Artificial intelligence and machine learning in cancer imaging

Communications Medicine, Journal Year: 2022, Volume and Issue: 2(1)

Published: Oct. 27, 2022

An increasing array of tools is being developed using artificial intelligence (AI) and machine learning (ML) for cancer imaging. The development an optimal tool requires multidisciplinary engagement to ensure that the appropriate use case met, as well undertake robust testing prior its adoption into healthcare systems. This review highlights key developments in field. We discuss challenges opportunities AI ML imaging; considerations algorithms can be widely used disseminated; ecosystem needed promote growth

Language: Английский

---

Review of Deep Learning Based Automatic Segmentation for Lung Cancer Radiotherapy

Frontiers in Oncology, Journal Year: 2021, Volume and Issue: 11

Published: July 8, 2021

Lung cancer is the leading cause of cancer-related mortality for males and females. Radiation therapy (RT) one primary treatment modalities lung cancer. While delivering prescribed dose to tumor targets, it essential spare tissues near targets—the so-called organs-at-risk (OARs). An optimal RT planning benefits from accurate segmentation gross volume surrounding OARs. Manual a time-consuming tedious task radiation oncologists. Therefore, crucial develop automatic image relieve oncologists contouring work. Currently, atlas-based technique commonly used in clinical routines. However, this depends heavily on similarity between atlas segmented. With significant advances made computer vision, deep learning as part artificial intelligence attracts increasing attention medical segmentation. In article, we reviewed based techniques related compared them with technique. At present, auto-segmentation OARs relatively large such heart etc. outperforms organs small esophagus. The average Dice coefficient (DSC) lung, liver are over 0.9, best DSC spinal cord reaches 0.9. esophagus ranges 0.71 0.87 ragged performance. terms volume, below 0.8. Although indicate superiority many aspects manual segmentation, various issues still need be solved. We discussed potential including low contrast, dataset size, consensus guidelines, network design. Clinical limitations future research directions were well.

Language: Английский

---

Implementation of deep learning-based auto-segmentation for radiotherapy planning structures: a workflow study at two cancer centers

Radiation Oncology, Journal Year: 2021, Volume and Issue: 16(1)

Published: June 8, 2021

Abstract Purpose We recently described the validation of deep learning-based auto-segmented contour (DC) models for organs at risk (OAR) and clinical target volumes (CTV). In this study, we evaluate performance implemented DC in radiotherapy (RT) planning workflow report on user experience. Methods materials were two cancer centers used to generate OAR CTVs all patients undergoing RT a central nervous system (CNS), head neck (H&N), or prostate cancer. Radiation Therapists/Dosimetrists Oncologists completed post-contouring surveys rating degree edits required DCs (1 = minimal, 5 significant) overall satisfaction poor, high). Unedited compared edited treatment approved contours using Dice similarity coefficient (DSC) 95% Hausdorff distance (HD). Results Between September 19, 2019 March 6, 2020, generated approximately 551 eligible cases. 203 collected 27 CNS, 54 H&N, 93 plans, resulting an survey compliance rate 32%. The majority minimal subjectively (mean editing score ≤ 2) objectively DSC HD was ≥ 0.90 2.0 mm). Mean 4.1 4.4 4.6 structures. Overall CTV (n 25), which encompassed prostate, seminal vesicles, lymph node volumes, 4.1. Conclusions Previously validated subjective objective resulted positive experience, although low concern. model evaluation even more limited, but high suggests that they may have served as appropriate starting points patient specific edits.

Language: Английский

---

Image Segmentation Techniques: Statistical, Comprehensive, Semi-Automated Analysis and an Application Perspective Analysis of Mathematical Expressions

Archives of Computational Methods in Engineering, Journal Year: 2022, Volume and Issue: 30(1), P. 457 - 495

Published: Aug. 22, 2022

Language: Английский

---

A Review of the Metrics Used to Assess Auto-Contouring Systems in Radiotherapy

Clinical Oncology, Journal Year: 2023, Volume and Issue: 35(6), P. 354 - 369

Published: Jan. 31, 2023

Auto-contouring could revolutionise future planning of radiotherapy treatment. The lack consensus on how to assess and validate auto-contouring systems currently limits clinical use. This review formally quantifies the assessment metrics used in studies published during one calendar year assesses need for standardised practice. A PubMed literature search was undertaken papers evaluating 2021. Papers were assessed types metric methodology generate ground-truth comparators. Our identified 212 studies, which 117 met criteria review. Geometric 116 (99.1%). includes Dice Similarity Coefficient 113 (96.6%) studies. Clinically relevant metrics, such as qualitative, dosimetric time-saving less frequently 22 (18.8%), 27 (23.1%) 18 (15.4%) respectively. There heterogeneity within each category metric. Over 90 different names geometric measures used. Methods qualitative all but two papers. Variation existed methods plans assessment. Consideration editing time only given 11 (9.4%) single manual contour a comparator 65 (55.6%) Only 31 (26.5%) compared auto-contours usual inter- and/or intra-observer variation. In conclusion, significant variation exists research accuracy automatically generated contours. are most popular, however their utility is unknown. perform Considering stages system implementation may provide framework decide appropriate metrics. analysis supports auto-contouring.

