Inter-observer variability of organ contouring for preclinical studies with cone beam Computed Tomography imaging

Deep Learning For Automatic Gross Tumor Volumes Contouring in Esophageal Cancer Based on Contrast-Enhanced Computed Tomography Images: A Multi-Institutional Study

PURPOSE

To develop and externally validate an automatic artificial intelligence (AI) tool for delineating gross tumor volume (GTV) in patients with esophageal squamous cell carcinoma (ESCC), which can assist in neo-adjuvant or radical radiation therapy treatment planning.

METHODS AND MATERIALS

In this multi-institutional study, contrast-enhanced CT images from 580 eligible ESCC patients were retrospectively collected. The GTV contours delineated by 2 experts via consensus were used as ground truth. A 3-dimensional deep learning model was developed for GTV contouring in the training cohort and internally and externally validated in 3 validation cohorts. The AI tool was compared against 12 board-certified experts in 25 patients randomly selected from the external validation cohort to evaluate its assistance in improving contouring performance and reducing variation. Contouring performance was measured using dice similarity coefficient (DSC) and average surface distance. Additionally, our previously established radiomics model for predicting pathologic complete response was used to compare AI-generated and ground truth contours, to assess the potential of the AI contouring tool in radiomics analysis.

RESULTS

The AI tool demonstrated good GTV contouring performance in multicenter validation cohorts, with median DSC values of 0.865, 0.876, and 0.866 and median average surface distance values of 0.939, 0.789, and 0.875 mm, respectively. Furthermore, the AI tool significantly improved contouring performance for half of 12 board-certified experts (DSC values, 0.794-0.835 vs 0.856-0.881, P = .003-0.048), reduced the intra- and interobserver variations by 37.4% and 55.2%, respectively, and saved contouring time by 77.6%. In the radiomics analysis, 88.7% of radiomic features from ground truth and AI-generated contours demonstrated stable reproducibility, and similar pathologic complete response prediction performance for these contours (P = .430) was observed.

CONCLUSIONS

Our AI contouring tool can improve GTV contouring performance and facilitate radiomics analysis in ESCC patients, which indicates its potential for GTV contouring during radiation therapy treatment planning and radiomics studies.

Segment anything model for medical image segmentation: Current applications and future directions

Due to the inherent flexibility of prompting, foundation models have emerged as the predominant force in the fields of natural language processing and computer vision. The recent introduction of the Segment Anything Model (SAM) signifies a noteworthy expansion of the prompt-driven paradigm into the domain of image segmentation, thereby introducing a plethora of previously unexplored capabilities. However, the viability of its application to medical image segmentation remains uncertain, given the substantial distinctions between natural and medical images. In this work, we provide a comprehensive overview of recent endeavors aimed at extending the efficacy of SAM to medical image segmentation tasks, encompassing both empirical benchmarking and methodological adaptations. Additionally, we explore potential avenues for future research directions in SAM's role within medical image segmentation. While direct application of SAM to medical image segmentation does not yield satisfactory performance on multi-modal and multi-target medical datasets so far, numerous insights gleaned from these efforts serve as valuable guidance for shaping the trajectory of foundational models in the realm of medical image analysis. To support ongoing research endeavors, we maintain an active repository that contains an up-to-date paper list and a succinct summary of open-source projects at https://github.com/YichiZhang98/SAM4MIS.

Deep learning-based automatic segmentation of cardiac substructures for lung cancers

PURPOSE

Accurate and comprehensive segmentation of cardiac substructures is crucial for minimizing the risk of radiation-induced heart disease in lung cancer radiotherapy. We sought to develop and validate deep learning-based auto-segmentation models for cardiac substructures.

MATERIALS AND METHODS

Nineteen cardiac substructures (whole heart, 4 heart chambers, 6 great vessels, 4 valves, and 4 coronary arteries) in 100 patients treated for non-small cell lung cancer were manually delineated by two radiation oncologists. The valves and coronary arteries were delineated as planning risk volumes. An nnU-Net auto-segmentation model was trained, validated, and tested on this dataset with a split ratio of 75:5:20. The auto-segmented contours were evaluated by comparing them with manually drawn contours in terms of Dice similarity coefficient (DSC) and dose metrics extracted from clinical plans. An independent dataset of 42 patients was used for subjective evaluation of the auto-segmentation model by 4 physicians.

