. As the early symptoms of malignant intestinal swelling are not obvious, most patients are not diagnosed with colorectal cancer until they are in a serious condition, and early diagnosis of colorectal cancer is a major challenge. In addition to genetic factors, colorectal cancer is caused by stress and poor dietary habits. To improve the effectiveness of treatment, we propose a model for multi-attribute decision-making based on medical data by combining the Delphi method, fuzzy logic theory, and the analytical hierarchy process. In the hierarchical structure, key risk factors and their values are proposed using fuzzy logic theory. The model suggests the importance of the medical team, medical equipment, post-operative care, and other factors for treating colorectal cancer effectively. Treatment and appropriate care must be integrated in the overall medical process to improve the effectiveness of colorectal cancer treatment.

of vertebral T12, and 75.7% of the LNs were seen above the border of 2 cm inferior to the left renal artery, which often near the bottom of vertebral L2. Horizontally, 80% of LNs in celiac level were located within the region confined by the anterior surface of aorta, inferior vena cava (IVC), portal vein, splenic vein and celiac branches. In the superior mesenteric artery (SMA) level, all the LNs except 1 post-aortic LNs were behind the anterior edge of the SMA and anterior to the posterior edge of aorta, 85% were left to the left edge of IVC, and 95% were within 1.5 cm of the left side of the SMA. In the region beneath SMA, 92.7% of LNs appeared right to the lateral edge of the left psoas major (LPM), and all the LNs lied behind the anterior surface of the left renal vein or adjacent to the anterior surface of the aorta, left to the right border of the IVC, and anteriorly to the vertebral and the anterior edge of the LPM. The proposed CTV from the middle level of vertebral T12 to the 2cm below the left renal artery or the bottom of vertebral L2 (depending on which one is lower) with aforementioned boundaries can cover 74.52% of the LNs in the studied group. The postoperative RT treatment plan which included the commonly used T-shield field and the proposed CTV for RetroLNs in three simulated cases all met with satisfactory in the dosimetric analysis. Conclusion: The proposed CTV is the first that focus on the high-risk area of RetroLNs failure in LTESCC patients treated after surgery, and is also feasible in terms of dosimetric performance. The result can serve as a reference for future study to better improve the disease control in LTESCC.

No abstract