Artificial intelligence in radiotherapy: Current applications and future trends.

Radiation therapy has dramatically changed with the advent of computed tomography and intensity modulation. This added complexity to the workflow but allowed for more precise and reproducible treatment. As a result, these advances required the accurate delineation of many more volumes, raising questions about how to delineate them, in a uniform manner across centers. Then, as computing power improved, reverse planning became possible and three-dimensional dose distributions could be generated. Artificial intelligence offers the opportunity to make such workflow more efficient while increasing practice homogeneity. Many artificial intelligence-based tools are being implemented in routine practice to increase efficiency, reduce workload and improve homogeneity of treatments. Data retrieved from this workflow could be combined with clinical data and omic data to develop predictive tools to support clinical decision-making process. Such predictive tools are at the stage of proof-of-concept and need to be explainatory, prospectively validated, and based on large and multicenter cohorts. Nevertheless, they could bridge the gap to personalized radiation oncology, by personalizing oncologic strategies, dose prescriptions to tumor volumes and dose constraints to organs at risk.

Deep learning for automated segmentation in radiotherapy: a narrative review.

The segmentation of organs and structures is a critical component of radiation therapy planning, with manual segmentation being a laborious and time-consuming task. Interobserver variability can also impact the outcomes of radiation therapy. Deep neural networks have recently gained attention for their ability to automate segmentation tasks, with convolutional neural networks (CNNs) being a popular approach. This article provides a descriptive review of the literature on deep learning (DL) techniques for segmentation in radiation therapy planning. This review focuses on five clinical sub-sites and finds that U-net is the most commonly used CNN architecture. The studies using DL for image segmentation were included in brain, head and neck, lung, abdominal, and pelvic cancers. The majority of DL segmentation articles in radiation therapy planning have concentrated on normal tissue structures. N-fold cross-validation was commonly employed, without external validation. This research area is expanding quickly, and standardization of metrics and independent validation are critical to benchmarking and comparing proposed methods.

A Radiomics Approach to Identify Immunologically Active Tumor in Patients with Head and Neck Squamous Cell Carcinomas.

BACKGROUND: We recently developed a gene-expression-based HOT score to identify the hot/cold phenotype of head and neck squamous cell carcinomas (HNSCCs), which is associated with the response to immunotherapy. Our goal was to determine whether radiomic profiling from computed tomography (CT) scans can distinguish hot and cold HNSCC. METHOD: We included 113 patients from The Cancer Genome Atlas (TCGA) and 20 patients from the Groupe Hospitalier Pitié-Salpêtrière (GHPS) with HNSCC, all with available pre-treatment CT scans. The hot/cold phenotype was computed for all patients using the HOT score. The IBEX software (version 4.11.9, accessed on 30 march 2020) was used to extract radiomic features from the delineated tumor region in both datasets, and the intraclass correlation coefficient (ICC) was computed to select robust features. Machine learning classifier models were trained and tested in the TCGA dataset and validated using the area under the receiver operator characteristic curve (AUC) in the GHPS cohort. RESULTS: A total of 144 radiomic features with an ICC >0.9 was selected. An XGBoost model including these selected features showed the best performance prediction of the hot/cold phenotype with AUC = 0.86 in the GHPS validation dataset. CONCLUSIONS AND RELEVANCE: We identified a relevant radiomic model to capture the overall hot/cold phenotype of HNSCC. This non-invasive approach could help with the identification of patients with HNSCC who may benefit from immunotherapy.

Impact of Neck Dissection in Head and Neck Squamous Cell Carcinomas of Unknown Primary.

PURPOSE: Management of head and neck cancers of unknown primary (HNCUP) combines neck dissection (ND) and radiotherapy, with or without chemotherapy. The prognostic value of ND has hardly been studied in HNCUP. METHODS: A retrospective multicentric study assessed the impact of ND extent (adenectomy, selective ND, radical/radical-modified ND) on nodal relapse, progression-free survival (PFS) or survival, taking into account nodal stage. RESULTS: 53 patients (16.5%) had no ND, 33 (10.2%) had lymphadenectomy, 116 (36.0%) underwent selective ND and 120 underwent radical/radical-modified ND (37.3%), 15 of which received radical ND (4.7%). With a 34-month median follow-up, the 3-year incidence of nodal relapse was 12.5% and progression-free survival (PFS) 69.1%. In multivariate analysis after adjusting for nodal stage, the risk of nodal relapse or progression was reduced with lymphadenectomy, selective or radical/modified ND, but survival rates were similar. Patients undergoing lymphadenectomy or ND had a better PFS and lowered nodal relapse incidence in the N1 + N2a group, but the improvement was not significant for the N2b or N2 + N3c patients. Severe toxicity rates exceeded 40% with radical ND. CONCLUSION: In HNCUP, ND improves PFS, regardless of nodal stage. The magnitude of the benefit of ND does not appear to depend on ND extent and decreases with a more advanced nodal stage.

