scholarly journals A deep learning-based auto-segmentation system for organs-at-risk on whole-body computed tomography images for radiation therapy

2021 ◽  
Author(s):  
Xuming Chen ◽  
Shanlin Sun ◽  
Narisu Bai ◽  
Hao Tang ◽  
Qianqian Liu ◽  
...  
Keyword(s):  
Computed Tomography ◽  
At Risk ◽  
Radiation Therapy ◽  
Deep Learning ◽  
Organs At Risk ◽  
Whole Body ◽  
Computed Tomography Images ◽  
Whole Body Computed Tomography

scholarly journals Clinically Applicable Segmentation of Head and Neck Anatomy for Radiotherapy: Deep Learning Algorithm Development and Validation Study

10.2196/26151 ◽  
2021 ◽  
Vol 23 (7) ◽  
pp. e26151
Author(s):  
Stanislav Nikolov ◽  
Sam Blackwell ◽  
Alexei Zverovitch ◽  
Ruheena Mendes ◽  
Michelle Livne ◽  
...  
Keyword(s):  
Computed Tomography ◽  
At Risk ◽  
Deep Learning ◽  
Clinical Practice ◽  
Head And Neck ◽  
Organs At Risk ◽  
Data Set ◽  
Computed Tomography Scans ◽  
Neck Anatomy ◽  
Organ At Risk

Background Over half a million individuals are diagnosed with head and neck cancer each year globally. Radiotherapy is an important curative treatment for this disease, but it requires manual time to delineate radiosensitive organs at risk. This planning process can delay treatment while also introducing interoperator variability, resulting in downstream radiation dose differences. Although auto-segmentation algorithms offer a potentially time-saving solution, the challenges in defining, quantifying, and achieving expert performance remain. Objective Adopting a deep learning approach, we aim to demonstrate a 3D U-Net architecture that achieves expert-level performance in delineating 21 distinct head and neck organs at risk commonly segmented in clinical practice. Methods The model was trained on a data set of 663 deidentified computed tomography scans acquired in routine clinical practice and with both segmentations taken from clinical practice and segmentations created by experienced radiographers as part of this research, all in accordance with consensus organ at risk definitions. Results We demonstrated the model’s clinical applicability by assessing its performance on a test set of 21 computed tomography scans from clinical practice, each with 21 organs at risk segmented by 2 independent experts. We also introduced surface Dice similarity coefficient, a new metric for the comparison of organ delineation, to quantify the deviation between organ at risk surface contours rather than volumes, better reflecting the clinical task of correcting errors in automated organ segmentations. The model’s generalizability was then demonstrated on 2 distinct open-source data sets, reflecting different centers and countries to model training. Conclusions Deep learning is an effective and clinically applicable technique for the segmentation of the head and neck anatomy for radiotherapy. With appropriate validation studies and regulatory approvals, this system could improve the efficiency, consistency, and safety of radiotherapy pathways.

scholarly journals Incidental Findings on Whole-body Computed Tomography in Major Trauma Patients: Who and What?

2021 ◽  
pp. 000313482199868
Author(s):  
Ping-Yuan Liu ◽  
Ling-Wei Kuo ◽  
Chien-Hung Liao ◽  
Chi-Hsun Hsieh ◽  
Francesco Bajani ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Incidental Findings ◽  
Major Trauma ◽  
Neck Region ◽  
Whole Body ◽  
Trauma Patients ◽  
Computed Tomography Images ◽  
Whole Body Computed Tomography ◽  
Head Neck

