Deep learning-based pelvic levator hiatus segmentation from ultrasound images

Purpose

To automatically segment and measure the levator hiatus with a deep learning approach and evaluate the performance between algorithms, sonographers, and different devices.

Methods

Three deep learning models (UNet-ResNet34, HR-Net, and SegNet) were trained with 360 images and validated with 42 images. The trained models were tested with two test sets. The first set included 138 images to evaluate the performance between the algorithms and sonographers. An independent dataset including 679 images assessed the performances of algorithms between different ultrasound devices. Four metrics were used for evaluation: DSC, HDD, the relative error of segmentation area, and the absolute error of segmentation area.

Results

The UNet model outperformed HR-Net and SegNet. It could achieve a mean DSC of 0.964 for the first test set and 0.952 for the independent test set. UNet was creditable compared with three senior sonographers with a noninferiority test in the first test set and equivalent in the two test sets collected by different devices. On average, it took two seconds to process one case with a GPU and 2.4 s with a CPU.

Conclusions

The deep learning approach has good performance for levator hiatus segmentation and good generalization ability on independent test sets. This automatic levator hiatus segmentation approach could help shorten the clinical examination time and improve consistency.

Value and prognostic impact of a deep learning segmentation model of COVID-19 lung lesions on low-dose chest CT

Objectives

1) To develop a deep learning (DL) pipeline allowing quantification of COVID-19 pulmonary lesions on low-dose computed tomography (LDCT). 2) To assess the prognostic value of DL-driven lesion quantification.

Methods

This monocentric retrospective study included training and test datasets taken from 144 and 30 patients, respectively. The reference was the manual segmentation of 3 labels: normal lung, ground-glass opacity(GGO) and consolidation(Cons). Model performance was evaluated with technical metrics, disease volume and extent. Intra- and interobserver agreement were recorded. The prognostic value of DL-driven disease extent was assessed in 1621 distinct patients using C-statistics. The end point was a combined outcome defined as death, hospitalization>10 days, intensive care unit hospitalization or oxygen therapy.

Results

The Dice coefficients for lesion (GGO+Cons) segmentations were 0.75±0.08, exceeding the values for human interobserver (0.70±0.08; 0.70±0.10) and intraobserver measures (0.72±0.09). DL-driven lesion quantification had a stronger correlation with the reference than inter- or intraobserver measures. After stepwise selection and adjustment for clinical characteristics, quantification significantly increased the prognostic accuracy of the model (0.82 vs. 0.90; p<0.0001).

Conclusions

A DL-driven model can provide reproducible and accurate segmentation of COVID-19 lesions on LDCT. Automatic lesion quantification has independent prognostic value for the identification of high-risk patients.

Interrater agreement of contouring of the neurovascular bundles and internal pudendal arteries in neurovascular-sparing magnetic resonance-guided radiotherapy for localized prostate cancer

•

Interrater DSC of the NVB was 0.60 and 0.61 for the left and right side respectively.

•

Interrater DSC of the IPA was 0.59 for the left and right side.

•

Agreement was best for the inferior half (i.e. prostate apex to midgland) of the NVB.

•

Agreement improved with MRI optimization and rater training.

Background and purpose

Radiation damage to neural and vascular tissue, such as the neurovascular bundles (NVBs) and internal pudendal arteries (IPAs), during radiotherapy for prostate cancer (PCa) may cause erectile dysfunction. Neurovascular-sparing magnetic resonance-guided adaptive radiotherapy (MRgRT) aims to preserve erectile function after treatment. However, the NVBs and IPAs are not routinely contoured in current radiotherapy practice. Before neurovascular-sparing MRgRT for PCa can be implemented, the interrater agreement of the contouring of the NVBs and IPAs on pre-treatment MRI needs to be assessed.

Materials and methods

Four radiation oncologists independently contoured the prostate, NVB, and IPA in an unselected consecutive series of 15 PCa patients, on pre-treatment MRI. Dice similarity coefficients (DSCs) for pairwise interrater agreement of contours were calculated. Additionally, the DCS of a subset of the inferior half of the NVB contours (i.e. approximately prostate midgland to apex level) was calculated.

Results

Median overall interrater DSC for the left and right NVB was 0.60 (IQR: 0.54 – 0.68) and 0.61 (IQR: 0.53 – 0.69) respectively and for the left and right IPA 0.59 (IQR: 0.53 – 0.64) and 0.59 (IQR: 0.52 – 0.64) respectively. Median overall interrater DSC for the inferior half of the left NVB was 0.67 (IQR: 0.58 – 0.74) and 0.67 (IQR: 0.61 – 0.71) for the right NVB.

Conclusion

We found that the interrater agreement for the contouring of the NVB and IPA improved with enhancement of the MRI sequence as well as further training of the raters. The agreement was best in the subset of the inferior half of the NVB, where a good agreement is clinically most relevant for neurovascular-sparing MRgRT for PCa.

