Search

Machine learning black boxes, exemplified by deep neural networks, often exhibit challenges in interpretability due to their reliance on complicated relationships involving numerous internal parameters and input features. This lack of transparency from a human perspective renders their predictions untrustworthy, particularly in critical applications. In this paper, we address this issue by introducing the design and implementation of CReEPy, an algorithm for symbolic knowledge extraction based on explainable clustering. Specifically, CReEPy leverages the underlying clustering performed by the ExACT or CREAM algorithms to generate human-interpretable Prolog rules that mimic the behaviour of opaque models. Additionally, we introduce CRASH, an algorithm for the automated tuning of hyper-parameters required by CReEPy. We present experiments evaluating both the human readability and predictive performance of the proposed knowledge-extraction algorithm, employing existing state-of-the-art techniques as benchmarks for comparison in real-world applications. [ABSTRACT FROM AUTHOR]

The XAI community is currently studying and developing symbolic knowledge-extraction (SKE) algorithms as a means to produce human-intelligible explanations for black-box machine learning predictors, so as to achieve believability in human-machine interaction. However, many extraction procedures exist in the literature, and choosing the most adequate one is increasingly cumbersome, as novel methods keep on emerging. Challenges arise from the fact that SKE algorithms are commonly defined based on theoretical assumptions that typically hinder practical applicability. This paper focuses on hypercube-based SKE methods, a quite general class of extraction techniques mostly devoted to regression-specific tasks. We first show that hypercube-based methods are flexible enough to support classification problems as well, then we propose a general model for them, and discuss how they support SKE on datasets, predictors, or learning tasks of any sort. Empirical examples are reported as well –based upon the PSyKE framework –, showing the applicability of hypercube-based methods to actual classification tasks. [ABSTRACT FROM AUTHOR]

A common practice in modern explainable AI is to post-hoc explain black-box machine learning (ML) predictors – such as neural networks – by extracting symbolic knowledge out of them, in the form of either rule lists or decision trees. By acting as a surrogate model, the extracted knowledge aims at revealing the inner working of the black box, thus enabling its inspection, representation, and explanation. Various knowledge-extraction algorithms have been presented in the literature so far. Unfortunately, running implementations of most of them are currently either proofs of concept or unavailable. In any case, a unified, coherent software framework supporting them all – as well as their interchange, comparison, and exploitation in arbitrary ML workflows – is currently missing. Accordingly, in this paper we discuss the design of PSyKE, a platform providing general-purpose support to symbolic knowledge extraction from different sorts of black-box predictors via many extraction algorithms. Notably, PSyKE targets symbolic knowledge in logic form, allowing the extraction of first-order logic clauses. The extracted knowledge is thus both machine- and human-interpretable, and can be used as a starting point for further symbolic processing—e.g. automated reasoning. [ABSTRACT FROM AUTHOR]

Short-term variations (<1 month) of the galactic cosmic-ray (GCR) flux in the inner heliosphere are mainly associated with the passage of high-speed solar wind streams (HSS) and interplanetary (IP) counterparts of coronal mass ejections (ICMEs). Data gathered with a particle detector flown on board the ESA LISA Pathfinder (LPF) spacecraft, during the declining part of solar cycle 24 (2016 February–2017 July) around the Lagrange point L1, have allowed us to study the characteristics of recurrent cosmic-ray flux modulations above 70 MeV n−1. It is shown that the amplitude and evolution of individual modulations depend in a unique way on both IP plasma parameters and particle flux intensity before HSS and ICME transit. By comparing the LPF data with those gathered contemporaneously with the magnetic spectrometer experiment AMS-02 on board the International Space Station and with those of Earth's polar neutron monitors, the GCR flux modulation was studied at different energies during recurrent short-term variations. It is also aimed to set the near real-time particle observation requirements to disentangle the role of long- and short-term variations of the GCR flux to evaluate the performance of high-sensitivity instruments in space such as the future interferometers for gravitational wave detection. Finally, the association between recurrent GCR flux variation observations in L1 and weak to moderate geomagnetic activity in 2016–2017 is discussed. Short-term recurrent GCR flux variations are good proxies of recurrent geomagnetic activity when the Bz component of the IP magnetic field is directed north. [ABSTRACT FROM AUTHOR]