The Role of Radiomics in Lung Cancer: From Screening to Treatment and Follow-Up.

PURPOSE: Lung cancer represents the first cause of cancer-related death in the world. Radiomics studies arise rapidly in this late decade. The aim of this review is to identify important recent publications to be synthesized into a comprehensive review of the current status of radiomics in lung cancer at each step of the patients' care. METHODS: A literature review was conducted using PubMed/Medline for search of relevant peer-reviewed publications from January 2012 to June 2020. RESULTS: We identified several studies at each point of patient's care: detection and classification of lung nodules (n=16), determination of histology and genomic (n=10) and finally treatment outcomes predictions (=23). We reported the methodology of those studies and their results and discuss the limitations and the progress to be made for clinical routine applications. CONCLUSION: Promising perspectives arise from machine learning applications and radiomics based models in lung cancers, yet further data are necessary for their implementation in daily care. Multicentric collaboration and attention to quality and reproductivity of radiomics studies should be further consider.

Investigation of Relation of Radiation Therapy Quality With Toxicity and Survival in LAP07 Phase 3 Trial for Locally Advanced Pancreatic Carcinoma.

PURPOSE: The LAP07 multicenter randomized study assessed whether chemoradiation therapy increases overall survival versus continuation chemotherapy in patients whose locally advanced pancreatic cancer was controlled after 4 months of induction chemotherapy. This analysis investigated whether failure to adhere to radiation therapy (RT) guidelines influenced survival and toxicity. METHODS AND MATERIALS: This is a planned analysis of secondary objectives in the framework of a randomized international phase 3 trial. The protocol included detailed written RT guidelines. All participating institutions undertook an initial benchmark case to check adherence to protocol guidelines. Centers with major deviation were not allowed to include patients until they achieved a significant improvement and rigorously followed the guidelines. On-trial RT quality assurance consisted of a central review of treatment plan with dose-volume histograms for each patient. Adherence to guidelines was graded as per protocol (PP), minor deviation (MiD), or major deviation (MaD). RESULTS: Fifty-seven benchmark cases were evaluated, 26% were classified as PP, 60% were MiD, and 14% were MaD. Among the 442 included patients, 133 patients were randomized in the chemoradiation therapy arm, and 117 patients were assessable for RT quality analysis. RT quality was graded as PP in 38.5% of patients, MiD in 43.6% of patients, and MaD in 17.9% of patients. The most frequent protocol violations were dose distribution heterogeneities. Median overall survival was 17 months with PP and MiD versus 13.4 months with MaD (hazard ratio [HR], 1.63; 95% confidence interval [CI], 0.99-2.71; P = .055). There was no difference in terms of progression-free survival (HR, 1.09; 95% CI, 0.66-1.8; P = .72). Patients with MaD had more nausea than patients treated PP or with MiD (P = .0045). CONCLUSIONS: MaD was associated with a trend for worst survival. There was no difference in terms of progression-free survival. Because of the low rate of major deviations, their effects on the LAP07 trial results may be negligeable.

Interpretable Machine Learning Model for Locoregional Relapse Prediction in Oropharyngeal Cancers.

BACKGROUND: There is no evidence to support surgery or radiotherapy as the best treatment for resectable oropharyngeal cancers with a negative HPV status. Predictive algorithms may help to decide which strategy to choose, but they will only be accepted by caregivers and European authorities if they are interpretable. As a proof of concept, we developed a predictive and interpretable algorithm to predict locoregional relapse at 18 months for oropharyngeal cancers as a first step towards that goal. METHODS: The model was based on clinical and Pyradiomics features extracted from the dosimetric CT scan. Intraclass correlation was used to filter out features dependant on delineation. Correlated redundant features were also removed. An XGBoost model was cross-validated and optimised on the HN1 cohort (79 patients), and performances were assessed on the ART ORL cohort (45 patients). The Shapley Values were used to provide an overall and local explanation of the model. RESULTS: On the ART ORL cohort, the model trained on HN1 yielded a precision-or predictive positive value-of 0.92, a recall of 0.42, an area under the curve of the receiver operating characteristic of 0.68 and an accuracy of 0.64. The most contributory features were shape Voxel Volume, grey level size zone matrix Small Area Emphasis (glszmSAE), gldm Dependence Non Uniformity Normalized (gldmDNUN), Sex and Age. CONCLUSIONS: We developed an interpretable and generalizable model that could yield a good precision-positive predictive value-for relapse at 18 months on a different test cohort.

