Search

Giant cell tumor of bone (GCTB) is a rare tumor of the bone that is locally invasive. Surgery is the primary treatment that is usually done by intralesional curettage. In pelvis and spine surgery may be associated with high rate of complications, recently, Denosumab has been proposed for the treatment of these tumors in latter anatomical regions. Denosumab may be administered alone or as an adjuvant to surgery. This study aimed to assess the treatment effects of Denosumab in patients with unresectable GCTB. This study was a case series. Patients with unresectable GCTB of vertebra and sacrum were enrolled in this study. Patients received 120 mg of monthly Denosumab and additional doses on days 8th and 15th of treatment. Images of patients before and after treatment were evaluated. Nine patients with a median age of 30 years with spine and sacrum GCTB were included in this study. The median time of treatment with denosumab was 28 months (range: 3–67). Tumor control was seen in all patients. According to Inverse Choi density/size (ICDS), criteria objective response (complete response and partial response) was seen in 8 patients, and one had stable disease. Based on CT scan images, in 4 patients (44.44%), less than 50% of the transverse diameter of the tumor became ossified, and in the other five patients (55.55%), more than 50% of the tumor's transverse diameter became ossified. The median tumor volume before treatment was 829 cm3, and after treatment was 504 cm3 which was significantly reduced (P = 0.005). No complication related to therapy was seen. Tumor response was seen in all patients, and tumor control according to ICDS criteria was evident in all cases. This finding was in line with previous studies. Clinical improvement of signs and symptoms was also seen in all patients. Generally, our study demonstrates a sustained clinical benefit and tumor response with Denosumab, as tumor response ≥ 24 weeks was evident in all cases. No side effects were seen in patients despite long-term treatment with Denosumab. [ABSTRACT FROM AUTHOR]

Background and objective: Pelvic bone tumors represent a harmful orthopedic condition, encompassing both benign and malignant forms. Addressing the issue of limited accuracy in current machine learning algorithms for bone tumor image segmentation, we have developed an enhanced bone tumor image segmentation algorithm. This algorithm is built upon an improved full convolutional neural network, incorporating both the fully convolutional neural network (FCNN-4s) and a conditional random field (CRF) to achieve more precise segmentation. Methodology: The enhanced fully convolutional neural network (FCNN-4s) was employed to conduct initial segmentation on preprocessed images. Following each convolutional layer, batch normalization layers were introduced to expedite network training convergence and enhance the accuracy of the trained model. Subsequently, a fully connected conditional random field (CRF) was integrated to fine-tune the segmentation results, refining the boundaries of pelvic bone tumors and achieving high-quality segmentation. Results: The experimental outcomes demonstrate a significant enhancement in segmentation accuracy and stability when compared to the conventional convolutional neural network bone tumor image segmentation algorithm. The algorithm achieves an average Dice coefficient of 93.31 %, indicating superior performance in real-time operations. Conclusion: In contrast to the conventional convolutional neural network segmentation algorithm, the algorithm presented in this paper boasts a more intricate structure, proficiently addressing issues of over-segmentation and under-segmentation in pelvic bone tumor segmentation. This segmentation model exhibits superior real-time performance, robust stability, and is capable of achieving heightened segmentation accuracy.

• Current machine learning algorithms for pelvic bone tumor image segmentation have limited accuracy. • Our proposed algorithm combines a fully convolutional neural network and a conditional random field to achieve more accurate segmentation of pelvic bone tumor images. • FCNN-4s is used to improve the precision and convergence speed of pelvic bone tumor segmentation. • FCNN-4s adopts operations like Crop and Fuse, padding, ReLU activation, and SoftMax loss with optimized hyperparameters for better performance. • Our algorithm demonstrated an improvement of 6.69% in terms of the Dice coefficient compared to other algorithms, with an average enhancement of 9.33% Pelvic bone tumors represent a harmful orthopedic condition, encompassing both benign and malignant forms. Addressing the issue of limited accuracy in current machine learning algorithms for bone tumor image segmentation, we have developed an enhanced bone tumor image segmentation algorithm. This algorithm is built upon an improved full convolutional neural network, incorporating both the fully convolutional neural network (FCNN-4s) and a conditional random field (CRF) to achieve more precise segmentation. The enhanced fully convolutional neural network (FCNN-4s) was employed to conduct initial segmentation on preprocessed images. Following each convolutional layer, batch normalization layers were introduced to expedite network training convergence and enhance the accuracy of the trained model. Subsequently, a fully connected conditional random field (CRF) was integrated to fine-tune the segmentation results, refining the boundaries of pelvic bone tumors and achieving high-quality segmentation. The experimental outcomes demonstrate a significant enhancement in segmentation accuracy and stability when compared to the conventional convolutional neural network bone tumor image segmentation algorithm. The algorithm achieves an average Dice coefficient of 93.31 %, indicating superior performance in real-time operations. In contrast to the conventional convolutional neural network segmentation algorithm, the algorithm presented in this paper boasts a more intricate structure, proficiently addressing issues of over-segmentation and under-segmentation in pelvic bone tumor segmentation. This segmentation model exhibits superior real-time performance, robust stability, and is capable of achieving heightened segmentation accuracy. [ABSTRACT FROM AUTHOR]