Medical image analysis remains a challenging application area for artificial intelligence. When applying machine learning, obtaining ground-truth labels for supervised learning is more difficult than ...in many more common applications of machine learning. This is especially so for datasets with abnormalities, as tissue types and the shapes of the organs in these datasets differ widely. However, organ detection in such an abnormal dataset may have many promising potential real-world applications, such as automatic diagnosis, automated radiotherapy planning, and medical image retrieval, where new multimodal medical images provide more information about the imaged tissues for diagnosis. Here, we test the application of deep learning methods to organ identification in magnetic resonance medical images, with visual and temporal hierarchical features learned to categorize object classes from an unlabeled multimodal DCE-MRI dataset so that only a weakly supervised training is required for a classifier. A probabilistic patch-based method was employed for multiple organ detection, with the features learned from the deep learning model. This shows the potential of the deep learning model for application to medical images, despite the difficulty of obtaining libraries of correctly labeled training datasets and despite the intrinsic abnormalities present in patient datasets.
The combination of MRI and radiotherapy on a single platform has the potential to revolutionise image-guided delivery of radiation doses. However, in order to realise these ambitions, good dosimetry ...must be available. The electron return effect gives rise to significant perturbations of dose at interfaces between tissue and air within the body, and this might lead to difficulties in dose compensation if air cavities move during treatment. In this article, I review briefly the ways in which the available methods of dosimetry are affected by the presence of magnetic fields and discuss the contribution that three-dimensional measurements can make to studies in this area. The methods of MRI and optical computed tomography have well known issues in imaging close to interfaces. These are described together with progress so far in providing solutions.
3-D dosimetry readout techniques Doran, Simon J
Journal of physics. Conference series,
08/2019, Letnik:
1305, Številka:
1
Journal Article
Recenzirano
Odprti dostop
This invited technical review will discuss the numerous options available for making 3-D measurements of radiation dose, including both the physical principles underlying them and the potential ...sources of error involved.
The misfolding and subsequent aggregation of amyloidogenic proteins is a classic pathological hallmark of neurodegenerative diseases. Aggregates of the α-synuclein protein (αS) are implicated in ...Parkinson’s disease (PD) pathogenesis, and naturally occurring autoantibodies to these aggregates are proposed to be potential early-stage biomarkers to facilitate the diagnosis of PD. However, upon misfolding, αS forms a multitude of quaternary structures of varying functions that are unstable ex vivo. Thus, when used as a capture agent in enzyme-linked immunosorbent assays (ELISAs), significant variance among laboratories has prevented the development of these valuable diagnostic tests. We reasoned that those conflicting results arise due to the high nonspecific binding and amyloid nucleation that are typical of ELISA platforms. In this work, we describe a multiplexed, easy-to-operate immunoassay that is generally applicable to quantify the levels of amyloid proteins and their binding partners, named Oxaziridine-Assisted Solid-phase Immunosorbent (OASIS) assay. The assay is built on a hydrophilic poly(ethylene glycol) scaffold that inhibits aggregate nucleation, which we show reduces assay variance when compared to similar ELISA measurements. To validate our OASIS assay in patient-derived samples, we measured the levels of naturally occurring antibodies against the αS monomer and oligomers in a cohort of donor plasma from patients diagnosed with PD. Using OASIS assays, we observed significantly higher titers of immunoglobulin G antibody recognizing αS oligomers in PD patients compared to those in healthy controls, while there was no significant difference in naturally occurring antibodies against the αS monomer. In addition to its development into a blood test to potentially predict or monitor PD, we anticipate that the OASIS assay will be of high utility for studies aimed at understanding protein misfolding, its pathology and symptomology in PD, and other neurodegenerative diseases.
Textural features extracted from MRI potentially provide prognostic information additional to volume for influencing surgical management of cervical cancer.
To identify textural features that differ ...between cervical tumors above and below the volume threshold of eligibility for trachelectomy and determine their value in predicting recurrence in patients with low-volume tumors.