Radiomics and Machine Learning for Radiotherapy in Head and Neck Cancers.

Introduction: An increasing number of parameters can be considered when making decisions in oncology. Tumor characteristics can also be extracted from imaging through the use of radiomics and add to this wealth of clinical data. Machine learning can encompass these parameters and thus enhance clinical decision as well as radiotherapy workflow. Methods: We performed a description of machine learning applications at each step of treatment by radiotherapy in head and neck cancers. We then performed a systematic review on radiomics and machine learning outcome prediction models in head and neck cancers. Results: Machine Learning has several promising applications in treatment planning with automatic organ at risk delineation improvements and adaptative radiotherapy workflow automation. It may also provide new approaches for Normal Tissue Complication Probability models. Radiomics may provide additional data on tumors for improved machine learning powered predictive models, not only on survival, but also on risk of distant metastasis, in field recurrence, HPV status and extra nodal spread. However, most studies provide preliminary data requiring further validation. Conclusion: Promising perspectives arise from machine learning applications and radiomics based models, yet further data are necessary for their implementation in daily care.

Comparison of 3D conformal radiation therapy and intensity-modulated radiation therapy in patients with endometrial cancer: efficacy, safety and prognostic analysis.

Introduction: Adjuvant whole-pelvic radiation therapy (WPRT) improves locoregional control for high-intermediate stages I-III endometrial cancer patients. Intensity modulated radiation therapy (IMRT) tends to replace the standard 3D conformal radiation therapy (3DCRT) technique used in trials. Material and methods: Consecutive patients with stages I-IIIc endometrial cancer treated between 2008 and 2014 in our department with post-operative 3DCRT or IMRT WPRT were studied retrospectively. Patients with cervical involvement underwent additional low-dose rate vaginal brachytherapy. The impact of the WPRT technique on local control, tolerance, disease-free survival (DFS) and overall survival (OS) was assessed. Clinicians evaluated routinely acute radiation toxicity each week during radiation therapy and late toxicity during standard follow-up consultations. Results: Median follow-up was 50 months (range: 6-158). Among the 83 patients included, 47 were treated with 3DCRT and 36 with IMRT. There was no difference in patient characteristics between groups. The 5-year locoregional control and DFS rates were 94.5% and 68%, respectively. No significant difference was found between the 3DCRT and IMRT groups in terms of survival, with 5-year OS rates of 74.6% and 78%, respectively. In multivariate analysis, age over 68, stage > T1 and grade 3 were independently associated with shorter DFS and OS. Seven patients (8.4%) had grades 3-4 acute gastrointestinal (GI) toxicity with five patients (10.6%) and two (5.4%) in the 3DCRT and IMRT groups, respectively (p = .69). One case (1.2%) of late grade 3 GI toxicity was observed treated in 3DCRT. Conclusions: IMRT seems to be a safe technique for the treatment of endometrial cancer with a trend towards decreased acute GI toxicities. Results of the phase 3 RTOG 1203 trial are needed to confirm these results.

Unilateral or bilateral irradiation in cervical lymph node metastases of unknown primary? A retrospective cohort study.

INTRODUCTION: Patients with cervical lymphadenopathy of unknown primary carcinoma (CUP) usually undergo neck dissection and irradiation. There is an ongoing controversy regarding the extent of nodal and mucosal volumes to be irradiated. We assessed outcomes after bilateral or unilateral nodal irradiation. METHODS: This retrospective multicentre study included patients with CUP and squamous cellular carcinoma who underwent radiotherapy (RT) between 2000 and 2015. RESULTS: Of 350 patients, 74.5% had unilateral disease and 25.5% had bilateral disease. Of 297 patients with available data on disease and irradiation sides, 61 (20.5%) patients had unilateral disease and unilateral irradiation, 155 (52.2%), unilateral disease and bilateral irradiation and 81 (27.3%), bilateral disease and bilateral irradiation. Thirty-four (9.7%) and 217 (62.0%) patients received neoadjuvant and/or concomitant chemotherapy, respectively. Median follow-up was 37 months. Three-year local, regional, locoregional failure rates and CUP-specific survival were 5.6%, 11.7%, 15.0% and 84.7%, respectively. In patients with unilateral disease, the 3-year cumulative incidence of regional/local relapse was 7.7%/4.3% after bilateral irradiation versus 16.9%/11.1% after unilateral irradiation (hazard ratio = 0.56/0.61, p = 0.17/0.32). The cumulative incidence of CUP-specific deaths was 9.2% after bilateral irradiation and 15.5% after unilateral irradiation (p = 0.92). In multivariate analysis, mucosal irradiation was associated with better local control, whereas no neck dissection, ≥N2b and interruption of RT for more than 4 days were associated with poorer regional control. Toxicity was higher after bilateral irradiation (p < 0.05). No positron-emission tomography-computed tomography, largest node diameter, ≥N2b, neoadjuvant chemotherapy and interruption of RT were associated with poorer cause-specific survival. CONCLUSION: Bilateral nodal irradiation yielded non-significant better nodal and mucosal control rates but was associated with higher rates of severe toxicity.

