Evaluation of auto-segmentation for brachytherapy of postoperative cervical cancer using deep learning-based workflow

doi:10.21203/rs.3.rs-2100195/v1

Download PDF

Article

Evaluation of auto-segmentation for brachytherapy of postoperative cervical cancer using deep learning-based workflow

https://doi.org/10.21203/rs.3.rs-2100195/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objective: The purpose of this study was to evaluate the accuracy of brachytherapy (BT) planning structures derived from Deep learning (DL) based auto-segmentation compared with standard manual delineation for postoperative cervical cancer.

Methods: We introduced a convolutional neural networks (CNN) which was developed and presented for auto-segmentation in cervical cancer radiotherapy. The dataset of 60 patients received BT of postoperative cervical cancer was used to train and test this model for delineation of high-risk clinical target volume (HRCTV) and organs at risk (OARs). Dice similarity coefficient (DSC), 95% hausdorff distance (95%HD), jaccard coefficient (JC) and dose-volume index (DVI) were used to evaluate the accuracy. The correlation between geometric metrics and dosimetric difference was performed by Spearman’s correlation analysis. The radiation oncologists scored the auto-segmented contours by rating the lever of satisfaction (no edits, minor edits, major edits).

Results: The mean DSC values of DL based model were 0.87, 0.94, 0.86, 0.79 and 0.92 for HRCTV, bladder, rectum, sigmoid and small intestine, respectively. The Bland-Altman test obtained dose agreement for HRCTV_D_90%, HRCTV_D_mean, bladder_D_2cc, sigmoid_D_2cc and small intestine_D_2cc. Wilcoxon’s signed-rank test indicated significant dosimetric differences in bladder_D_0.1cc, rectum_D_0.1cc and rectum_D_2cc(P<0.05). A strong correlation between HRCTV_D_90% with its DSC (R=-0.842, P=0.002) and JC (R=-0.818, P=0.004) were found in Spearman’s correlation analysis. From the physician review, 80% of HRCTVs and 72.5% of OARs in the test dataset were shown satisfaction (no edits).

Conclusion: The proposed DL based model achieved a satisfied agreement between the auto-segmented and manually defined contours of HRCTV and OARs, although the clinical acceptance of small volume dose of OARs around the target was a concern. DL based auto-segmentation was an essential component in cervical cancer workflow which would generate the accurate contouring.

Deep learning

Auto-segmentation

Postoperative cervical cancer

Brachytherapy

There is strong evidence^[1] supporting the use of radiotherapy (RT) as an adjuvant treatment for postoperative cervical cancer in the presence of risk factors, and brachytherapy (BT) is strongly recommended for all women receiving definitive RT. Mauro et al^[2] reported vaginal cuff BT was associated with a reduced recurrence rate (disease free survival of 86.9 months) in the postoperative setting of high-risk patients with early-stage cervical cancer. Indeed, BT is an critical treatment modalities and closely associated with improvements in clinical outcomes^[3]. Modern technique of BT allows adaptive treatment planning based on the three-dimensional (3D) image, and shows more advantageous than conventional two-dimensional (2D) image[1]based approach^[4]. Currently, the application of 3D image-guided BT (IGBT) for each applicator implantation is still limited because of the complex treatment workflow and high workload pressures in China. This highlights the importance of a rapid and accurate method which would improve the IGBT workflow through automation^[5].

Artificial intelligence (AI) has been applied for RT procedures during several steps, starting from initial treatment decision-making till delivery of radiation therapy^[6]. Many studies have demonstrated the accuracy and availability about auto-segmentation of target and organs at risk (OARs) in external beam radiation therapy (EBRT) ^[7-10]. However, relevant research was rarely mentioned in BT^[11-12]. The delineation of planning structures is an important task but also such a labor-intensive part in RT workflow, and always suffers from the inter- and intra-observer variability^[13]. Meanwhile, the contouring uncertainties would be magnified especially in high-dose rate BT owing to the high steep dose gradient^[14]. Therefore, improving the precision of delineation and simplifying the RT procedure are shown practical significance in BT workflow.