This paper presents a comprehensive study of hydrogenated amorphous silicon (a-Si)-based detectors, utilizing electrical characterization, Raman spectroscopy, photoemission, and inverse photoemission techniques. The unique properties of a-Si have sparked interest in its application for radiation detection in both physics and medicine. Although amorphous silicon (a-Si) is inherently a highly defective material, hydrogenation significantly reduces defect density, enabling its use in radiation detector devices. Spectroscopic measurements provide insights into the intricate relationship between the structure and electronic properties of a-Si, enhancing our understanding of how specific configurations, such as the choice of substrate, can markedly influence detector performance. In this study, we compare the performance of a-Si detectors deposited on two different substrates: crystalline silicon (c-Si) and flexible Kapton. Our findings suggest that detectors deposited on Kapton exhibit reduced sensitivity, despite having comparable noise and leakage current levels to those on crystalline silicon. We hypothesize that this discrepancy may be attributed to the substrate material, differences in film morphology, and/or the alignment of energy levels. Further measurements are planned to substantiate these hypotheses.

Objective. Detectors that can provide accurate dosimetry for microbeam radiation therapy (MRT) must possess intrinsic radiation hardness, a high dynamic range, and a micron-scale spatial resolution. In this work we characterize hydrogenated amorphous silicon detectors for MRT dosimetry, presenting a novel combination of flexible, ultra-thin and radiation-hard features. Approach. Two detectors are explored: an n-type/intrinsic/p-type planar diode (NIP) and an NIP with an additional charge selective layer (NIP + CSC). Results. The sensitivity of the NIP + CSC detector was greater than the NIP detector for all measurement conditions. At 1 V and 0 kGy under the 3T Cu-Cu synchrotron broadbeam, the NIP + CSC detector sensitivity of (7.76 ± 0.01) pC cGy -1 outperformed the NIP detector sensitivity of (3.55 ± 0.23) pC cGy -1 by 219%. The energy dependence of both detectors matches closely to the attenuation coefficient ratio of silicon against water. Radiation damage measurements of both detectors out to 40 kGy revealed a higher radiation tolerance in the NIP detector compared to the NIP + CSC (17.2% and 33.5% degradations, respectively). Percentage depth dose profiles matched the PTW microDiamond detector's performance to within ±6% for all beam filtrations except in 3T Al-Al due to energy dependence. The 3T Cu-Cu microbeam field profile was reconstructed and returned microbeam width and peak-to-peak values of (51 ± 1) μ m and (405 ± 5) μ m, respectively. The peak-to-valley dose ratio was measured as a function of depth and agrees within error to the values obtained with the PTW microDiamond. X-ray beam induced charge mapping of the detector revealed minimal dose perturbations from extra-cameral materials. Significance. The detectors are comparable to commercially available dosimeters for quality assurance in MRT. With added benefits of being micron-sized and possessing a flexible water-equivalent substrate, these detectors are attractive candidates for quality assurance, in-vivo dosimetry and in-line beam monitoring for MRT and FLASH therapy., (Creative Commons Attribution license.)