Surface functionalization-induced photoresponse characteristics of monolayer MoS2 for fast flexible photodetectors.

Monolayered, semiconducting molybdenum disulfide (MoS2) is of considerable interest for its potential applications in next-generation flexible, wearable, and transparent photodetectors because it has outstanding physical properties coupled with unique atomically thin dimensions. However, there is still a lack of understanding in terms of the underlying mechanisms responsible for the photoresponse dynamics, which makes it difficult to identify the appropriate device design strategy for achieving a fast photoresponse time in MoS2 photodetectors. In this study, we investigate the importance of surface functionalization on controlling the charge carrier densities in a MoS2 monolayer and in turn the corresponding behavior of the photoresponse in relation to the position of the Fermi-level and the energy band structure. We find that the p-doping and n-doping, which is achieved through the surface functionalization of the MoS2 monolayer, leads to devices with different photoresponse behavior. Specifically, the MoS2 devices with surface functional groups contributing to p-doping exhibited a faster response time as well as higher sensitivity compared to that observed for the MoS2 devices with surface functional groups contributing to n-doping. We attribute this difference to the degree of bending in the energy bands at the metal-semiconductor junction as a result of shifting in the Fermi-level position, which influences the optoelectronic transport properties as well as the recombination dynamics leading to a low dark and thus high detectivity and fast decay time. Based upon these findings, we have also demonstrated the broad applicability of surface functionalization by fabricating a flexible MoS2 photodetector that shows an outstanding decay time of 0.7 s, which is the fastest response time observed in flexible MoS2 detectors ever reported.

Balancing Charge Carrier Transport in a Quantum Dot P-N Junction toward Hysteresis-Free High-Performance Solar Cells.

In a quantum dot solar cell (QDSC) that has an inverted structure, the QD layers form two different junctions between the electron transport layer (ETL) and the other semiconducting QD layer. Recent work on an inverted-structure QDSC has revealed that the junction between the QD layers is the dominant junction, rather than the junction between the ETL and the QD layers, which is in contrast to the conventional wisdom. However, to date, there have been a lack of systematic studies on the role and importance of the QD heterojunction structure on the behavior of the solar cell and the resulting device performance. In this study, we have systematically controlled the structure of the QD junction to balance charge transport, which demonstrates that the position of the junction has a significant effect on the hysteresis effect, fill factor, and solar cell performance and is attributed to balanced charge transport.

Field effect transistors and phototransistors based upon p-type solution-processed PbS nanowires.

We demonstrate the fabrication of solution processed highly crystalline p-type PbS nanowires via the oriented attachment of nanoparticles. The analysis of single nanowire field effect transistor (FET) devices revealed a hole conduction behaviour with average mobilities greater than 30 cm2 V-1 s-1, which is an order of magnitude higher than that reported to date for p-type PbS colloidal nanowires. We have investigated the response of the FETs to near-infrared light excitation and show herein that the nanowires exhibited gate-dependent photo-conductivities, enabling us to tune the device performances. The responsivity was found to be greater than 104 A W-1 together with a detectivity of 1013 Jones, which benefits from a photogating effect occurring at negative gate voltages. These encouraging detection parameters are accompanied by relatively short switching times of 15 ms at positive gate voltages, resulting from a combination of the standard photoconduction and the high crystallinity of the nanowires. Collectively, these results indicate that solution-processed PbS nanowires are promising nanomaterials for infrared photodetectors as well as p-type nanowire FETs.

Thermodynamically Stable Synthesis of Large-Scale and Highly Crystalline Transition Metal Dichalcogenide Monolayers and their Unipolar n-n Heterojunction Devices.