Mohammadi et al^[15] reported the ResU-Net deep learning (DL) model achieved a good agreement between the predicted and manually defined contours of bladder, rectum and sigmoid. Jiang et al^[16] used a RefineNet-based DL model to obtain rapid and accurate automatic delineation of CTVs and OARs. As a novel technique, DL has replaced most of the machine learning models due to the superior ability of human equivalent performance^[17]. In automatic delineation of BT workflow, the quality and reliability of DL models should be further evaluated by multiple goals with geometric metrics, dosimetric metrics and even subjective assessment.

The primary purpose of this study was to evaluate the accuracy of DL model to automatically delineate the high risk clinical target volume (HRCTV) and OARs for patients with postoperative cervical cancer in BT workflow.

Experiments

The evaluation of this work was divided into 2 sections. Section 1, the accuracy of DL based auto-segmentation was assessed using objective indicators, including geometric and dosimetric metrics. Section 2, subjective assessment of auto-segmented contours was evaluated by two experienced radiation oncologists. This study was approved by the Ethical Committee of Women’s Hospital, School of Medicine, Zhejiang University, and all methods were performed in accordance with the relevant guidelines and regulation.

Clinical datasets

60 patients of postoperative cervical cancer collected between August 2021 and June 2022 were included in this study, and informed consent was obtained from all participants. The enrolled patients were diagnosed with 2018 International Federation of Gynecology and Obstetrics (FIGO) stage IB1- IIIC1, treated with EBRT (45Gy-50.4Gy, 1.8Gy/fraction) and IGBT (12Gy-30Gy, 6Gy/fraction) as a boost treatment. The average age ± standard deviation of these patients was 48.20±14.52 years old. For each patient, the oral contrast (diatrizoate meglumine) was required for small intestine preparations before each BT implantation session. Meanwhile, the bladder was filled with 150cc of normal saline in the CT scanning room, and the CT images were reconstructed with 512×512 matrix size and 3 mm slice thickness using a Philips Brilliance Big Bore CT scanner system (Philips Healthcare,Best, the Netherlands).

Definitions for volume-based targets (HRCTVs) were established by the Groupe Européen de Curiethérapie-European Society for Radiotherapy and Oncology (GEC-ESTRO)^[18], the boost dose was prescribed to the vaginal surface with multi-needles to the upper vagina. Relevant OARs included for IGBT plans were bladder, rectum, sigmoid and small intestine. All of the manual contours were reviewed and approved by senior radiation oncologists specialized in cervical cancer to generate the standard delineation.

Deep Learning for Segmentation

We introduced a robust deep learning model based on CNN to delineate the CTVs and OARs for cervical cancer patients with EBRT, which has been proven to provide high performance in automatic segmentation of planning structures^[19]. This model is an end-to-end segmentation architecture that can predict pixel class labels in CT images. The inputs to the DL based model were the 2D CT images, and the outputs were the corresponding labels of HRCTV and OARs. The training and testing were performed using a graphics card equipped with IntelCore i7 processor. The final performance of DL based model was assessed using ten-fold cross-validation, in which each fold consisted of randomly selected 40 subjects for training, 10 for validation, and 10 for testing.

Objective indicators

The geometric and dosimetric metrics were used for quantitative analysis. Segmentation was assessed by Dice Similarity Coefficient (DSC), 95% Hausdorff Distance (HD) and Jaccard Coefficient (JC) for geometric calculations. The definitions are as follows:

DSC and JC describe the relative overlap between segmentation A and B. HD is used to quantify the 3D distance between two segmentation surfaces. The 95%HD is the distance that indicates the largest surface-to-surface separation among the closest 95% of surface points. For the complete overlap, the value of HD is 0, and the values of DSC and JC are 1. For the incomplete overlap, the value of HD is large, and the values of DSC and JC are close to 0.