Background: The increasing use of complex and high dose-rate treatments in radiation therapy necessitates advanced detectors to provide accurate dosimetry. Rather than relying on pre-treatment quality assurance (QA) measurements alone, many countries are now mandating the use of in vivo dosimetry, whereby a dosimeter is placed on the surface of the patient during treatment. Ideally, in vivo detectors should be flexible to conform to a patient's irregular surfaces., Purpose: This study aims to characterize a novel hydrogenated amorphous silicon (a-Si:H) radiation detector for the dosimetry of therapeutic x-ray beams. The detectors are flexible as they are fabricated directly on a flexible polyimide (Kapton) substrate., Methods: The potential of this technology for application as a real-time flexible detector is investigated through a combined dosimetric and flexibility study. Measurements of fundamental dosimetric quantities were obtained including output factor (OF), dose rate dependence (DPP), energy dependence, percentage depth dose (PDD), and angular dependence. The response of the a-Si:H detectors investigated in this study are benchmarked directly against commercially available ionization chambers and solid-state diodes currently employed for QA practices., Results: The a-Si:H detectors exhibit remarkable dose linearities in the direct detection of kV and MV therapeutic x-rays, with calibrated sensitivities ranging from (0.580 ± 0.002) pC/cGy to (19.36 ± 0.10) pC/cGy as a function of detector thickness, area, and applied bias. Regarding dosimetry, the a-Si:H detectors accurately obtained OF measurements that parallel commercially available detector solutions. The PDD response closely matched the expected profile as predicted via Geant4 simulations, a PTW Farmer ionization chamber and a PTW ROOS chamber. The most significant variation in the PDD performance was 5.67%, observed at a depth of 3 mm for detectors operated unbiased. With an external bias, the discrepancy in PDD response from reference data was confined to ± 2.92% for all depths (surface to 250 mm) in water-equivalent plastic. Very little angular dependence is displayed between irradiations at angles of 0° and 180°, with the most significant variation being a 7.71% decrease in collected charge at a 110° relative angle of incidence. Energy dependence and dose per pulse dependence are also reported, with results in agreement with the literature. Most notably, the flexibility of a-Si:H detectors was quantified for sample bending up to a radius of curvature of 7.98 mm, where the recorded photosensitivity degraded by (-4.9 ± 0.6)% of the initial device response when flat. It is essential to mention that this small bending radius is unlikely during in vivo patient dosimetry. In a more realistic scenario, with a bending radius of 15-20 mm, the variation in detector response remained within ± 4%. After substantial bending, the detector's photosensitivity when returned to a flat condition was (99.1 ± 0.5)% of the original response., Conclusions: This work successfully characterizes a flexible detector based on thin-film a-Si:H deposited on a Kapton substrate for applications in therapeutic x-ray dosimetry. The detectors exhibit dosimetric performances that parallel commercially available dosimeters, while also demonstrating excellent flexibility results., (© 2024 The Authors. Medical Physics published by Wiley Periodicals LLC on behalf of American Association of Physicists in Medicine.)

Objective . Microbeam radiation therapy (MRT) is an alternative emerging radiotherapy treatment modality which has demonstrated effective radioresistant tumour control while sparing surrounding healthy tissue in preclinical trials. This apparent selectivity is achieved through MRT combining ultra-high dose rates with micron-scale spatial fractionation of the delivered x-ray treatment field. Quality assurance dosimetry for MRT must therefore overcome a significant challenge, as detectors require both a high dynamic range and a high spatial resolution to perform accurately. Approach . In this work, a series of radiation hard a-Si:H diodes, with different thicknesses and carrier selective contact configurations, have been characterised for x-ray dosimetry and real-time beam monitoring applications in extremely high flux beamlines utilised for MRT at the Australian Synchrotron. Results . These devices displayed superior radiation hardness under constant high dose-rate irradiations on the order of 6000 Gy s -1 , with a variation in response of 10% over a delivered dose range of approximately 600 kGy. Dose linearity of each detector to x-rays with a peak energy of 117 keV is reported, with sensitivities ranging from (2.74 ± 0.02) nC/Gy to (4.96 ± 0.02) nC/Gy. For detectors with 0.8 μ m thick active a-Si:H layer, their operation in an edge-on orientation allows for the reconstruction of micron-size beam profiles (microbeams). The microbeams, with a nominal full-width-half-max of 50 μ m and a peak-to-peak separation of 400 μ m, were reconstructed with extreme accuracy. The full-width-half-max was observed as 55 ± 1 μ m. Evaluation of the peak-to-valley dose ratio and dose-rate dependence of the devices, as well as an x-ray induced charge (XBIC) map of a single pixel is also reported. Significance . These devices based on novel a-Si:H technology possess a unique combination of accurate dosimetric performance and radiation resistance, making them an ideal candidate for x-ray dosimetry in high dose-rate environments such as FLASH and MRT., (Creative Commons Attribution license.)