Purpose Whole-body computed tomography (WBCT) scans are frequently used for trauma patients, and sometimes, nontraumatic findings are observed. We aimed to investigate the characteristics of patients with nontraumatic findings on WBCT. Methods From 2013 to 2016, adult trauma patients who underwent WBCT were enrolled. The proportions of nontraumatic findings in different anatomical regions were studied. Nontraumatic findings were classified and evaluated as clinically important findings and findings that needed no further follow-up or treatment. The characteristics of the patients with nontraumatic findings were analyzed and compared with those of patients without nontraumatic findings. Results Two hundred seventeen patients were enrolled in this study during the 3-year study period, and 89 (41.0%) patients had nontraumatic findings. Nontraumatic findings were found more frequently in the abdomen (69.2%) than in the head/neck (17.3%) and chest regions (13.5%). In total, 31.3% of the findings needed further follow-up or treatment. Patients with nontraumatic findings that needed further management were significantly older than those without nontraumatic findings (57.3 vs. 38.9; P < .001), particularly those with abdominal nontraumatic findings (57.9 vs. 41.3; P < .001). A significantly higher proportion of women were observed in the group with head/neck nontraumatic findings that needed further management than in the group without nontraumatic findings (56.3% vs 24.9%; P = .015). Conclusions Whole-body computed tomography could provide alternative benefits for nontraumatic findings. Whole-body computed tomography images should be read carefully for nontraumatic findings, particularly for elderly patients or the head/neck region of female patients. A comprehensive program for the follow-up of nontraumatic findings is needed.

scholarly journals Analysis of Synchronization of Respiration-induced Motions in the Duodenum and Stomach for the Primary Tumor in Pancreatic Cancer Using Four-Dimensional Computed Tomography

2020 ◽  
Author(s):  
Rei Umezawa ◽  
Akihisa Wakita ◽  
Yoshiyuki Katsuta ◽  
Yoshinori Ito ◽  
Satoshi Nakamura ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Pancreatic Cancer ◽  
At Risk ◽  
Primary Tumor ◽  
Organs At Risk ◽  
Free Breathing ◽  
Computed Tomography Images ◽  
Significant Difference ◽  
Respiratory Phase ◽  
Respiratory Phases

Abstract Background: We investigated the synchronization of respiration-induced motions at the primary tumor and organs at risk at radiation planning for pancreatic cancer.Methods: Four-dimensional (4D) computed tomography images were acquired under the condition of shallow free breathing in patients with pancreatic cancer. The gross tumor volume (GTV), duodenum (DU) and stomach (ST) were contoured. With reference to the 50% phase (exhale phase), excursions of respiration-induced motions of the GTV, DU and ST were measured. Based on the shift of the GTV, we investigated the synchronization of respiration-induced motions between each contouring target. We examined the differences in the overlap volumes between PTV and the planning organ at risk volume (PRV) in ST and DU and the differences in mean doses to the ST and DU in each respiratory phase.Results: Nine patients with pancreatic cancer were analyzed in this study. The mean maximum three-dimensional excursions at the GTV, DU and ST were 9.6, 9.8 and 11.4 mm, respectively. At phase 0% and 90% (inhale phases), mean distance changes in the positional relationship with the GTV were 0.3 and 0.7 mm, respectively, for the DU, and -2.5 and -2.4 mm, respectively for the ST, respectively. There was no significant difference in distance changes between each respiratory phase in the DU (p = 0.568), while there was a significant difference in distance changes in the ST (p < 0.001). There was a slight increase in the overlap volume between PTV and PRV in the expiratory phase in ST, but there were no significant differences between the respiratory phases either in ST or DU (p = 0.101 and 0.559). There was a significant difference in the change rates of mean doses in the respiratory phases in ST (p = 0.023), but there was no significant difference in DU (p = 0.933). Conclusions: Our results indicate that the DU but not the ST might move synchronously with GTV due to respiration.

scholarly journals Clinically Applicable Segmentation of Head and Neck Anatomy for Radiotherapy: Deep Learning Algorithm Development and Validation Study (Preprint)

2020 ◽  
Author(s):  
Stanislav Nikolov ◽  
Sam Blackwell ◽  
Alexei Zverovitch ◽  
Ruheena Mendes ◽  
Michelle Livne ◽  
...  
Keyword(s):  
Computed Tomography ◽  
At Risk ◽  
Deep Learning ◽  
Clinical Practice ◽  
Head And Neck ◽  
Organs At Risk ◽  
Data Set ◽  
Computed Tomography Scans ◽  
Neck Anatomy ◽  
Organ At Risk