Dose-volume index (DVI) were used to compare the dosimetric metrics. Original BT plans were designed and optimized based on the standard manual contours by using Oncentra Treatment Planning System (Nucletron, Elekta AB, Stockholm, Sweden, V.4.3), and the auto-segmentation structures were transmitted to original BT plans for dosimetric evaluation. For HRCTV, we mainly focused on D_meanand D_90%, where D_meanand D_90% are defined as the average dose and minimal dose to 90% of the HRCTV, respectively. For OARs, we mainly focused on D_0.1cc, D_2cc and D_5cc, where Dxcc represents the minimum dose received by xcm³of an OAR. A prescription dose of ≥60Gy_EQD2 (45-50.4Gy as EBRT) was considered for the target, equivalent dose of 2Gy with an α-to-β ratio of 10, and maximum dose (D_2cc) of 80Gy_EQD2, 65Gy_EQD2, 70Gy_EQD2 and 70Gy_EQD2 to the bladder, rectum, sigmoid and small intestine, respectively, assuming α-to-β ratio of 3. Table 1 is presented the constraints and dosimetric metrics.

Table 1. The constraints and dosimetric metrics for BT planning structures.

Structures	Constraints	Dosimetric metrics
HRCTV	D_90%≥60Gy（EQD2₁₀）^[20]	D_mean, D_90%
Bladder	D_2cc<80Gy（EQD2₃）^[21]	D_0.1cc, D_2cc and D_5cc
Rectum	D_2cc<65Gy（EQD2₃）^[21]	D_0.1cc, D_2cc and D_5cc
Sigmoid	D_2cc<70Gy（EQD2₃）^[22]	D_0.1cc, D_2cc
Small intestine	D_2cc<70Gy（EQD2₃）^[22]	D_0.1cc, D_2cc

Subjective assessment

For qualitative analysis, two experienced radiation oncologists would blindly evaluate the results of manual and automatic delineations on 10 tested patients, and were required to score as needing no edits, minor edits, or major edits. If the auto-segmentation scored as no edits, the contour was considered as suitable for employ in the clinic.

Statistical analysis

Bland-Altman test was calculated for the test of agreement between manual and DL based methods, P>0.05 means agreement of two segmented methods. Wilcoxon’s paired nonparametric signed-rank test was performed to compare the dosimetric differences, P<0.05 indicates the difference is statistically significant. The correlations between geometric metrics and dosimetric difference were evaluated with Spearman’s correlation analysis. All the statistical analyses were performed using IBM SPSS Statistics software (version 19.0, IBM Inc., Armonk, NY, USA) and Python software (version 3.6.5,Anaconda Inc.).

Analysis of geometric metrics

Figure 1 shows the quantitative indicators of segmentation results on the testing dataset with our proposed DL based model. Automatic delineation produced the results for HRCTV with average DSC value of 0.87±0.06, 95%HD of 1.45±0.27mm and JC of 0.78±0.10. The mean DSC values of OARs were all greater than 0.85 except for the sigmoid (0.79). Among all the planning structures, the superior results were found in bladder delineation with average DSC and JC values of 0.94 and 0.89, respectively.

Analysis of dosimetric metrics

The comparisons of dosimetric metrics between two methods using Wilcoxon’s paired nonparametric signed-rank test are presented in Table 2. Significant dosimetric differences were found in bladder_D_0.1cc, rectum_D_0.1cc and rectum_D_2cc (P＜0.05), and the remaining dosimetric indicators were shown no significant differences. For all of the planning structures, both the manual and automatic delineation were able to meet the clinical dose constraints (EQD2₁₀ or EQD2₃). Examples of segmentation and dose distributions from manual and DL based methods are illustrated in Figure 2.

Table 2. Dosimetric parameters of manual and DL based methods in the original clinical BT plans.