RESULTS

The average DSCs were 0.95 (+/- 0.01) for the whole heart, 0.91 (+/- 0.02) for 4 chambers, 0.86 (+/- 0.09) for 6 great vessels, 0.81 (+/- 0.09) for 4 valves, and 0.60 (+/- 0.14) for 4 coronary arteries. The average absolute errors in mean/max doses to all substructures were 1.04 (+/- 1.99) Gy and 2.20 (+/- 4.37) Gy. The subjective evaluation revealed that 94% of the auto-segmented contours were clinically acceptable.

CONCLUSION

We demonstrated the effectiveness of our nnU-Net model for delineating cardiac substructures, including coronary arteries. Our results indicate that this model has promise for studies regarding radiation dose to cardiac substructures.

Effect of Dataset Size and Medical Image Modality on Convolutional Neural Network Model Performance for Automated Segmentation: A CT and MR Renal Tumor Imaging Study

The aim of this study is to investigate the use of an exponential-plateau model to determine the required training dataset size that yields the maximum medical image segmentation performance. CT and MR images of patients with renal tumors acquired between 1997 and 2017 were retrospectively collected from our nephrectomy registry. Modality-based datasets of 50, 100, 150, 200, 250, and 300 images were assembled to train models with an 80-20 training-validation split evaluated against 50 randomly held out test set images. A third experiment using the KiTS21 dataset was also used to explore the effects of different model architectures. Exponential-plateau models were used to establish the relationship of dataset size to model generalizability performance. For segmenting non-neoplastic kidney regions on CT and MR imaging, our model yielded test Dice score plateaus of [Formula: see text] and [Formula: see text] with the number of training-validation images needed to reach the plateaus of 54 and 122, respectively. For segmenting CT and MR tumor regions, we modeled a test Dice score plateau of [Formula: see text] and [Formula: see text], with 125 and 389 training-validation images needed to reach the plateaus. For the KiTS21 dataset, the best Dice score plateaus for nn-UNet 2D and 3D architectures were [Formula: see text] and [Formula: see text] with number to reach performance plateau of 177 and 440. Our research validates that differing imaging modalities, target structures, and model architectures all affect the amount of training images required to reach a performance plateau. The modeling approach we developed will help future researchers determine for their experiments when additional training-validation images will likely not further improve model performance.

Development of a PTV margin for preclinical irradiation of orthotopic pancreatic tumors derived from a well-known recipe for humans

In human radiotherapy a safety margin (PTV margin) is essential for successful irradiation and is usually part of clinical treatment planning. In preclinical radiotherapy research with small animals, most uncertainties and inaccuracies are present as well, but according to the literature a margin is used only scarcely. In addition, there is only little experience about the appropriate size of the margin, which should carefully be investigated and considered, since sparing of organs at risk or normal tissue is affected. Here we estimate the needed margin for preclinical irradiation by adapting a well-known human margin recipe from van Herck et al. to the dimensions and requirements of the specimen on a small animal radiation research platform (SARRP). We adjusted the factors of the described formula to the specific challenges in an orthotopic pancreatic tumor mouse model to establish an appropriate margin concept. The SARRP was used with its image-guidance irradiation possibility for arc irradiation with a field size of 10 × 10 mm² for 5 fractions. Our goal was to irradiate the clinical target volume (CTV) of at least 90% of our mice with at least 95% of the prescribed dose. By carefully analyzing all relevant factors we gain a CTV to planning target volume (PTV) margin of 1.5 mm for our preclinical setup. The stated safety margin is strongly dependent on the exact setting of the experiment and has to be adjusted for other experimental settings. The few stated values in literature correspond well to our result. Even if using margins in the preclinical setting might be an additional challenge, we think it is crucial to use them to produce reliable results and improve the efficacy of radiotherapy.

Automated Contouring and Planning in Radiation Therapy: What Is 'Clinically Acceptable'?

Developers and users of artificial-intelligence-based tools for automatic contouring and treatment planning in radiotherapy are expected to assess clinical acceptability of these tools. However, what is 'clinical acceptability'? Quantitative and qualitative approaches have been used to assess this ill-defined concept, all of which have advantages and disadvantages or limitations. The approach chosen may depend on the goal of the study as well as on available resources. In this paper, we discuss various aspects of 'clinical acceptability' and how they can move us toward a standard for defining clinical acceptability of new autocontouring and planning tools.