Comparison of atlas-based auto-segmentation accuracy for radiotherapy in prostate cancer

•

Auto-contouring accuracy and contouring time were evaluated using two procedures.

•

Dice coefficient was better for the multiple atlases procedure than for one atlas.

•

Contouring time of the multiple atlases procedure is clinically acceptable.

Atlas-based auto-segmentation (ABS) procedure used in radiotherapy can be classified into two groups, one using one atlas per patient (sSM) and the other using multiple atlases (sMM). This study evaluated auto-contouring accuracy and contouring time in patients with prostate cancer using the two procedures. The Dice similarity coefficient of sMM was significantly better than that of sSM (prostate [median, 0.81 (range, 0.66–0.91) vs. 0.64 (0.27–0.71), p < 0.01], seminal vesicles [0.49 (0.31–0.80) vs. 0.18 (0.01–0.60), p < 0.05], and rectum [0.81 (0.37–0.91) vs. 0.57 (0.31–0.77), p < 0.01]). The median contouring times were 2.6 (sMM) and 1.3 min (sSM).

The GIRAFE phase II trial on MVCT-based "volumes of the day" and "dose of the day" addresses when and how to implement adaptive radiotherapy for locally advanced head and neck cancer

During exclusive curative radiotherapy for head and neck tumors, the patient's organs at risk (OAR) and target volumes frequently change size and shape, leading to a risk of higher toxicity and lower control than expected on planned dosimetry. Adaptive radiotherapy is often necessary but 1) tools are needed to define the optimal time for replanning, and 2) the subsequent workflow is time-consuming. We designed a prospective study to evaluate 1) the validity of automatically deformed contours on the daily MVCT, in order to safely use the "dose-of the day" tool to check daily if replanning is necessary; 2) the automatically deformed contours on the replanning CT and the time gained in the replanning workflow. Forty-eight patients with T3-T4 and/or involved node >2 cm head and neck squamous cell carcinomas, planned for curative radiotherapy without surgery, will be enrolled. They will undergo treatment with helical IMRT including daily repositioning MVCTs. The contours proposed will be compared weekly on intermediate planning CTs (iCTs) on weeks 3, 4, 5 and 6. On these iCTs both manual recontouring and automated deformable registration of the initial contours will be compared with the contours automatically defined on the MVCT. The primary objective is to evaluate the Dice similarity coefficient (DSC) of the volumes of each parotid gland. The secondary objectives will evaluate, for target volumes and all OARs: the DSC, the mean distance to agreement, and the average surface-to-surface distance. Time between the automatic and the manual recontouring workflows will be compared.

Variability of clinical target volume delineation for rectal cancer patients planned for neoadjuvant radiotherapy with the aid of the platform Anatom-e

Objective

Delineation of treatment volumes is a major source of uncertainties in radiotherapy (RT). This is also true for rectal cancer patients undergoing neoadjuvant RT, with a potential impact on treatment quality. We investigated the role of the digital platform Anatom-e (Anatom-e Information Sytems Ltd., Houston, Texas) in increasing the compliance to follow a specific treatment protocol in a multicentric setting.

Materials and methods

Two clinical cases of locally advanced rectal cancer were chosen. Participants were instructed to follow the 2009 Radiation Therapy Oncology Group consensus atlas and asked to manually segment clinical target volumes (CTVs), for both patient 1 and 2, on day 1 with and without the use of Anatom-e. After one week (day 2), the same radiation oncologist contoured again, with and without Anatom-e, the same CT series. Intraobserver (Intra-OV) and interobserver (Inter-OV) variability were evaluated with the Dice similarity coefficient (DSC), the Hausdorff distance (HD) and mean distance to agreement (MDA).

Results

For clinical case 1, no significant difference was found for Intra-OV and Inter-OV. For clinical case 2, no significant difference was found for Intra-OV but a statistically significant difference was found for Inter-OV in DSC when using or not the platform. Mean DCS was 0.65 (SD: ±0.64; range: 0.58-0.79) for day 1 vs reference volume without Anatom-e and 0.72 (SD: ±0.39; range: 0.67-0.77) (p = 0.03) with it. Mean MDA was lower with Anatom-e (3.61; SD: ±1.33; range: 2.85-4.78) than without (4.14; SD: ±2.97; range: 2.18-5.21), with no statistical significance (p = 0.21) The use of Anatom-e decreased the SD from 2.97 to 1.33. Mean HD was lower with Anatom-e (26.06; SD: ±2.05; range: 24.08-32.62), with no statistical significance (p = 0.14) compared to that without (31.39; SD: ±1.31; range: 26.14-48.72).

Conclusions

The use of Anatom-e decreased the Inter-OV in the CTV delineation process for locally advanced rectal cancer with complex disease presentation planned for neoadjuvant RT. This system may be potentially helpful in increasing the compliance to follow shared guidelines and protocols.