Structure	Dosimetric parameters	Manual	DL	Z	P
		mean±standard deviation
HRCTV	D_mean	1311.10±17.22	1293.40±84.19	-0.26	0.799
	D_90%	635.90±10.56	646.80±11.86	-1.08	0.193
Bladder	D_0.1cc	412.80±89.43	369.30±71.07	-1.99	0.047
	D_2cc	269.30±76.65	260.70±72.72	-1.79	0.074
	D_5cc	203.40±46.90	197.90±49.13	-1.74	0.083
Rectum	D_0.1cc	553.10±132.23	448.70±56.94	-2.70	0.007
	D_2cc	288.90±29.03	273.10±26.56	-2.81	0.005
	D_5cc	223.11±20.40	217.10±20.12	-1.85	0.068
Sigmoid	D_0.1cc	114.10±45.18	130.30±33.71	-1.12	0.262
	D_2cc	74.20±26.86	71.40±21.36	-0.82	0.415
Small intestine	D_0.1cc	96.20±65.08	90.70±46.97	-0.51	0.625
	D_2cc	61.80±40.29	60.10±32.53	-0.56	0.574

Correlation analysis between geometric and dosimetric metrics

The results of Spearman’s correlation analysis between geometric metrics and dosimetric metrics (Δdose) are presented in Table 3. No structures showed strong correlation except for the ΔD_90%of HRCTV with its DSC and JC (R=-0.842,P=0.002 and R=-0.818,P=0.004). The correlation heatmap demonstrated the weak link between all of the dosimetric difference and its geometric metrics in the remaining BT planning structures (figure 3).

Table 3. The Spearman’s correlation analysis between geometric metrics and dosimetric metrics.

Structure	ΔDose	Geometric metrics	Correlation analysis
HRCTV	ΔDmean	DSC	R=-0.030,P=0.934
		95%HD	R=0.212,P=0.556
		JC	R=-0.006,P=0.987
HRCTV	ΔD90%	DSC	R=-0.842,P=0.002
		95%HD	R=0.321,P=0.365
		JC	R=-0.818,P=0.004
Bladder	ΔD0.1cc	DSC	R=0.224,P=0.533
		95%HD	R=-0.345,P=0.328
		JC	R=-0.479,P=0.162
Bladder	ΔD2cc	DSC	R=0.122,P=0.738
		95%HD	R=0.280,P=0.434
		JC	R=-0.152,P=0.675
Bladder	ΔD5cc	DSC	R=-0.152,P=0.674
		95%HD	R=0.299,P=0.402
		JC	R=-0.421,P=0.226
Rectum	ΔD0.1cc	DSC	R=-0.164,P=0.651
		95%HD	R=0.248,P=0.489
		JC	R=0.285,P=0.425
Rectum	ΔD2cc	DSC	R=-0.213,P=0.554
		95%HD	R=0.079,P=0.828
		JC	R=-0.396,P=0.257
Rectum	ΔD5cc	DSC	R=-0.239,P=0.506
		95%HD	R=0.031,P=0.933
		JC	R=-0.350,P=0.322
Sigmoid	ΔD0.1cc	DSC	R=0.334,P=0.345
		95%HD	R=-0.419,P=0.228
		JC	R=0.334,P=0.345
Sigmoid	ΔD2cc	DSC	R=-0.152,P=0.675
		95%HD	R=0.395,P=0.258
		JC	R=-0.152,P=0.675
Small Intestine	ΔD0.1cc	DSC	R=-0.554,P=0.097
		95%HD	R=0.248,P=0.489
		JC	R=-0.541,P=0.106
Small Intestine	ΔD2cc	DSC	R=-0.544,P=0.104
		95%HD	R=0.388,P=0.268
		JC	R=-0.534,P=0.112

DSC: dice similarity coefficient; HD: hausdorff distance; JC: jaccard coefficient; ΔDose: dosimetric differences between two segmented methods.