Dose-volume-based evaluation of convolutional neural network-based auto-segmentation of thoracic organs at risk

•

Dice score and Hausdorff distance do not correlate with dose-volume-based results.

•

Auto-contours close to the tumor or in entry/exit beams should be checked.

•

Heart and esophagus must be checked for locally advanced non-small cell lung cancer.

•

Bronchi must be checked for peripheral early-stage non-small cell lung cancer.

•

Every treatment plan still passed the clinical goals for the manual organs at risk.

Background and purpose

The geometrical accuracy of auto-segmentation using convolutional neural networks (CNNs) has been demonstrated. This study aimed to investigate the dose-volume impact of differences between automatic and manual OARs for locally advanced (LA) and peripherally located early-stage (ES) non-small cell lung cancer (NSCLC).

Material and methods

A single CNN was created for automatic delineation of the heart, lungs, main left and right bronchus, esophagus, spinal cord and trachea using 55/10/40 patients for training/validation/testing. Dice score coefficient (DSC) and 95th percentile Hausdorff distance (HD95) were used for geometrical analysis. A new treatment plan based on the auto-segmented OARs was created for each test patient using 3D for ES-NSCLC (SBRT, 3–8 fractions) and IMRT for LA-NSCLC (24–35 fractions). The correlation between geometrical metrics and dose-volume differences was investigated.

Results

The average (±1 SD) DSC and HD95 were 0.82 ± 0.07 and 16.2 ± 22.4 mm, while the average dose-volume differences were 0.5 ± 1.5 Gy (ES) and 1.5 ± 2.8 Gy (LA). The geometrical metrics did not correlate with the observed dose-volume differences (average Pearson for DSC: −0.27 ± 0.18 (ES) and −0.09 ± 0.12 (LA); HD95: 0.1 ± 0.3 mm (ES) and 0.2 ± 0.2 mm (LA)).

Conclusions

After post-processing, manual adjustments of automatic contours are only needed for clinically relevant OARs situated close to the tumor or within an entry or exit beam e.g., the heart and the esophagus for LA-NSCLC and the bronchi for ES-NSCLC. The lungs do not need to be checked further in detail.

Automatic contouring of normal tissues with deep learning for preclinical radiation studies

Objective.Delineation of relevant normal tissues is a bottleneck in image-guided precision radiotherapy workflows for small animals. A deep learning (DL) model for automatic contouring using standardized 3D micro cone-beam CT (μCBCT) volumes as input is proposed, to provide a fully automatic, generalizable method for normal tissue contouring in preclinical studies.Approach.A 3D U-net was trained to contour organs in the head (whole brain, left/right brain hemisphere, left/right eye) and thorax (complete lungs, left/right lung, heart, spinal cord, thorax bone) regions. As an important preprocessing step, Hounsfield units (HUs) were converted to mass density (MD) values, to remove the energy dependency of theμCBCT scanner and improve generalizability of the DL model. Model performance was evaluated quantitatively by Dice similarity coefficient (DSC), mean surface distance (MSD), 95th percentile Hausdorff distance (HD_95p), and center of mass displacement (ΔCoM). For qualitative assessment, DL-generated contours (for 40 and 80 kV images) were scored (0: unacceptable, manual re-contouring needed - 5: no adjustments needed). An uncertainty analysis using Monte Carlo dropout uncertainty was performed for delineation of the heart.Main results.The proposed DL model and accompanying preprocessing method provide high quality contours, with in general median DSC > 0.85, MSD < 0.25 mm, HD_95p < 1 mm and ΔCoM < 0.5 mm. The qualitative assessment showed very few contours needed manual adaptations (40 kV: 20/155 contours, 80 kV: 3/155 contours). The uncertainty of the DL model is small (within 2%).Significance.A DL-based model dedicated to preclinical studies has been developed for multi-organ segmentation in two body sites. For the first time, a method independent of image acquisition parameters has been quantitatively evaluated, resulting in sub-millimeter performance, while qualitative assessment demonstrated the high quality of the DL-generated contours. The uncertainty analysis additionally showed that inherent model variability is low.