BACKGROUND Over half a million individuals are diagnosed with head and neck cancer each year globally. Radiotherapy is an important curative treatment for this disease, but it requires manual time to delineate radiosensitive organs at risk. This planning process can delay treatment while also introducing interoperator variability, resulting in downstream radiation dose differences. Although auto-segmentation algorithms offer a potentially time-saving solution, the challenges in defining, quantifying, and achieving expert performance remain. OBJECTIVE Adopting a deep learning approach, we aim to demonstrate a 3D U-Net architecture that achieves expert-level performance in delineating 21 distinct head and neck organs at risk commonly segmented in clinical practice. METHODS The model was trained on a data set of 663 deidentified computed tomography scans acquired in routine clinical practice and with both segmentations taken from clinical practice and segmentations created by experienced radiographers as part of this research, all in accordance with consensus organ at risk definitions. RESULTS We demonstrated the model’s clinical applicability by assessing its performance on a test set of 21 computed tomography scans from clinical practice, each with 21 organs at risk segmented by 2 independent experts. We also introduced surface Dice similarity coefficient, a new metric for the comparison of organ delineation, to quantify the deviation between organ at risk surface contours rather than volumes, better reflecting the clinical task of correcting errors in automated organ segmentations. The model’s generalizability was then demonstrated on 2 distinct open-source data sets, reflecting different centers and countries to model training. CONCLUSIONS Deep learning is an effective and clinically applicable technique for the segmentation of the head and neck anatomy for radiotherapy. With appropriate validation studies and regulatory approvals, this system could improve the efficiency, consistency, and safety of radiotherapy pathways.

scholarly journals Correlation of age, sex and season with the state of human decomposition as quantified by postmortem computed tomography

2021 ◽  
Author(s):  
Dominic L. C. Guebelin ◽  
Akos Dobay ◽  
Lars Ebert ◽  
Eva Betschart ◽  
Michael J. Thali ◽  
...  
Keyword(s):  
Computed Tomography ◽  
Demographic Data ◽  
Sociodemographic Factors ◽  
The State ◽  
The Body ◽  
Seasonal Effects ◽  
Whole Body ◽  
Postmortem Computed Tomography ◽  
Whole Body Computed Tomography ◽  
Urban Study

AbstractDead bodies exhibit a variable range of changes with advancing decomposition. To quantify intracorporeal gas, the radiological alteration index (RAI) has been implemented in the assessment of postmortem whole-body computed tomography. We used this RAI as a proxy for the state of decomposition. This study aimed to (I) investigate the correlation between the state of decomposition and the season in which the body was discovered; and (II) evaluate the correlations between sociodemographic factors (age, sex) and the state of decomposition, by using the RAI as a proxy for the extent of decomposition. In a retrospective study, we analyzed demographic data from all autopsy reports from the Institute of Forensic Medicine of Zurich between January 2017 to July 2019 and evaluated the radiological alteration index from postmortem whole-body computed tomography for each case. The bodies of older males showed the highest RAI. Seasonal effects had no significant influence on the RAI in our urban study population with bodies mostly being discovered indoors. Autopsy reports contain valuable data that allow interpretation for reasons beyond forensic purposes, such as sociopolitical observations.

Deep Learning-Based Delineation of Head and Neck Organs at Risk: Geometric and Dosimetric Evaluation

2019 ◽  
Vol 104 (3) ◽  
pp. 677-684 ◽  
Author(s):  
Ward van Rooij ◽  
Max Dahele ◽  
Hugo Ribeiro Brandao ◽  
Alexander R. Delaney ◽  
Berend J. Slotman ◽  
...  
Keyword(s):  
At Risk ◽  
Deep Learning ◽  
Head And Neck ◽  
Organs At Risk
Sign in / Sign up

Export Citation Format

Share Document