Test agreement between manual and DL based approaches

The Bland-Altman test was mainly calculated for HRCTV_D_90%, HRCTV_D_mean, bladder_D_2cc, rectum_D_2cc, sigmoid_D_2cc and small intestine_D_2cc. The figure 4 showed 95% consistent limits for all of the BT planning structures between two methods. The test of agreement for DL based method can be evaluated according to the number of the points outside the 95% consistent limits and the maximum difference within the consistent limits. From the Bland-Altman plot, HRCTV_D_90%, HRCTV_D_mean, bladder_D_2cc, sigmoid_D_2cc and small intestine_D_2ccshowed no significant inconsistency (P>0.05) between two segmented methods.

Subjective assessment

Two radiation oncologists scoring of the automatically generated contours on the 10 tested patients is shown in Table 4. For the HRCTV, 80% of the segmentation was clinical acceptance (no edits) and 20% was scored as needing minor edits. For bladder, rectum, sigmoid and small intestine, 90%, 70%, 50% and 80% were clinical acceptance, respectively, while 10% of rectum and 30% of sigmoid were scored as needing major edits.

Table 4. Qualitative scores of the auto-segmented contours on 10 tested CT scans.

Structure	No edits	Minor edits	Major edits
HRCTV	8 (80%)	2 (20%)	0 (0%)
Bladder	9 (90%)	1 (10%)	0 (0%)
Rectum	7 (70%)	2 (20%)	1 (10%)
Sigmoid	5 (50%)	2 (20%)	3 (30%)
Small intestine	8 (80%)	2(20%)	0 (0%)

IGBT using 3D images for RT planning allows for more precise and individualized treatment compared even to stereotactic body radiotherapy (SBRT)^[23-24]. However, modern IGBT procedures with increasing real-time steps and complexities will require for more technical and manpower resources before confident applications. As for the economic and efficient factors, BT provides fewer reimbursement and higher fractional treatment time compared with EBRT, which is obvious in the declining trend of BT in the last decades^[25]. The development of AI has the potential possibilities to change this phenomenon of BT due to its ability of solving most challenging tasks, and then might contribute to the progress of IGBT.

Segmentation of targets and OARs is an essential task in the treatment planning process of IGBT. GEC-ESTRO recommends magnetic resonance imaging (MRI) as the gold standard in IGBT because of its excellent soft tissue contrast for HRCTV and OARs delineation^[26]. Compared with MRI, there is more challenging for auto segmentation in CT images owing to the lack of visible anatomical edges, especially in the definition of target volume^[27]. At present, CT images are still the most applied for IGBT in China due to the technical limitation of MRI simulation. To the best of our knowledge, few work has been studied for CT based automatic delineation of BT planning structures. The purpose of this study is trying to compare the performance of DL based auto-segmentation against standard contours in postoperative cervical cancer patients with CT images.

We observed that DL based model generated structures with average DSC of 0.87, 0.94, 0.86, 0.79 and 0.92 for HRCTV, bladder, rectum, sigmoid and small intestine, respectively. The performances of auto segmentation for target and OARs in CT based IGBT from other DL model is presented in Table 5. The geometric similarity of all BT planning structures were equivalent to or better than other published literature^[28,15,16]. Among all the DL based models, the superior results were obtained for the bladder delineation (DSC of 0.86-0.96, HD of 4.05-19.98). This is mainly because of the bladder filling protocol with relatively consistent volume as well as the high contrast in CT images. The geometric metrics of sigmoid in our study were worse than the other BT planning structures, which were reported with the similar results for other DL model. The spatial variation of sigmoid in fractional BT and low contrast in CT images may explain this reason^[29]. Another interest finding in this work was small intestine with superior geometric results (DSC of 0.92, 95%HD of 8.83 and JC of 0.85). As we all know, the location of small intestine is always different in the RT process and confusion with other soft tissue in CT images. Therefore, the representative features of small intestine was difficult to extract and classify using CNN model. In order to recognize the small intestine in CT images, we used oral contrast protocol for small intestine preparations to obtain the distinct training dataset of DL based model. The accuracy of small intestine was satisfying and could be as supplement for our previous study which was absent to assess this organ^[19]. Usually, the better auto-segmented results are achieved through increasing the number of training dataset. Nevertheless, our DL model used fewer enrolled patients for training and generated higher overlap with standard manual contours compared to other model in target volume segmentation.

