Search

In this work, we aim to predict the survival time (ST) of glioblastoma (GBM) patients undergoing different treatments based on preoperative magnetic resonance (MR) scans. The personalized and precise treatment planning can be achieved by comparing the ST of different treatments. It is well established that both the current status of the patient (as represented by the MR scans) and the choice of treatment are the cause of ST. While previous related MR-based glioblastoma ST studies have focused only on the direct mapping of MR scans to ST, they have not included the underlying causal relationship between treatments and ST. To address this limitation, we propose a treatment-conditioned regression model for glioblastoma ST that incorporates treatment information in addition to MR scans. Our approach allows us to effectively utilize the data from all of the treatments in a unified manner, rather than having to train separate models for each of the treatments. Furthermore, treatment can be effectively injected into each convolutional layer through the adaptive instance normalization we employ. We evaluate our framework on the BraTS20 ST prediction task. Three treatment options are considered: Gross Total Resection (GTR), Subtotal Resection (STR), and no resection. The evaluation results demonstrate the effectiveness of injecting the treatment for estimating GBM survival., Comment: SPIE Medical Imaging 2024: Computer-Aided Diagnosis

Deep learning (DL) models for segmenting various anatomical structures have achieved great success via a static DL model that is trained in a single source domain. Yet, the static DL model is likely to perform poorly in a continually evolving environment, requiring appropriate model updates. In an incremental learning setting, we would expect that well-trained static models are updated, following continually evolving target domain data -- e.g., additional lesions or structures of interest -- collected from different sites, without catastrophic forgetting. This, however, poses challenges, due to distribution shifts, additional structures not seen during the initial model training, and the absence of training data in a source domain. To address these challenges, in this work, we seek to progressively evolve an ``off-the-shelf" trained segmentation model to diverse datasets with additional anatomical categories in a unified manner. Specifically, we first propose a divergence-aware dual-flow module with balanced rigidity and plasticity branches to decouple old and new tasks, which is guided by continuous batch renormalization. Then, a complementary pseudo-label training scheme with self-entropy regularized momentum MixUp decay is developed for adaptive network optimization. We evaluated our framework on a brain tumor segmentation task with continually changing target domains -- i.e., new MRI scanners/modalities with incremental structures. Our framework was able to well retain the discriminability of previously learned structures, hence enabling the realistic life-long segmentation model extension along with the widespread accumulation of big medical data., Comment: Early Accept to MICCAI 2023

Proton therapy systems without a gantry can be more compact and less expensive in terms of capital cost, and therefore more available to a larger patient population. Would the advances in pencil beam scanning and robotics make gantry-less treatment possible? In this study, we explore if high-quality treatment plans can be obtained without a gantry. We recently showed that proton treatments with the patient in an upright position may be feasible with a new soft robotic immobilization device and imaging which enables multiple possible patient orientations during a treatment. In this study, we evaluate if this new treatment geometry could enable high quality treatment plans without a gantry. We created pencil beam scanning (PBS) treatment plans for seven patients with head-and-neck or brain tumors. Each patient was planned with two scenarios: one with a gantry with the patient in supine position and the other with a gantry-less fixed horizontal beam-line with the patient sitting upright. For the treatment plans, dose-volume-histograms (DVHs), target homogeneity index (HI), mean dose, D_2 and D_98 are reported. A robustness analysis of one plan was performed with +/-2.5 mm setup errors and +/-3.5% range uncertainties with nine scenarios. Most of the PBS-gantry-less plans had similar target HI and OAR mean dose as compared to PBS-gantry plans, and similar robustness with respect to range uncertainties and setup errors. Pencil beam scanning provides sufficient power to deliver high quality treatment plans without requiring a gantry for head-and-neck or brain tumors. In combination with the development of the new positioning and immobilization methods required to support this treatment geometry, this work suggests the feasibility of further development of a compact proton therapy system with a fixed horizontal beam-line to treat patients in sitting and reclined positions.

Definition of the clinical target volume (CTV) is one of the weakest links in the radiation therapy chain. In particular, inability to account for uncertainties is a severe limitation in the traditional CTV delineation approach. Here, we introduce and test a new concept for tumor target definition, the clinical target distribution (CTD). The CTD is a continuous distribution of the probability of voxels to be tumorous. We describe an approach to incorporate the CTD in treatment plan optimization algorithms, and implement it in a commercial treatment planning system. We test the approach in two synthetic and two clinical cases, a sarcoma and a glioblastoma case. The CTD is straightforward to implement in treatment planning and comes with several advantages. It allows one to find the most suitable tradeoff between target coverage and sparing of surrounding healthy organs at the treatment planning stage, without having to modify or redraw a CTV. Owing to the variable probabilities afforded by the CTD, a more flexible and more clinically meaningful sparing of critical structure becomes possible. Finally, the CTD is expected to reduce the inter-user variability of defining the traditional CTV.

Glioblastoma are known to infiltrate the brain parenchyma instead of forming a solid tumor mass with a defined boundary. Only the part of the tumor with high tumor cell density can be localized through imaging directly. In contrast, brain tissue infiltrated by tumor cells at low density appears normal on current imaging modalities. In clinical practice, a uniform margin is applied to account for microscopic spread of disease. The current treatment planning procedure can potentially be improved by accounting for the anisotropy of tumor growth: Anatomical barriers such as the falx cerebri represent boundaries for migrating tumor cells. In addition, tumor cells primarily spread in white matter and infiltrate gray matter at lower rate. We investigate the use of a phenomenological tumor growth model for treatment planning. The model is based on the Fisher-Kolmogorov equation, which formalizes these growth characteristics and estimates the spatial distribution of tumor cells in normal appearing regions of the brain. The target volume for radiotherapy planning can be defined as an isoline of the simulated tumor cell density. A retrospective study involving 10 glioblastoma patients has been performed. To illustrate the main findings of the study, a detailed case study is presented for a glioblastoma located close to the falx. In this situation, the falx represents a boundary for migrating tumor cells, whereas the corpus callosum provides a route for the tumor to spread to the contralateral hemisphere. We further discuss the sensitivity of the model with respect to the input parameters. Correct segmentation of the brain appears to be the most crucial model input. We conclude that the tumor growth model provides a method to account for anisotropic growth patterns of glioblastoma, and may therefore provide a tool to make target delineation more objective and automated.