Lumbar and Thoracic Vertebrae Segmentation in CT Scans Using a 3D Multi-Object Localization and Segmentation CNN.

Radiation treatment of cancers like prostate or cervix cancer requires considering nearby bone structures like vertebrae. In this work, we present and validate a novel automated method for the 3D segmentation of individual lumbar and thoracic vertebra in computed tomography (CT) scans. It is based on a single, low-complexity convolutional neural network (CNN) architecture which works well even if little application-specific training data are available. It is based on volume patch-based processing, enabling the handling of arbitrary scan sizes. For each patch, it performs segmentation and an estimation of up to three vertebrae center locations in one step, which enables utilizing an advanced post-processing scheme to achieve high segmentation accuracy, as required for clinical use. Overall, 1763 vertebrae were used for the performance assessment. On 26 CT scans acquired for standard radiation treatment planning, a Dice coefficient of 0.921 ± 0.047 (mean ± standard deviation) and a signed distance error of 0.271 ± 0.748 mm was achieved. On the large-sized publicly available VerSe2020 data set with 129 CT scans depicting lumbar and thoracic vertebrae, the overall Dice coefficient was 0.940 ± 0.065 and the signed distance error was 0.109 ± 0.301 mm. A comparison to other methods that have been validated on VerSe data showed that our approach achieved a better overall segmentation performance.

Initial clinical applications treating pediatric and adolescent patients using MR-guided radiotherapy.

Purpose: To demonstrate the clinical applications and feasibility of online adaptive magnetic resonance image guided radiotherapy (MRgRT) in the pediatric, adolescent and young adult (AYA) population. Methods: This is a retrospective case series of patients enrolled onto a prospective study. All pediatric (age < 18) and AYA patients (age< 30), treated on the Elekta Unity MR linear accelerator (MRL) from 2019 to 2021 were enrolled onto a prospective registry. Rationale for MRgRT included improved visualization of and alignment to the primary tumor, re-irradiation in a critical area, ability to use smaller margins, and need for daily adaptive replanning to minimize dose to adjacent critical structures. Step-and-shoot intensity-modulated radiation treatment (IMRT) plans were generated for all Unity patients with a dose grid of 3 mm and a statistical uncertainty of < 1% per plan. Results: A total of 15 pediatric and AYA patients have been treated with median age of 13 years (range: 6 mos - 27 yrs). Seven patients were <10 yo. The clinical applications of MRgRT included Wilms tumor with unresectable IVC thrombus (n=1), Ewing sarcoma (primary and metastatic, n=3), recurrent diffuse intrinsic pontine glioma (DIPG, n=2), nasopharyngeal carcinoma (n=1), clival chordoma (n=1), primitive neuroectodermal tumor of the pancreas (n=1), recurrent gluteo-sacral germ cell tumor (n=1), C-spine ependymoma (n=1), and posterior fossa ependymoma (n=1). Two children required general anesthesia. One AYA patient could not complete the MRgRT course due to tumor-related pain exacerbated by longer treatment times. Two AYA patients experienced anxiety related to treatment on the MRL, one of which required daily Ativan. No patient experienced treatment interruptions or unexpected toxicity. Conclusion: MRgRT was well-tolerated by pediatric and AYA patients. There was no increased use of anesthesia outside of our usual practice. Dosimetric advantages were seen for patients with tumors in critical locations such as adjacent to or involving optic structures, stomach, kidney, bowel, and heart.

Quantification of uptake in pelvis F-18 FLT PET-CT images using a 3D localization and segmentation CNN.

PURPOSE: The purpose of this work was to develop and validate a deep convolutional neural network (CNN) approach for the automated pelvis segmentation in computed tomography (CT) scans to enable the quantification of active pelvic bone marrow by means of Fluorothymidine F-18 (FLT) tracer uptake measurement in positron emission tomography (PET) scans. This quantification is a critical step in calculating bone marrow dose for radiopharmaceutical therapy clinical applications as well as external beam radiation doses. METHODS: An approach for the combined localization and segmentation of the pelvis in CT volumes of varying sizes, ranging from full-body to pelvis CT scans, was developed that utilizes a novel CNN architecture in combination with a random sampling strategy. The method was validated on 34 planning CT scans and 106 full-body FLT PET-CT scans using a cross-validation strategy. Specifically, two different training and CNN application options were studied, quantitatively assessed, and statistically compared. RESULTS: The proposed method was able to successfully locate and segment the pelvis in all test cases. On all data sets, an average Dice coefficient of 0.9396 ± $\pm$ 0.0182 or better was achieved. The relative tracer uptake measurement error ranged between 0.065% and 0.204%. The proposed approach is time-efficient and shows a reduction in runtime of up to 95% compared to a standard U-Net-based approach without a localization component. CONCLUSIONS: The proposed method enables the efficient calculation of FLT uptake in the pelvis. Thus, it represents a valuable tool to facilitate bone marrow preserving adaptive radiation therapy and radiopharmaceutical dose calculation. Furthermore, the method can be adapted to process other bone structures as well as organs.

Using [(18)F]Fluorothymidine Imaged With Positron Emission Tomography to Quantify and Reduce Hematologic Toxicity Due to Chemoradiation Therapy for Pelvic Cancer Patients.

PURPOSE: The purpose of the present prospective clinical trial was to determine the efficacy of [(18)F]fluorothymidine (FLT)-identified active bone marrow sparing for pelvic cancer patients by correlating the FLT uptake change during and after chemoradiation therapy with hematologic toxicity. METHODS AND MATERIALS: Simulation FLT positron emission tomography (PET) images were used to spare pelvic bone marrow using intensity modulated radiation therapy (IMRT BMS) for 32 patients with pelvic cancer. FLT PET scans taken during chemoradiation therapy after 1 and 2 weeks and 30 days and 1 year after completion of chemoradiation therapy were used to evaluate the acute and chronic dose response of pelvic bone marrow. Complete blood counts were recorded at each imaging point to correlate the FLT uptake change with systemic hematologic toxicity. RESULTS: IMRT BMS plans significantly reduced the dose to the pelvic regions identified with FLT uptake compared with control IMRT plans (P<.001, paired t test). Radiation doses of 4 Gy caused an ∼50% decrease in FLT uptake in the pelvic bone marrow after either 1 or 2 weeks of chemoradiation therapy. Additionally, subjects with more FLT-identified bone marrow exposed to ≥4 Gy after 1 week developed grade 2 leukopenia sooner than subjects with less marrow exposed to ≥4 Gy (P<.05, Cox regression analysis). Apparent bone marrow recovery at 30 days after therapy was not maintained 1 year after chemotherapy. The FLT uptake in the pelvic bone marrow regions that received >35 Gy was 18.8% ± 1.8% greater at 30 days after therapy than at 1 year after therapy. The white blood cell, platelet, lymphocyte, and neutrophil counts at 1 year after therapy were all lower than the pretherapy levels (P<.05, paired t test). CONCLUSIONS: IMRT BMS plans reduced the dose to FLT-identified pelvic bone marrow for pelvic cancer patients. However, reducing hematologic toxicity is challenging owing to the acute radiation sensitivity (∼4 Gy) and chronic suppression of activity in bone marrow receiving radiation doses >35 Gy, as measured by the FLT uptake change correlated with the complete blood cell counts.