Table 5. Summary of DL based auto segmentation results for target and OARs in CT based IGBT from other published literature.

Group	DL model	Enrolled patients	Structures	DSC	HD(mm)
Zhang et al.(2020)	DSD-UNET	91 patients with cervical cancer; 73 cases for training and 18 cases for testing.	HRCTV	0.83	8.1	[28]
			Bladder	0.87	12.1
			Rectum	0.82	9.2
			Sigmoid	0.65	19.6
			Small intestine	0.80	27.8
Mohammadi et al.(2021)	ResU-Net	113 patients with cervical cancer; 83 cases for training and 30 cases for testing.	Bladder	0.96	4.05	[15]
			Rectum	0.97	1.96
			Sigmoid	0.93	3.15
Jiang et al.(2021)	RefineNet	210 patients with cervical cancer; 200 cases for training and 10 cases for testing.	HRCTV	0.86	6.0	[16]
			Bladder	0.86	19.98
			Rectum	0.86	12.27
			Sigmoid	0.66	98.41
			Small intestine	0.56	68.12
Our method	Modified CNN	60 patients with cervical cancer; 50 cases for training and 10 cases for testing.	HRCTV	0.87	1.45
			Bladder	0.94	4.52
			Rectum	0.86	2.52
			Sigmoid	0.79	10.92
			Small intestine	0.92	8.83

CNN: convolutional neural network; DL model: deep learning model; HD: including 95%HD

Yoganathan et al^[12] demonstrated the importance of dosimetric evaluation over geometric evaluation for an automatic segmentation problem. They found the poor geometric metrics of sigmoid and small intestine, while closely matched dosimetric metrics with the manual segmentation. Similar results were presented in our work, particularly in auto segmentation of sigmoid. In addition, significant dosimetric differences were found in bladder_D_0.1cc, rectum_D_0.1cc and rectum_D_2cc (P＜0.05), and the remaining dosimetric parameters were shown no significant differences. These data indicated the auto-segmented OARs inside of the high-dose region (Fig.2) remains necessary to be reviewed by senior radiation oncologists rather than geometric values. The heatmap of correlation analysis showed there was no clear strong relationship between geometric metrics and dosimetric differences for OARs. However, the only strong correlation was found for the D_90% of HRCTV with its DSC and JC (R=-0.842 and -0.818, respectively). This quantitative analysis proved that more accurate delineation of target volume would carry out more precise delivery of radiation dose in IGBT workflow.

Besides of evaluation for dosimetric differences, we also used the test of agreement to verify the performance of DL based auto segmentation. The automatic method has been proven to obtain dose consistency for HRCTV_D_90%, HRCTV_D_mean, bladder_D_2cc, sigmoid_D_2cc and small intestine_D_2ccwhich were calculated by Bland-Altman test. Certainly, the test would still achieve the dose agreement even for the situation of segmentation inaccuracies (outside of the high-dose region) such as sigmoid. Therefore, the multiple objective indicators should be used for assessing the accuracy of DL based auto segmentation which would generate the meaningful results. As for subjective assessment, the clinical acceptance rates were 80% for the HRCTVs and 72.5% for the OARs. The quality of 10% of rectum and 30% of sigmoid were unsatisfactory with scored as needing major edits. Overall, the quantitative and qualitative analyses evaluated the automatic delineation were relatively consistent.

In this work, we investigated the performance of DL based auto segmentation in postoperative cervical cancer patients treated with IGBT. In fact, automatic approach would reduce the waiting time for patients in adaptive IGBT process, which may decrease the uncertainties such as displacement of applicator. As a novel technology, DL based method has the potential ability to link the EBRT and BT for full workflow of cervical cancer, especially in the dosimetric prediction for cold point of the target and hot point of the OARs in combined therapy. This work was still focus on the problem of auto-segmented accuracy which would be an essential part implemented in the IGBT workflow of cervical cancer.