Of 378 patients with Stage1–2 cervical cancer imaged prospectively (3T, endovaginal coil), 125 had well-defined, histologically-confirmed squamous or adenocarcinomas with >100 voxels (>0.07 cm3) suitable for radiomic analysis. Regions-of-interest outlined the whole tumor on T2-W images and apparent diffusion coefficient (ADC) maps. Textural features based on grey-level co-occurrence matrices were compared (Mann-Whitney test with Bonferroni correction) between tumors greater (n = 46) or less (n = 79) than 4.19 cm3. Clustering eliminated correlated variables. Significantly different features were used to predict recurrence (regression modelling) in surgically-treated patients with low-volume tumors and compared with a model using clinico-pathological features.
Textural features (Dissimilarity, Energy, ClusterProminence, ClusterShade, InverseVariance, Autocorrelation) in 6 of 10 clusters from T2-W and ADC data differed between high-volume (mean ± SD 15.3 ± 11.7 cm3) and low-volume (mean ± SD 1.3 ± 1.2 cm3) tumors. (p < 0.02). In low-volume tumors, predicting recurrence was indicated by: Dissimilarity, Energy (ADC-radiomics, AUC = 0.864); Dissimilarity, ClusterProminence, InverseVariance (T2-W-radiomics, AUC = 0.808); Volume, Depth of Invasion, LymphoVascular Space Invasion (clinico-pathological features, AUC = 0.794). Combining ADC-radiomic (but not T2-radiomic) and clinico-pathological features improved prediction of recurrence compared to the clinico-pathological model (AUC = 0.916, p = 0.006). Findings were supported by bootstrap re-sampling (n = 1000).
Textural features from ADC maps and T2-W images differ between high- and low-volume tumors and potentially predict recurrence in low-volume tumors.
•Texture features differed significantly between high-compared to low-volume cervical tumors (p < 0.02).•In low-volume tumors predicting recurrence from ADC-radiomics was superior to T2W-radiomics or clinico-pathologic features.•Combining ADC-radiomics and clinico-pathologic features together improved recurrence prediction further.
Over recent decades, modern protocols of external beam radiotherapy have been developed that involve very steep dose gradients and are thus extremely sensitive to errors in treatment delivery. A ...recent credentialling study by the Radiological Physics Center at the MD Anderson Cancer Center (Texas, USA) has noted potentially significant inaccuracies in test treatments at a variety of institutions. 3-D radiation dosimetry (often referred to as “gel dosimetry”) may have an important role in commissioning new treatment protocols, to help prevent this type of error. This article discusses the various techniques of 3-D radiation dosimetry, with a focus on the types of radiosensitive samples used and on the optical computed tomography readout technique.
Dosimetric quality assurance (QA) of the new Elekta Unity (MR-linac) will differ from the QA performed of a conventional linac due to the constant magnetic field, which creates an electron return ...effect (ERE). In this work we aim to validate PRESAGE® dosimetry in a transverse magnetic field, and assess its use to validate the research version of the Monaco TPS of the MR-linac. Cylindrical samples of PRESAGE® 3D dosimeter separated by an air gap were irradiated with a cobalt-60 unit, while placed between the poles of an electromagnet at 0.5 T and 1.5 T. This set-up was simulated in EGSnrc/Cavity Monte Carlo (MC) code and relative dose distributions were compared with measurements using 1D and 2D gamma criteria of 3% and 1.5 mm. The irradiation conditions were adapted for the MR-linac and compared with Monaco TPS simulations. Measured and EGSnrc/Cavity simulated profiles showed good agreement with a gamma passing rate of 99.9% for 0.5 T and 99.8% for 1.5 T. Measurements on the MR-linac also compared well with Monaco TPS simulations, with a gamma passing rate of 98.4% at 1.5 T. Results demonstrated that PRESAGE® can accurately measure dose and detect the ERE, encouraging its use as a QA tool to validate the Monaco TPS of the MR-linac for clinically relevant dose distributions at tissue-air boundaries.