Language: Английский

---

LLM-driven multimodal target volume contouring in radiation oncology

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: Oct. 24, 2024

Target volume contouring for radiation therapy is considered significantly more challenging than the normal organ segmentation tasks as it necessitates utilization of both image and text-based clinical information.Inspired by recent advancement large language models (LLMs) that can facilitate integration textural information images, here we present an LLM-driven multimodal artificial intelligence (AI), namely LLMSeg, utilizes applicable to task 3-dimensional context-aware target delineation oncology.We validate our proposed LLMSeg within context breast cancer radiotherapy using external validation data-insufficient environments, which attributes highly conducive real-world applications.We demonstrate exhibits markedly improved performance compared conventional unimodal AI models, particularly exhibiting robust generalization data-efficiency.

Language: Английский

---

Review of Deep Learning Based Autosegmentation for Clinical Target Volume: Current Status and Future Directions

Advances in Radiation Oncology, Journal Year: 2024, Volume and Issue: 9(5), P. 101470 - 101470

Published: Feb. 8, 2024

PurposeManual contour work for radiation treatment planning takes significant time to ensure volumes are accurately delineated. The use of artificial intelligence with deep learning based autosegmentation (DLAS) models has made itself known in recent years alleviate this workload. It is used organs at risk (OAR) contouring consistency performance and saving. purpose study was evaluate the current published data DLAS clinical target volume (CTV) contours, identify areas improvement, discuss future directions.MethodologyA literature review performed by utilizing key words "Deep Learning" AND ("Segmentation" OR "Delineation") "Clinical Target Volume" an indexed search into PubMed. A total 154 articles on criteria were reviewed. considered model used, disease site, targets contoured, guidelines utilized, overall performance.ResultsOf 53 investigating CTV, only 6 before 2020. Publications have increased years, 46 between 2020-2023. cervix (n=19) prostate (n=12) studied most frequently. Most studies (n=43) involved a single institution. Median sample size 130 patients (range: 5-1,052). common metrics utilized measure Dice similarity coefficient (DSC) followed Hausdorff distance. Dosimetric seldom reported (n=11). There also variability specific (RTOG, ESTRO, others). had good CTV multiple sites, showing DSC values >0.7. delineated faster compared manual contouring. However, some contours still required least minor edits, require improvement.ConclusionsDLAS demonstrates capability completing plans efficiency accuracy. developed validated institutions using developing institutions. about years. Future need include larger datasets different patient demographics, stages, validation multi-institutional settings, inclusion dosimetric performance. Manual directions. Of improvement.

Language: Английский

---

Dosimetric impact of deep learning-based CT auto-segmentation on radiation therapy treatment planning for prostate cancer

Radiation Oncology, Journal Year: 2022, Volume and Issue: 17(1)

Published: Jan. 31, 2022

The evaluation of automatic segmentation algorithms is commonly performed using geometric metrics. An analysis based on dosimetric parameters might be more relevant in clinical practice but often lacking the literature. aim this study was to investigate impact state-of-the-art 3D U-Net-generated organ delineations dose optimization radiation therapy (RT) for prostate cancer patients.A database 69 computed tomography images with prostate, bladder, and rectum used single-label U-Net training dice similarity coefficient (DSC)-based loss. Volumetric modulated arc (VMAT) plans have been generated both manual segmentations same settings. These were chosen give consistent when applying perturbations segmentations. Contours evaluated terms DSC, average 95% Hausdorff distance (HD). Dose distributions as reference volume histogram (DVH) a 3%/3 mm gamma-criterion 10% cut-off. A Pearson correlation between DSC metrics, i.e. gamma index DVH parameters, has calculated.3D U-Net-based achieved 0.87 (0.03) 0.97 (0.01) bladder 0.89 (0.04) rectum. mean HD below 1.6 (0.4) 5 (4) mm, respectively. V[Formula: see text] rectum, showed agreement within [Formula: text], D[Formula: its 3 expansion (surrogate target volume) distribution 2% Gy exception one case. pass-rate 85%. comparison metrics no strong statistically significant correlation.The developed work geometrical performance. Analysis clinically VMAT demonstrated neither excessive increase OARs nor substantial under/over-dosage all Yet indicated several cases low pass rates. highlighted importance adding standard evaluation.

Language: Английский

---

Training, validation, and clinical implementation of a deep-learning segmentation model for radiotherapy of loco-regional breast cancer

Radiotherapy and Oncology, Journal Year: 2022, Volume and Issue: 173, P. 62 - 68

Published: May 23, 2022

Language: Английский

---

First Report On Physician Assessment and Clinical Acceptability of Custom-Retrained Artificial Intelligence Models for Clinical Target Volume and Organs-at-Risk Auto-Delineation for Postprostatectomy Patients

Practical Radiation Oncology, Journal Year: 2023, Volume and Issue: 13(4), P. 351 - 362

Published: April 6, 2023

Language: Английский

---