There are still several limitations in this study. First, our DL based model was trained and verified using the clinical dataset treated with multiple metal needles, which may not be suitable for other situation such as vaginal applicator applied for treating patients. It mostly affects the HRCTV definition generated by DL based method when different catheters are implanted. Second, we used the bladder and small intestine protocols for preparation of treatment while different centers employ their own methods such as contrast agent inside the rectum. Therefore, increasing the amount of training data including various applicators or protocols in IGBT workflow could make the DL model more robust to improve performance further.

Segmentation of planning structures is an important component of RT treatment planning, and automatic approach would relieve physicians from the labor-intensive tasks in adaptive IGBT workflow. The proposed DL based model achieved a satisfied agreement between the auto-segmented and manually defined contours of HRCTV and OARs in postoperative cervical cancer, although the clinical acceptance of small volume of OARs inside of the high-dose region was a concern.

Data availability

The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.

Chino J, Annunziata CM, Beriwal S, et al.Radiation Therapy for Cervical Cancer: Executive Summary of an ASTRO Clinical Practice Guideline. Pract Radiat Oncol. 2020 Jul-Aug; 10(4): 220–234.
Mauro GP, Kleine RT, Costa SC, et al. Vaginal cuff brachytherapy in the adjuvant setting for patients with high-risk early-stage cervical cancer. Brachytherapy. 2019 Nov-Dec; 18(6): 747–752.
Contreras J, Srivastava A, Chundury A, et al. Long-term outcomes of intensity-modulated radiation therapy (IMRT) and high dose rate brachytherapy as adjuvant therapy after radical hysterectomy for cervical cancer. Int J Gynecol Cancer. 2020 Aug;30(8):1157–1161.
Harkenrider MM, Alite F, Silva SR, et al. Image-based brachytherapy for the treatment of cervical cancer. Int J Radiat Oncol Biol Phys. 2015;92:921–934.
Banerjee S, Goyal S, Mishra S, et al. Artificial intelligence in brachytherapy: a summary of recent developments. Br J Radiol. 2021 Jun 1;94(1122):20200842.
Huynh E, Hosny A, Guthier C, et al. Artificial intelligence in radiation oncology. Nat Rev Clin Oncol. 2020 Dec;17(12):771–781.
Liu Z, Liu X,Guan H,et al. Development and Validation of a Deep Learning Algorithm for Auto-Delineation of Clinical Target Volume and Organs at Risk in Cervical Cancer Radiotherapy. Radiother Oncol. 2020; 153:172–179.
Rhee DJ,Jhingran A,Rigaud B,et al. Automatic contouring system for cervical cancer using convolutional neural networks.Phys Med. 2020;47:5648–5658.
Wang Z,Chang YK,Peng Z,et al. Evaluation of deep learning-based auto-segmentation algorithms for delineating clinical target volume and organs at risk involving data for 125 cervical cancer patients. J Appl Clin Med Phys. 2020; 21(12): 272–279.
Liu Z, Liu X, Xiao B, et al. Segmentation of organs-at-risk in cervical cancer CT images with a convolutional neural network. Med Phys. 2020;69:184–191.
Cao YF, Vassantachart A, Ragab O, et al. Automatic segmentation of high-risk clinical target volume for tandem-and-ovoids brachytherapy patients using an asymmetric dual-path convolutional neural network. Med Phys. 2022 Mar;49(3):1712–1722.
Yoganathan SA, Paul SN, Paloor S, et al. Automatic segmentation of magnetic resonance images for high-dose-rate cervical cancer brachytherapy using deep learning. Med Phys. 2022 Mar; 49(3): 1571–1584.