Transition metal dichalcogenide (TMDC) monolayers are considered to be potential materials for atomically thin electronics due to their unique electronic and optical properties. However, large-area and uniform growth of TMDC monolayers with large grain sizes is still a considerable challenge. This report presents a simple but effective approach for large-scale and highly crystalline molybdenum disulfide monolayers using a solution-processed precursor deposition. The low supersaturation level, triggered by the evaporation of an extremely thin precursor layer, reduces the nucleation density dramatically under a thermodynamically stable environment, yielding uniform and clean monolayer films and large crystal sizes up to 500 µm. As a result, the photoluminescence exhibits only a small full-width-half-maximum of 48 meV, comparable to that of exfoliated and suspended monolayer crystals. It is confirmed that this growth procedure can be extended to the synthesis of other TMDC monolayers, and robust MoS2 /WS2 heterojunction devices are easily prepared using this synthetic procedure due to the large-sized crystals. The heterojunction device shows a fast response time (≈45 ms) and a significantly high photoresponsivity (≈40 AW-1 ) because of the built-in potential and the majority-carrier transport at the n-n junction. These findings indicate an efficient pathway for the fabrication of high-performance 2D optoelectronic devices.

Monolayer optical memory cells based on artificial trap-mediated charge storage and release.

Monolayer transition metal dichalcogenides are considered to be promising candidates for flexible and transparent optoelectronics applications due to their direct bandgap and strong light-matter interactions. Although several monolayer-based photodetectors have been demonstrated, single-layered optical memory devices suitable for high-quality image sensing have received little attention. Here we report a concept for monolayer MoS2 optoelectronic memory devices using artificially-structured charge trap layers through the functionalization of the monolayer/dielectric interfaces, leading to localized electronic states that serve as a basis for electrically-induced charge trapping and optically-mediated charge release. Our devices exhibit excellent photo-responsive memory characteristics with a large linear dynamic range of ∼4,700 (73.4 dB) coupled with a low OFF-state current (<4 pA), and a long storage lifetime of over 104 s. In addition, the multi-level detection of up to 8 optical states is successfully demonstrated. These results represent a significant step toward the development of future monolayer optoelectronic memory devices.

Synergistic Effects of a Multifunctional Graphene Based Interlayer on Electrochemical Behavior and Structural Stability.

The ability to rationally design and manipulate the interfacial structure in lithium ion batteries (LIBs) is of utmost technological importance for achieving desired performance requirements as it provides synergistic effects to the electrochemical properties and cycling stability of electrode materials. However, despite considerable efforts and progress made in recent years through the interface engineering based on active electrode materials, relatively little attention has been devoted to address the physical aspects of the interface and interfacial layer between the anode materials layer and the current collector. Here, we propose and successfully grow unique graphene directly on a Cu current collector as an ideal interfacial layer using the modified chemical vapor deposition (CVD). The anode with an engineered graphene interlayer exhibits remarkably improved electrochemical performances, such as large reversible specific capacity (921.4 mAh g(-1) at current density of 200 mA g(-1)), excellent Coulombic efficiency (close to approximately 96%), and superior cycling capacity retention and rate properties compared to the bare Cu. These excellent electrochemical features are discussed in terms of multiple beneficial effects of graphene on interfacial stability and adhesion between the anode and the collector, oxidation or corrosion resistance of the graphene grown Cu current collector, and electrical contact conductance during the charge/discharge process.

Enhanced charge carrier transport properties in colloidal quantum dot solar cells via organic and inorganic hybrid surface passivation.

Colloidal quantum dots (CQDs) are extremely promising as photovoltaic materials. In particular, the tunability of their electronic band gap and cost effective synthetic procedures allow for the versatile fabrication of solar energy harvesting cells, resulting in optimal device performance. However, one of the main challenges in developing high performance quantum dot solar cells (QDSCs) is the improvement of the photo-generated charge transport and collection, which is mainly hindered by imperfect surface functionalization, such as the presence of surface electronic trap sites and the initial bulky surface ligands. Therefore, for these reasons, finding effective methods to efficiently decorate the surface of the as-prepared CQDs with new short molecular length chemical structures so as to enhance the performance of QDSCs is highly desirable. Here, we suggest employing hybrid halide ions along with the shortest heterocyclic molecule as a robust passivation structure to eliminate surface trap sites while decreasing the charge trapping dynamics and increasing the charge extraction efficiency in CQD active layers. This hybrid ligand treatment shows a better coordination with Pb atoms within the crystal, resulting in low trap sites and a near perfect removal of the pristine initial bulky ligands, thereby achieving better conductivity and film structure. Compared to halide ion-only treated cells, solar cells fabricated through this hybrid passivation method show an increase in the power conversion efficiency from 5.3% for the halide ion-treated cells to 6.8% for the hybrid-treated solar cells.