Duane FK, Langan B, Gillham C, et al. Impact of delineation uncertainties on dose to organs at risk in CT-guided intracavitary brachytherapy. Brachytherapy. 2014;13:210–218.
Saarnak AE, Boersma M, van Bunningen BN, et al. Inter-observer variation in delineation of bladder and rectum contours for brachytherapy of cervical cancer. Radiother Oncol. 2000; 56: 37–42.
Mohammadi R, Shokatian I, Salehi M, et al. Deep learning-based auto-segmentation of organs at risk in high-dose rate brachytherapy of cervical cancer. Radiother Oncol. 2021 Jun; 159: 231–240.
Jiang X, Wang F, Chen Y, et al. RefineNet-based automatic delineation of the clinical target volume and organs at risk for three-dimensional brachytherapy for cervical cancer. Ann Transl Med. 2021 Dec; 9(23):1721.
LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015; 521(7553): 436–444.
Haie-Meder C, Potter R, Limbergen EV, et al. Recommendations from Gynaecological (GYN) GEC-ESTRO Working Group (I): Concepts and terms in 3D image based 3D treatment planning in cervix cancer brachytherapy with emphasis on MRI assessment of GTV and CTV. Radiother Oncol. 2005; 74:235–245.
Wang JH, Chen YY, Xie HL, et al. Evaluation of auto-segmentation for EBRT planning structures using deep learning-based workflow on cervical cancer. Sci Rep. 2022 Aug 11; 12(1): 13650.
Prescribing, recording, and reporting brachytherapy for cancer of the cervix ICRU report 89. Bethesda:ICRU,2016;1-258.
Georg P, Potter R, Georg D, et al. Dose effect relationship for late side effects of the rectum and urinary bladder in magnetic resonance image-guided adaptive cervix cancer brachytherapy. Int J Radiat Oncol Biol Phys. 2012; 82: 653–657.
Potter R, Tanderup K, Kirisits C, et al. The EMBRACE II study: The outcome and prospect of two decades of evolution within the GEC-ESTRO GYN working group and the EMBRACE studies. Clin Transl Radiat Oncol. 2018;9:48–60.
Lee TH, Song C, Kim IA, et al. Stereotactic ablative body radiotherapy boost for cervical cancer when brachytherapy boost is not feasible. Radiat Oncol. 2021 Aug 12; 16(1): 148.
Shenker R, Stephens SJ, Davidson B, et al. Role of stereotactic body radiotherapy in gynecologic radiation oncology. Int J Gynecol Cancer. 2022 Mar;32(3):372–379.
Petereit DG, Frank SJ, Viswanathan AN, et al. Brachytherapy: where has it gone? J Clin Oncol. 2015 Mar 20;33(9):980–982.
Dimopoulos JC, Petrow P, Tanderup K, et al. Recommendations from Gynaecological (GYN) GEC-ESTRO Working Group (IV): basic principles and parameters for MR imaging within the frame of image based adaptive cervix cancer brachytherapy. Radiother Oncol. 2012;103:113–122.
Cardenas CE, McCarroll RE, Court LE, et al. Deep learning algorithm for auto-delineation of high-risk oropharyngeal clinical target volumes with built-in dice similarity coefficient parameter optimization function. Int J Radiat Oncol Biol Phys. 2018;101: 468–478.
Zhang DG, Yang ZY, Jiang S, et al. Automatic segmentation and applicator reconstruction for CT-based brachytherapy of cervical cancer using 3D convolutional neural networks. J Appl Clin Med Phys. 2020 Oct; 21(10): 158–169.
Jamema SV, Mahantshetty U, Tanderup K, et al. Inter-application variation of dose and spatial location of D2cm3 volumes of OARs during MR image based cervix brachytherapy. Radiother Oncol. 2013;107(1):58–62.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Evaluation of auto-segmentation for brachytherapy of postoperative cervical cancer using deep learning-based workflow

Status:

Version 1

Abstract

Figures

Introduction

Methods and materials

Results

Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1